Computer Systems
Documentation
A limited partnership since 1977 Tuesday January 1, 2002 17:04
Service 1 Wednesday November 13, 2002 09:42 Technical people 1 history added Thursday September 26, 2002 09:54 Critical links walmart is less expensive at
barnesandnoble
2
3
4 Thursday September 26, 2002 07:00 Stocks and Hi tech fraud Tuesday December 10, 2002 06:53 |
Tuesday January 14, 2003 18:20
Companies will spend their 2003 technology dollars on networking gear and systems to store data but handheld computers and wireless gadgets will fall by the wayside, a survey of global tech buyers said Tuesday.
The pension fund money did not evaporate. The money was stolen. Some of it using high tech!
JDS Uniphase is a financial disaster!
For example, Cronos, having previously heen sold to JDS Uniphase for $750 million in early 2000,
JSD Uniphase makes even Palm look good!
The study, by Wall Street brokerage firm UBS Warburg, also found that businesses intend to spend more on tech products this year than they did in 2002, going against a recent survey by rival Goldman Sachs that showed a spending decline.
UBS Warburg questioned 85 Chief Information Officers, who hold the technology purse strings of global corporations, during the month of December.
According to the survey, the CIO's expect tech spending to grow 4 percent in 2003 over 2002, with spending on storage products and networking equipment topping the bill.
Unfortunately, since the peak in the economy and the stock market two years ago, the expansive credit and monetary policy has not been able to compensate for the fall in stock prices and the credit defaults resulting from the mal-investments during the boom. However, low interest rates and lax credit requirements have resulted in residential real estate appreciation and massive mortgage refinancing. So far, this has kept the consumer part of the economy relatively healthy. It appears that this credit generating engine has about run its course. Overall though, the credit destruction through bankruptcies and the fall in asset prices in the private sector has put our financial system in jeopardy. There are also immense nascent threats to our financial system from many different sources including possible rating reductions in the fig leaf AAA ratings of credit insurance companies that guarantee $1.5 trillion in debt, the potential for massive dislocations in the $75 trillion derivative market, the uncertain financial condition of Fanny Mae, Freddy Mac and the Federal Home Loan Bank that have financed the $2 trillion in residential mortgages and the potential drop in home prices. There is a threat of a run on the U.S. dollar by foreigners who own more that $7 trillion of our assets ($2 trillion more than we own of their assets) and must continue to be net purchasers of our paper at the rate of $1.4 billion per day to finance our balance of payment deficits which now amounts to 5% of GDP. A run could be effectuated by either political or economic reasons. A run could even be precipitated by our own institutions and individual seeking profits or diversification. And, there appears to be impending wars and their aftermath with indeterminate costs. Perhaps, the government (government and Fed are used interchangeably just as Mr. Bernanke does in his speech) is thinking along similar lines. In spite of what they say, they are not worried about the prices of goods but financial deflation which would lead to the mother of all depressions.
Responding to a piece by Gene Epstein, who writes a weekly column about economic issues for Barron's, which stated, "Over the 56 years since 1946, consumer borrowing habits don't appear to have changed at all," The Prudent Bear's Doug Noland recently produced the following figures: In 1946, as a percentage of national income, total personal sector liabilities were 31%, non-farm mortgages 13%, and consumer credit 5%.
At the end of 2001, however, total personal sector liabilities were 133% of national income, non-farm mortgages 70%, and consumer credit 21%. Non-farm corporate liabilities stood at 44% of national income in 1946, compared to 101% at the end of 2001; total mortgage debt was 23% compared to 93%; security credit 3% versus 10%; state and local government debt 7% versus 17%; and total credit market debt 192% versus 359%.
In addition, in 1946, the personal savings rate stood at 9%, compared to around 2% now.
Many high tech businesses are merely fronts for ripping off pension fund money.
The object is to get the investor money then attempt to create the appearance of great technical achievement while taking, say 5% of one billion dollars.
Crace [read other Grace article below] laments investors finally wising up high tech fraud. Want to be injected with a micromachine?
The American Stock Exchange biotech index, a widely used measure of the industry's health, plunged roughly 40 percent in 2002. That put biotech on an even steeper decline than the embattled Nasdaq, which slumped 30 percent, or the S&P 500, which fell "only" 20 percent.
If the closely watched Amex biotech index is any measure, last week's conference gave the industry a mild boost at best. Executives from more than 250 biotech, drug and medical device firms took turns last week delivering 30- minute pep talks to pension fund managers and other institutional investors. The index was flat until a last-minute flurry on Friday pushed it up 10 points to close at 356, a gain of nearly 3 percent. It was better than nothing, but less than a ringing endorsement.
You need a team of highly intelligent high tech swindlers to convince lowly unintelligent money managers to invest.
We will be exposing more of the peripheral microcontroller fraud in the future. They don't have the bucks or talent to produce the peripheral side software!
But there is good stuff too.
But, of course, you are wisely advised about 80C32s and USB 2.0
Read-up below, Moe has something important to say.
Get the right numbers as opposed to wrong or random numbers. Remote digitization works but not very well.
- Digitize at the sensor Monday January 13, 2003 09:59
- Gigahertz PCs
- Microsoft's Visual Basic 6.0 integrated into Microsoft Office, Visual C/C++ with exquisite in-line assembler
- USB 2.0
- Super fast 80C32 USB 2.0 SOCs running BASIC-52++ and 80C32 Forth operating systems, of course!
- Internet software distribution and maintenance.
- Interactive hardware and software problem diagnosis over Internet with Netmeeting ... using BASIC-52++ and 80C32 Forth operating systems, of course!, on the peripheral end.
Digitizing at the sensor is very important if you want to stay in business - random or wrong numbers shouldn't sell very well.
Monday January 13, 2003 20:32
Ray Duncan furnished Sandia labs its 8085 Forth using Jerry Boutelle's metacompliler.
We needed interactive incremental assembler support.
Boutelle's metacompiler did not implement <BUILDS DOES> which was required for incremental assembler support.
The morning bill, Joseph, and Konrad arrived to pick up the software at Laboratory Microsystems, Duncan received software from England which implemented <BUILDS DOES> with Boutelle's metacompiler.
Here's who did it in England
----- Original Message -----
From: "Stephen Pelc"
To: Sent: Friday, October 25, 2002 8:30 AM Subject:
Re: Nautilus 1 metacompiler, ray duncan, andBill
>Are you the guy that did the for the Nautilus 1?
Yes. Our V6.1 compiler is a linear descendant of v1.Regards,
Stephen
"did the", oops I omitted <BUILDS DOES> but, hey, we all make mistakes.
So if you want to do an ARM project, you fool, and have fun too, buy MPE FORTH!
Bill
Thanks for the link.Despite your qualms about ARM pricing, one of my clients is paying US$7.50 for a 50MHz ARM with 2 serial ports, 2 HDLC ports, the usual I/O and timers, and an integrated 10/100 Ethernet controller.
These days, the CPU core is less than 10% of the silicon.
MPE also sells 8051 Forth!
We can't do it all by ourselves!
So let's have some fun doing it together and sharing code that we would otherwise have to try to reinvent. And probably screw up.
Most windows driver writers are from Russia.
Bill was told he is one of the very few in the US.
Litigation caused bill to get into the driver writing business!
Russian equities continued their world-beating run in 2002, with the benchmark RTS index gaining nearly 40 percent to trump all but Pakistan's main bourse. But market players see little chance of similar returns in 2003. ...
Except for the Pakistan Stock Index's impressive 118.21 percent gain, the dollar-denominated RTS outperformed all other indices in the world, followed closely by the main exchanges in Prague and Budapest, which gained 38 percent and 34 percent, respectively. Indonesia's Jakarta Composite Index rounded out the top five with a gain of 26 percent.
The world's two largest emerging markets -- Argentina and Brazil -- suffered the most, with Buenos Aires' Merval and Sao Paulo's Bovespa benchmark indices losing 47.11 and 45.83 percent of their value on the year.
Beware, say two finance experts in an alarming study recently published in the Financial Analysts Journal. They say that history--or at least the interpretation usually put on stock market history--is bunk. If they are right, the risk premium in coming decades won't be anything like 5% and you shouldn't count on the S&P 500 to provide you with a fat and comfortable retirement. Let's tour their findings and, for those who accept them, suggest an alternative to abandoning equities.
The U.S. has a bad debt problem. Debts cant be paid from cash flows. Total Credit Market Debt for the U.S. is up to $31 Trillion that is clearly more debt than a $10 Trillion economy can pay back. Cash flows that are needed to sustain asset prices arent there. The debt is bad. As the debt fails, the asset prices it support fall to true value. Indeed, deflation in massively overly valued assets cant be prevented. Many individual stocks will drop. The general level of stock market prices remain extraordinarily expensive. Moreover, so much new credit has been extended on real estate, that prices have been driven up to unsustainable levels. Deflation in real estate markets has already started. ...
What ought to follow a spectacularly absurd boom is a spectacularly absurd bust....
For the moment, the U.S. economy continues to run ahead of 'all reasonable expectations.' Eventually, reasonable expectations will catch up. Or, at least they ought to.
"This is the most dangerous period in financial history." - John Templeton, The Miami Herald, April 2000
As President Bush courted Democratic and Republican lawmakers Wednesday at the White House, the controversies swirling about his tax-cutting plan include a debate about deficits --- which the plan is almost certain to deepen.
In general, the mix of tax cuts and increased spending spells deficits. And 2001 saw both: a 10-year, $1.35 trillion tax cut and a huge, post-Sept. 11 boost to the military budget.
The equation became more lopsided Tuesday with Bush's announcement of a $674 billion package. The 10-year plan combines with increased government spending to virtually guarantee that those deficits will deepen --- at least in the short term. ...
"The deficit is now only about 1.5 percent of GDP [gross domestic product].
We survived with deficits of 2 or 2.5 percent during the mid-90s," Dhawan said.
Reagan became president with promises of smaller government and less taxes, but then he spent money like there was no tomorrow. From the end of World War II until the year Reagan was elected president, the national debt had slowly increased by $649 billion over thirty-five years but during his administration it skyrocketed by 1.7 trillion in just eight years. When he left, the debt was $2.6 trillion-almost three times what it was before he became president. The amount owed for each U.S. citizen shot up to $10,534. As a result, much of today's national debt is the interest owed on the money spent during the Reagan years. This also resulted in the United states becoming a debtor nation in 1985, losing its status as the world's financial leader - a position it had maintained since 1914 - to Japan.
We getting too much stuff on the main page! We know but we are too busy most of the time with our other big project.
Some states are selling bonds to try to cover pension underfunding.
The PBGC bailed out 150 bankrupt plans in the past year, compared with 104 the year before. The size of your insured benefit depends on when you retire and what kind of pension you take. Workers can get up to $3,665 a month if they quit at 65 with a single-life pension. But if you retire at 55 with a lifetime annuity for your spouse, your maximum drops to $1,484. The PBGC does not insure state and local government pensions, where underfunding may be twice as bad as in the private sector. Taxpayers blindly defer this cost.
Each of these instances reflects a powerful wave sweeping the economy -- overcapacity.
In the factory and in the office, it has a more concrete dimension: Businesses can produce far more than we need. Supply has simply outstripped demand. When that happens, production slows, equipment sits idle, costs go up, workers are laid off, and investments are postponed. ...
U.S. companies overbuilt in the 1990s, believing that the good times would never end. But the ``bubble'' popped with the turn of a new century and with the economy tipping into recession in March 2001 and then stumbling more after the Sept. 11 terrorist attacks.
Where did the pension fund money go? It did not evaporate or get erased as some economic journalist would like to have us believe.
These problems have little to do with any change in the number of people retiring, or an increase in their benefits. Rather, investments by the pension funds have fared poorly in recent years. As the prices of stocks and other investments have fallen, so has the return on the money set aside for the more than 44 million current and future private-sector retirees who qualify for traditional pensions.
At the same time, unusually low interest rates are further undermining pension plans. The effect of the bond rates is on the financial calculations used to determine the present value of the pension liabilities, not on the pension funds' return. Falling rates make future pension obligations look bigger on current balance sheets. To meet their obligations to workers, and to stay in compliance with pension laws, companies have been forced to set aside more money. ...
Still other companies, including Lucent Technologies, Boeing and Delta Air Lines, have been forced to reduce their net worth to reflect the way their growing pension obligations have outstripped their assets. ...
The sickest pension plans, by far, are at steel companies, followed by airlines. Other ailing sectors include the auto industry and its suppliers, the rubber and tire industries, and telecommunications. At worst, having to make a large pension contribution can push a company with insufficient cash flow into bankruptcy
Sunday January 12, 2003 19:46
"I don't think we're seeing depreciation in the low- to mid-price ranges," he said. "In fact, it's actually a hot market in the lower-price ranges in many areas. But if you look at the upper-price ranges -- starting at $200,000 to $250,000 for example -- you may see that prices are off by 10 to 20 percent, or more, compared to where they were several years ago."
Thousands of homeowners could soon be facing the prospect of negative equity, the Halifax is warning.
The bank, with two and a half million customers the largest mortgage lender in Britain, said a cooling housing market will soon see many with homes worth less than the value of their loans, the Sunday Telegraph reported.
Frank 't Hart, general counsel to SNS, said that when the bank confronted Citibank with its poor performance, it was assured that the portfolio would not fall much further. But by late 2001, SNS had lost three-quarters of its $15 million investment. Now it is worth $3 million. "With respect to this kind of investment, Citibank had a certain expertise we did not have," Mr. 't Hart said. "We hired Citibank because we believed them to have a certain level of professionalism and integrity and they have violated the trust we had in them." ...
Of the seven debt issues in the Captiva portfolio that have defaulted, five were bought by the Citibank manager. These include obligations of 360networks, a manager of fiber optic networks; Big V Supermarkets, a privately held food retailer; ICG Communications and Winstar Communications, both defunct; and the LTV Corporation, the steel maker. The LTV bond was underwritten by Salomon Smith Barney, another Citigroup unit. ...
Once again: Buyer beware.
Guys, we are interested in things other than computers, 8051 Forth, BASICs, assemblers, real world economics, pension fund fraud, microcontrollers, microcontroller operating systems, killing Iranian kids, counterfeit PhDs,
Bill, Lewis's MS and Ph.D thesis advisor, warned about professional birth control 29 years ago.
William H. Payne: Graduate Education: The Ph.D. Glut. CACM 16(3): 181-182 (1973)
deficient crypto and authentication algorithms, digitizing at the sensor, reloading, Alaska, fishing, hunting, fixing old VWs and Ford trucks, Internet. ..
But there are other interesting things too
February 1, 2003. It was the first day in the Year of the Horse. It is Year 4700 by Chinese calendar.
For those interested in astrology, it is the year of the Goat (sheep, ram).
Patty and bill are working out celebrations plans.
Random numbers or wrong numbers are sure interesting. And can be produced lots of different ways!
We got this thing about random numbers. And their avoidance in measuring sensor outputs.
But this leads to discussion of purple engineers. Those who advocate and use remote digitization.
This doesn't work very well.
In nearly all cases, there are better and more positive things to do.
Like
Patty and bill and working are working our celebration plans.
Aclara, Affymetrix, Caliper, Cepheid, Orchid, and, of course, JDS Uniphase
Then there are peripheral microcontroller companies. Like ARM.
Let's look at the numbers to make a guess what may happen in 2003. Like 3COM.
Below is going to cause a big shake-out in the peripheral micrcontroller industry .... and those do don't digitize at the sensor too, of course.
Tax cuts, while nice, do not increase corporate earnings nor increase the ability of corporations to pay dividends. Overvaluation is the mighty demon that todays perma-bulls are vainly battling, not socialist taxation issues that can be fixed with the mere stroke of a pen.
The Great Bear is not over and there remains a long, long way to fall yet before we reach The Bottom. Caveat Emptor!
"If there is a fourth down year, many investors are likely to look for a fifth -- or a case of them," says James Dines, editor of the Dines Letter.
"What bugs me is that economists insist America is still in an economic upturn, but I find myself disbelieving their numbers and conclusions," he says. "Are they fools or knaves?" ...
"One of the notions is that the market can't decline four years in a row since the only other time it has done so was during the Depression, and we're not in a depression," says John Hussman, professor of economics at the University of Michigan and head of Hussman Econometrics, a research firm.
"The difficulty is that the market's decline from its 2000 peak began when stocks commanded a price/peak earnings multiple that was fully 50 percent higher than the valuation at the 1929 peak," he says.
In March 2000 when the speculative bubble burst, the P/E ratio was more than 30, the highest of any bull market in history.
What's a P/E? It's a ratio that gives investors an idea of how much they are paying for a company's earnings power. So the higher the P/E, the more investors pay for the stock and therefore, the more earnings growth they are expecting.
Here's another story of a success, not a failure.
This success is having gotten away with the $1 billion ... of likely pension fund money.
In some ways, it's another verse in the dirge heard everywhere in Silicon Valley.
Let's look at the financials of some of the micromachine biomed companies mentioned by Grace.
Aclara, Affymetrix, Caliper, Cepheid, and Orchid
Only Affymetrix finally showed a profit.
2003 looks to be the year for more spectacular high tech bankruptcies.
But, hey, these companies succeeded in what they were trying to do all along. Steal the pension fund money.
The pension fund money did not evaporate which some financial journalist would like us to believe. And Grace has helped us identify four, maybe five, biomed companies which pocketed the money.
Lewis has concerns about bonds as investments.
Moody's Investors Service said today that bond defaults hit an all-time quarterly record $34 billion in the first period of 2002, although the global speculative grade bond default rate fell for a second consecutive month in March to 10.3%. By the end of the year, Moody's forecasts a default rate of 7.4%.
Late 2003?
Pierre Belec, of course, should write about this.
Some companies and even states are trying to fund missing pension fund payments by selling bonds.
Lewis warned of the Crash of 2000.
But got, not 2001, 2002 totally wrong. Where did bill fail?
Let's see what's going to happen in 2003.
Bill, Lewis's MS and Ph.D thesis advisor, warned about professional birth control 29 years ago.
William H. Payne: Graduate Education: The Ph.D. Glut. CACM 16(3): 181-182 (1973)
We must keep in mind that some of these high tech ventures were merely fronts to get at the pension fund money. And they worked really well help steal the money!
The pension fund money did not evaporate. The money was stolen. Some of it using high tech!
JDS Uniphase is a financial disaster!
For example, Cronos, having previously heen sold to JDS Uniphase for $750 million in early 2000,
JSD Uniphase makes even Palm look good!
Let's hope money managers are going to be a bit more careful where they invest the pension funds.
But there's good stuff too. USB 2.0 and may, if Lewis is right, Firewire
> Apple's new portables sport Firewire 800 which runs circles around USB 2.0
> and 802.11g which is compatible with 802.11b but 5 times faster. So, once
> again, Microsoft and Intel will have to back off from 802.11a and complete
> dependence on USB 2.0.
Here's a related article by Grace.
Financial Woes Mar the Road to
M3 Commercialization Roger H. Grace, Roger Grace Associates The industry known as M3 (that is, MEMS, MST, and micromachines-the microsize sensors that promise to change our world so dramatically) experienced its first slowdown in a decade during 2002. Besides the recent market dynamics affecting high tech in general, many M3 companies- even those that have been in business for years-suffered shortfalls typically associated with early-stage business. The M3 Commercialization Report Card (Figure 1), which I introduced in 1998 and have updated each year since, tracks the industry's performance in overcoming harriers to commercialization by covering the major success factors for M3 companies. (For detailed versions of the two most recent report cards, see Sensors Expo & Conference Proceedings, Fall 2001 and Fall 2002.) The Change column quantifies the difference in grade between 2001 and 2002. The most significant jumps are associated with the loss of gains achieved during 1998-2001. Other variances are positive, however. Here's my grading rationale. Market Research Improvement here is due in large part to the European Union organization Nexus, which in February published Market Analysis for Microsystems II: 2000-2005, a follow-up to its 1998 study. Developed with input from the Nexus User Supplier Clubs, the study provides hoth a top-down and bottom-up classical approach, and appears to he the most comprehensive report yet on M. At least 10 other reports have come out since 1992, when the seminal microsystems market study by Battelle Frankfurt was released, hut most (if not all) were conducted before the economic downturn and the optical telecom sector's extreme plunge. The Nexus report, because of its recent publication, avoids this error. The MEMS Industry Croup formed in 2001 as an association to support the advancement of M3. In addition, two other groups, while not formal industry associations, perform many of the functions of a classical trade croon. MANCEF. the Micro- and Nanotechnology Commercialization Education Foundation (established in 2000), which develops and manages the annual Commercialization of Micro- and Nanosystems (COMS) conference, has created an extensive industry roadmap and is a major supporter of M3 standards with the Semiconductor Equipment and Materials International (SEMI) organization. And the European group Nexus has many initiatives, including its very successful User Supplier Clubs, which operate similar to those of trade associations. Venture Capital Attraction Before 1999, M3 companies were considered too risky for most investors, and did not satisfy a number of the critical criteria for venture capitalist and private investor (angel) funding. Among these are a unique product concept, a strong management team with demonstrated industry successes, a large and well-defined (vis-a-vis formal research) market size and growth rate, a served market, a target market definition, and a significant return on capital in the first five years of operation. The advent of MEMS-based optical switches, however, changed this situation. BioMEMS companies, including Aclara, Affymetrix, Caliper, Cepheid, and Orchid, went public in the past three years, and some enjoyed large capitalization in early 2000 when the stock market exploded and biotechnology was so generously valued. But the downturn, especially the technology-rich NASDAQ market, has greatly reduced their values. First-round funding is minimal in all high-tech sectors, including M3. Many of the acquiring companies' stock prices have fallen 75%-80% from their highs of March 2000. For example, Cronos, having previously heen sold to JDS Uniphase for $750 million in early 2000, was sold to MEMSCAP in July 2002 for $8 million. JDS stock now lists for less than 10% of its previous high. In all business areas, venture capital investment in 2002 is expected to he the worst in the last decade: $20-$30 billion, down from $55 billion in 2001 and $108 billion in 2000. In early 2000, seed and first rounds accounted for 60% of total investments; they are currently running at 30%. A Price Waterhouse Cooper/ Venture Economics/National Venture Capital association Money Tree survey done in August shows that for every dollar invested in a new company, five to seven times this amount was invested in existing portfolio companies (although new deals in biotech and software received one third and one fourth, respectively, of all new funding in their categories). This is not good news for startups in general and certainly not for M3 startups-unless, perhaps, they are bioMEMS startups. Bucking the trend of traditional venture capitalists is Ardesta, the first MEMS-specific venture capital/technology accelerator firm. Formed in 2000, Ardesta has helped to launch a number of exciting startups that might otherwise not have been funded, and to accelerate companies such as Sensicore, IonOptics, and Tessera. Ardesta's strategy is to license intellectual property from leading institutions, including Sandia National Laboratories, University of Michigan, and University of California, Berkeley. Creation of Wealth Many people who became millionaires on paper as a result of optical telecom buyouts (based largely in stock and on future performance) saw their fortunes evaporate over the past two years. Acquisitions still continue, as with General Electric acquiring TRW Novasensor in September. The IPO of MEMSCAP in 2001, the first in the history of M3, has not fared well as a result of the dip in high tech, especially in the worldwide stock markets. Industry Roadmap M3 roadmaps have been developed by three organizations: Nexus (issued in late 2000), MEMS Industry Group (issued in late 2001), and MANCEF (issued in late 2002). Each takes a different approach, with MANCEF's being the most thorough. In 14 chapters it addresses such factors as foundries, standards, EDA tools, market analysis, and reliability. The Future The year 2002 was pivotal for high tech worldwide. Within M3, expansion in many important areas-including infrastructure, roadmaps, standards, and industry associations-portends maturation. But the business climate worldwide is still turbulent. Many U.S. M3 companies that had hoped to emulate the early successes of Cronos, Xros, and Intellisense are finding themselves out of business or canceling development in optical MEMS telecom systems. If M3 companies are to succeed during these trying times, they must define exactly what the market wants and what their competitive and sustainable advantages are. Roger H. Grace is President, Roger Grace Associates, Naples, FL; 415-436-9101, rgrace@rgrace.com, www.rgrace.com. Sensors January 2003 www.sensormag.com |
Saturday January 11, 2003 20:50
2003 may be the year of technical occupational migration.
Determine how committed you are to your old field. Do you actually want another job like the one you had, or would you rather try something different? If you want to stay in the same field, get as much information as you can to determine how realistic that is. Are jobs in your area expected to come back in Silicon Valley, or are they being automated or moved overseas? If they do come back, do you have the qualifications employers want? ...
Consider a career change. Perhaps you never really liked the field you were in. Or you realize that you're not likely to get another job like the one you had, no matter how much you want it.
Csd bought a CEN-TECH 35017 for the reason our previous multimeters didn't have capacitance and frequency measurements.
Deflation?
First use was measuring tachometer output of an MSD electronic ignition connect to white ford.
Search for Here's the compressor
The CEN-TECH works great. As well as do their cheaper, about $4.95, digital counterparts!
We've measured electronic components with it and our previous expensive multimeters. The CEN-TECH gets the same answers!
The CEN-TECH is a microcontroller product!!!
Bill ordered gobs of chips while at Sandia labs. As a result he know lots of chip reps.
A former [he was over 50] TI employee alerted bill to TI DLP technology.
We are using high tech for our projects.
And, of course, csd is watching TI/BB 80C32 mixed signal technology very carefully.
And firewire v USB 2.0 technology!
What the hell is DLP and why am I about to buy
it?!?
The beginners guide to better video through projection! Digital Light Processing is HOT! Over 75% of new projectors are now using this radical technology from Texas Instruments that uses hundreds of thou- sands of tiny "micromirrors" to produce the image. LCD is about 20% of sales and less than 5% are CRTs (actually. I believe it is now less than 1%!). And just about 5 years ago. DLP was barely even a blip on the radar screen. When the DreamVision DLP projector came out about 3 years ago, we picked up on it right away because it had a "breakthrough price" of $6800 and, combined with the then new $700 Scan line doubler, it provided an unprecendented level of picture qual- ity. Two years later, the Plus Piano came out and produced a picture that was twice as good for less than half the price AND included the $700 iScan linedoubler internally for free. Now we have an improved version of the Plus Piano and the new InFocus 7200 featuring the amazing new HD2 "Mustang" chip for twice the detail and twice the contrast ratio. Buckle up. DLP it's getting exciting! The big thing about DLP front projeciion is that it looks DRAMATICALLY betrer and more filmlike (even beyond film in many ways) than ANY rear projection TV. Even the finest, most expensive rear projectors look bad after seeing a DLP image. Why? Well, several reasons. One is that rear projectors have to bounce the video off of several mirrors which adds refraction distortion. Then, the image is splashed onto the back of a fresnel screen designed to scatter the light so that you can actually see the image. However, this screen adds even more refraction distortion and blurs the image to the point where computer data looks visibly fuzzy. And, because the screen can't scatter the light evenly and isn't even truly flat, you get "hotspotting" that makes some parts of the screen look brighter than others. Front projectors do NOT have any of these prob- lems. When you look at a front projected image on a screen, you see a crispness so precise, you can count the pixels and even see the dimple in the middle of the DLP mirrors IF you get close enough to the screen. Rear projection TVs simply don't have the kind of resolution required to show this level of detail. And computer generated data looks "pixel-per- fect". More than a few recent buyers of rear projection TVs have seen the Plus Piano front projector and said "Aw....crap. NOW you tell me." Hey we DID send out the info last springl Don't you read the newsletter?!? So, here are our thoughts on video. We DO NOT recommend, nor will we sell Plasma "flat screen" TVs. They are too expensive, too unreliable and not very good looking. We arelukewarm about rear projectors. They are a good compromise of big, day-to-day use video, but do not really pro- duce a "high-end" image because of optical problems. New TFI LCD panels (the new flat screen computer monitors) look promising and already produce a better image than either Plasma or CRT, but are too expensive and are not fast enough to produce a clean image when there is fast motion. However, at the rate they are going, they WILL be price/size competitive with Plasma by next year and will almost certainly supplant it within a few years. In fact. I personally believe that Plasma. rertainly as we know it. won't exist as a technology in 3-5 years. TFT LCD is moving so fast, conventional tube televisions will be in mortal danger in a few years. Until then, we think the BEST way to get great video is to keep your old tube TV for normal day to day use and a DLP front proector for primetime and movie watching in a somewhat darkened room. Then you have the best of both worlds. A good bright image for daylight watching and a huge. high-end theater-like image for movies for more serious night-time watching. TOP 10 REASONS TO DO A DLP FRONT PROJECTION SYSTEM 1. A picture up to 4 times the size of a "big screen" TV A DLP front projector can put out a picture up to 120" (bigger if you have the wall space). That is 4 times the size of a typical 60" rear projector. In a typical house. that gives you an image that appears as large or larger than the largest movie theater. 2. Razor sharp graphics for computer video games - These projectors do double duty with computer graphics and produce a sharper. more precise image than ANY other technology. Simply tell your computer to put out the same resolution as the projector and you have a virtually perfect representation of your computer screen. And there's no burn-in like conventional CRTs! 3. Better reliability than other technologies - DLP is the most reliable technology to ever come out for video reproduction. Each DMD is designed to work flawlessly for approximately 100.000 hours. That's 11 years of continuous use with no LCD pixel drop-outs. no CRT fade or burn-in, no catastrophic failure. All you have to do is replace the bulb every 1000-3000 hours and keep the projector clean. We can't guarantee that the projector itself will last 100.000 hours as that would be about 100 years in normal use, but you get what I'm talking about! 4. Virtually no impact on the sound of your speakers - Unlike a big TV. a projection screen on the wall doesn't significantly impact the imaging of your speakers. And. after all, creating a realistic soundstage is the entire point of 5.1 surround sound. Not only is front projection the ultimate in video quality, it actually can make your audio system sound better! 5. Gives you back the space in your room - Even though you can easily get a 100" or larger diagonal screen, the screen itself, including the frame is less than 2" thick. That's thinner than plasma! And it's bigger AND it's better. Get rid of that enormous rear projector and get something bigger. And smaller! 6. With a drop down screen, disappears when not in use - If you think TVs are ugly, you'll LOVE the idea of a drop down screen. Many of our customers have a beautiful painting on the wall, but when it's time to watch a movie, the screen simply comes out of the ceiling or out of a thin tube on the wall and gives you the movie theater experience of your dreams. 7. Screen investment remains, even if you upgrade the projector - Projector technology is improving almost as rapidly as computer technology, but your screen is an investment you can keep for decades. So. while a newer, more exotic projector may catch your eyes in 4 or 5 years, your screen will be with you for a long time. 8. Looks as good in 20 years as it does today - Unlike a typical TV that begins degrading the moment you turn it on. DLP won't fade or lose color. Simply replace the user-installable metal halide bulb when it burns out. DLPs don't lose quality For as long as they work, they produce their max- imum quality. 9. Better overall picture quality than any other technology or design - DLP is at the cutting edge of image quality. While other technologies are attempting to catch up. DLP is maintaining its edge. Only ultra-expensive CRT's can claim to be as good, and that's only if you have the patience and skills to tune them and converge them properly And if you don't think DLP is good enough yet, you haven't seen the new InFocus 7200! 10. The price/technology has finally become ~right" - Our first DLP projector showed the promise of this exciting technology. The new Plus Piano significantly improved on that and dropped the price to 40% of the award-winning DreamVision. And now, the InFocus has removed all doubts with double the resolution, contrast and brightness of the Piano (for twice the price!). If you're on a budget. you'll be shocked at the per- formance of the Plus Piano and InFocus projectors. especially compared to "big screens". If you're lusting after the hottest picture available this side of $20,000. the new InFocus 7200 will blow you away and keep you sat- isfied for years and years to come. If ever there was a time to "get in" to DLP. NOW is the time. Composite Video - Composite is exactly what it sounds like. It is a "composite" of all of the signals that make up the image. It uses a single RCA type cable and is analog land therefore easily susceptible to distortion I. Unfortunately, to create an image. the signal has to be split out into a three-color RGB signal. Composite is the lowest form of video-only signals because it is difficult to sepa- rate out the signals without significant distortion. However, almost all video components have either composite in or output because it is so common. S-Video - This is a noticeable leap over composite video. By keeping the Y or "brightness' signal separate from the C or "color' signal. there is a cleaner video with much fewer artifacts. S-Video is always preferred over composite. Satellite. new cable boxes. DVD and S-VHS all use S-Video for a better, smoother picture. S-Video uses a fragile DIN connector similar to a computer mouse cable, so be careful when plugging it ml Component Video - Component is the best consumer form of analog video. It has a Y signal. presumably identical to that of S-Video, but the color is created through a "color difference" method. One signal. R-Y or "Pcr' creates colors towards the red end of the spectrum and B-Y or "Pcb" creates colors towards the blue end of the spectrum. This is easily translated into the RC,B standard that TV actually runs on. Component uses three RCA cables. typically marked red. green and blue (green being Y). DVI - Digital Video Interface is a newer. computer-oriented digital connection that is gaining popularity for HDTV. A newly adopted copy-protection scheme (HDCP) is needed at both ends to send and receive high-definition digital signals. DVI is now being used on many new DLP projectors and now for high-end HDTV boxes. both of the oft-air and satellite varieties. Upcoming HD1'V cable boxes may also use DVI. DVI is an exceptional method of transferring video as it is com- pletely lossless. and since DLP is all digital. the signal goes through no analog cir- cuitry or filters whatsoever. Firewire/EE1394 - Firewire is even more advanced than DVI as it can handle two-way communication and audio as well. It is more complicated and a little less standardized. but it is poised to be the other standard for the future of audio/video. Firewire is the standard for HDTV recorders such as D-VHS and hard drive recorders.
NTSC/ATSC - Traditionally. video standards are named after the committee that standardized them. The National Television Standards Committee developed the old broadcast standard. The Advanced Television Systems Committee developed HDTV. NTSC is an analog "interlaced" format. It produces the odd lines of an image in one "field" and then produces the even lines in the following field. ATSC HDTV is digital can be either interlaced or "progressive". Progressive runs at double the speed and can show the entire frame in one field. Therefore, it looks dramatically smoother than interlaced and has dramatically less noticeable "motion artifacts" which is the edginess that occurs when things move on the screen. HDTV is also scaleable to different resolutions. One channel of 120 or 1080 lines (three times NTSC TV) can be transmitted or up to 5 channels of 480 lines can be transmitted simultaneously. You'll see terms like 480~ and 72 Op with HDTV. The first number stands for the lines of resolution. The 'i' or 'p stands for interlaced or progressive. CRT- Cathode Ray Tubes (hence the phrase "the Tube") fire electrons at a grid of posphers to create an image on a screen or via three large "guns" to produce a projected image. These are bulky and are quickly being replaced by a host of new technologies. Within 5 years. there almost certainly will be no more tube TVs or projectors. Once newer. more glamorous technologies reach price parity with CRTs. there will be no desire on the part of consumers to buy the old-fashioned "tube" any more. Plasma - Plasma uses millions of small phosphor coated bubbles that are activated by electrical currents to glow in the right color. In a sense. it is similar to a CRT. but instead of one gun that aims the electrons at each phosphor one at a time, the phospors are fired by electrically and individually operated plasma inside the bubble. Plasma's major problem is expense and reliability. They also do not work well and squeal at high altitudes. It also seems to be stagnating in price/development and is in danger of being overcome by TFT LCDs (see below). We don't recommend or sell Plasma for these reasons. TFT LCD - LCD panels have been around since the first portable computer. However, new models use a 'Thin Film Transistor" design that has individual transisters to drive each pixel element. TFT LCDs are signficantly brighter, faster and have a wider viewing angle than older LCDs. Most, if not all, new portable computers use TFTs. The only major problem with TFTs right now is that they are a little too slow for fast action and therefore blur a bit at times and the price. However. TFT's size/price ratio is advancing at a rapid pace and some of the newer ones don't seem to blur at all. We believe, barring a more advanced technology. TFT LCD will eclispe Plasma within a few years. We are considering selling these, but are waiting for the price to become more attractive. They will never be as large as a projection system. however. DLP - Digital Light Processing uses Digital Micromirror Devices ( DMDs) to produce startlingly clear images on a screen. It is a reflective technology and uses hundreds of thousands. soon to be millions of mirrors only microns across (they all fit on a small chip) to reflect an image through a lens. Despite the apparent complexity, reliability is exceptional. The picture needs to be seen to be believed. Current popular chips are the 480x848 and 720x1280 16:9 chips. Working prototypes for 1080x1920 projectors have been built. but there is no immediate plans for producing these for mass sale. yet. However, it is almost certainly only a few years away as production technologies and the competition improves. DLP now accounts for nearly 80% of all video projectors now being sold. The next big evolution will be more affordable 3-chip designs which will be even brighter and smoother looking, but those are more than $30,000 at the moment. Front Projection - Front projectors send an image that is reflected off of the front of an opaque video screen. It is the best way of producing an accurate image as it is optically simple and there is no distortion due to refraction or the absorption by the screen. It's main liability is that it needs to be relatively dark for a good image. It is best used in tandem with a standard TV for daytime use. Rear Projection - Rear projectors are a good compromise of size and all around use and project an image THROUGH a screen from the rear. However, they ARE a compromise. Video quality is generally poor relative to other technologies. including CRTs and LCDs. This is because the plastic screen that disperses the image adds a significant amount of distortion seen as a blurring of the signal. Also. like most TVs. it is sensitive to bright light and can be washed Out almost as much as a front projector if there is any direct light in the room. If you blew up the size of a rear projection TV to front projection size. it would be immediately apparent how much more blurred the image is compared to a reflective image. Once you see a good DLP front projector. you'll never want to own a rear projection TV again. 4:3/16:9 Aspect Ratios - The aspect ratio is the ratio of the width to the height of the image. Standard TV uses 4:3 ratio. In other words, a screen that is 3' high would be 4' wide. It is also known as 1.33:1. A 16:9 ratio has a wider aspect ratio and is more commonly referred to as "widescreen" or "HDTV format . 16:9 is also known as 1.78:1 and is closer to film's common 1.85:1 ratio (often seen as a "letterboxed" format). 16:9 is the main standard for HDTV. although 4:3 ratios are supported. Even computers are now moving to widescreen format. If you buy a projector, you should get a wide 16:9 screen. Letterbox - Letterboxing is the common method of shrinking a movie onto a relatively narrow ratio screen. When the image is narrowed to fit, the height also shrinks, and black bars appear at the top and bottom of the screen. Buying a widescreen TV does not always eliminate the bars as many DVDs are shown in the wider 1:85 and even as wide as the 2:35:1 aspect ratio. However, most HDTV material and many DVDs are in 16:9 ratio and therefore would not be lerterboxed on a widescreen TV. but would be on a standard TV. Anamorphic - Anamorphic widescreen is a method of squeezing a wider image into a narrower profile. DVD is actually a 4:3 aspect ratio format. at least until a more advanced version arrives in a few years. That means a widescreen film either needs to be shrunk in both height and width and then blown back up (letterboxing) by the TV/DVD OR the picture can be squeezed only in the horizontal domain, making everyone skinny in the DVD and then widened again by the TV/DVD to produce and accurate widescreen format. The advantage of anamorphic is that, while letterboxing loses both vertical and horizontal resolution. anamorphic only loses vertical lines of resolution and the image uses all of the available horizontal lines of resolution. Anamorphic DVD looks better on a widescreen TV than letterbox movies which are inherently lower in resolution. Pan and San/Full Screen - These are methods of showing a widescreen movie on a 4:3 TV and filling the screen. Full screen means that the movie is blown up the TV. but the sides of the image are chopped off by the TV. Full screen typically shows the middle 3/4s of the movie from a fixed postion. Pan and Scan is a more active way of focusing on the more important elements of the movie. Since filling the TV screen means throwing out 25% or more of the image. pan and scan hunts and pecks the scene so that the most important action stays on screen and only the "less important" elements of the movie are discarded. However. letterboxing or widescreen TVs allow you to see ALL of the movie, even if you have to shrink the image slightly to fit all of the width onto your screen. If you were to pan and scan Ben Hur. which is shot in a 2.3 5:1 ratio, you would be throwing out 44% of the moviel If you think letterboxing is a problem. you need a bigger. wider TV. let us help? adnm.com |
Pierre Belec strikes again at Yahoo! Finance.
In Austin the prices of houses once doubled, then fell back to less that where it started!
How many people will have been lured into a debt trap once the value of their home starts falling? Trust me, unless you've had it happen to you, you have no idea what it feels like to owe $130,000 on an $80,000 house, as happened to many people in the Texas real estate bust. (In most areas, you can probably double or triple those dollar amounts for today. How many people have that much cash laying around to make up the difference?) Thinking that asset prices can forever be levitated is a pipe dream. Ultimately, the payments on loans have to be paid, and that depends on how well capital has been allocated to meet the real demand of the economy. ...
The U.S. economy suffered a surprise 101,000 jobs drop in December, the government said on Friday, underscoring gloom among retailers and other employers and raising the political stakes for President Bush as he touts a $674 billion stimulus plan.
Here's neat email recently received from Kim Komando
MY WEEKLY COLUMN: Making the PC-Gadget Connection
So, you got that new computer up and running. And now you're looking at all the wondrous things you could add to it--digital cameras, music players, printers, memory card readers, wireless keyboards and mice, to name a few. Unfortunately, the boxes tell you to hook it up with USB 2.0, or Firewire, or, maybe even, Bluetooth. These are all protocols to link things to a computer.
Let's demystify them, starting with USB (Universal Serial Bus) 2.0. Of the really high-speed connectors, this is the most common. As its name implies, USB 2.0 is the second iteration of the universal serial bus standard. The original, 1.1, is still around.
The difference? Speed. USB 1.1 moves 12 megabits of data per second. USB 2.0 moves 480Mb of data per second. This means USB 2.0 is about 40 times faster than its predecessor, USB 1.1.
Why is this important? Think about MP3 music files. These are normally 2.5 to 3 megabytes per song. If you were transferring 1,000 songs (some people have that many!) to an MP3 player, you could spend a half hour using USB 1.1. If you used USB 2.0, it would take less than a minute!
How can you tell which you type of USB port is in your computer? Unfortunately, the slots look the same. Check the documentation that came with your computer. If the computer is new, at least some of your USB slots should be 2.0. If it is more than six months old, you may be stuck with 1.1. If that's the case, and you need a faster slot, you can install a 2.0 card in your computer.
Some peripherals don't need a screaming-fast port. Keyboards and mice come in USB versions. They're normally 1.1, because they just don't move that much data. You can hook 1.1 peripherals to 2.0 ports, but they will still run at the 1.1 speed.
Firewire is another screamer. It is known formally as IEEE 1394. It was developed by Apple, and is often used to download video. Those are huge files, and like a slew of MP3s, can take forever with a slower method. Firewire is less common on PCs than USB 2.0. It is built into all Apple machines. Firewire is slightly slower than USB 2.0, running at 400Mbps. That is still awfully fast.
Firewire comes in two flavors: four-wire and six-wire. Six-wire is handier, because it uses power from the computer to run the peripheral. So, if you're downloading pictures from a camera, for instance, you don't have to use the camera's batteries. Four-wire ports require the peripheral to be powered. You can buy Firewire cards to add this port to a computer. If you buy a card, get six-wire.
Both Firewire and USB 2.0 peripherals must be connected by cable to the computer. Bluetooth is entirely different. It is a wireless standard. It is talked about much more often than it is installed. If you decide to retrofit a computer with Bluetooth, you'll find few parts. And there are even fewer peripherals equipped for Bluetooth.
Nonetheless, Bluetooth has been the subject of continuing hype for several years. It is a low-power system, good for about 30 feet. Therefore, it is impractical for networking computers. But it should work well with peripherals.
Bluetooth was named for King Harald II of Denmark. He reigned in the 10th century and had--guess what?--a blue tooth. Used on a computer, Bluetooth runs at 721 kilobits per second one way. A fast parallel port, traditionally used for printers, runs about as fast. Bluetooth is much, much slower than USB or Firewire. When you look behind your computer, what do you see? Mine is a jungle of wiring. Conceivably, Bluetooth could alleviate that condition.
The bottom-line: If you're using a PC and want to buy a peripheral, concentrate on USB (hopefully, 2.0) and Firewire, if you have a Firewire port. If you're using an Apple machine, concentrate on Firewire. Bluetooth is a good choice for wireless keyboards and mice, but not much more than that.
--> MORE ABOUT ME--KIM:
Whew! In addition to my weekly newspaper column, I have a weekly column on MSN's bCentral, this newsletter, a weekly column for USA Today, and other editorial products. Oh, and I have a weekly radio show heard by almost 8 million people.
--> HAVE A WEB SITE? Link your site to my site. The easy instructions are here: http://www.komando.com/fanlinks.asp
California has to pay about $800 million into the California Public Employees' Retirement System, the largest U.S. pension fund, and the California State Teachers' Retirement System, the third largest. Both funds lost money in the stock market during the past three years. If the state decides to borrow, the pension bonds would be sold in the fiscal year starting July 1.
Assumption v reality!
In slides released in advance of an analyst meeting, the world's second-largest automaker said its worldwide pension underfunding totaled $14.5 billion. Its U.S. pension funds had a negative return of 9.7 percent in 2002 as U.S. stock markets suffered their third straight year of decline
Ford said it was changing the assumptions for future returns on its U.S., Canadian and British plans to 8.75 percent from 9.5 percent.
Over a long period of time the S&P 500 has sold at an average P/E of 15 with a range of 7 to 22 excluding the late 1990s boom. At bear market bottoms the P/E lows have averaged about 11 with none higher than 16. Now the index sells at 30 times estimated earnings for 2002 and 24 times the estimated consensus earnings for 2003, which are probably too high. This is not only far above the level at any bear market bottom in history, but is higher than the P/E at the 1929 peak.
The valuations for leading technology stocks are even worse. Fred Hickey, publisher of the excellent monthly, The High-Tech Strategist calculates that 16 leading tech stocks with earnings are now selling at an average of 52 times trailing earnings and 42 times estimated earnings for the following fiscal year. While some will say that this is a result of depressed earnings, the average P/E in 1990, when earnings were also depressed, was 17. For those who dont trust earnings numbers, the four-quarter trailing price-to-sales ratio is now 7.7 compared to only 2.3 in 1990. We note that the 16 stocks in the calculation account for 32% of Nasdaqs total capitalization.
Thursday January 9, 2003 18: 46
There are the batteries too.
U.S. technology companies lag foreign rivals in reducing hazardous materials in electronic devices, exposing gadget-hungry Americans to toxins whenever they use computers, according to a new report.
Within the next five years, up to 680 million computers will become obsolete in the United States, producing more than 4 billion pounds of plastic, 1 billion pounds of lead and millions of pounds of other waste products, according to the National Safety Council. According to the CTC report, less than 10 percent of outdated computer products will be refurbished or recycled.
GM said it expects pretax pension expenses to increase from $1 billion in 2002 to $3 billion this year. The company said an analysis showed that its U.S. plans were underfunded by $19.3 billion at the end of 2002.
GM's U.S. plans were underfunded $9.1 billion at the end of 2001.
Based on a study by asset managers and actuaries, the automaker has decided to reduce its assumption for how much it will earn from investments to nine per cent in 2003 from 10 per cent in 2002.
The stock market just wrapped up its third straight year of losses, but when it comes to pension accounting, that reality has yet to catch up with much of corporate America.
Many companies are feeding their bottom line by assuming their pension assets will grow by at least 10 percent in the future, even though the stock market slump shows few signs of abating. ...
International Business Machines Corp. (nyse: IBM - news - people), for example, notes in its 2001 annual report that since 1986 its actual returns exceeded its projected returns by $5.3 billion. In 2001, the company's pension plans lost $2.41 billion, despite projecting returns of $1.45 billion.
"Phantom pension earnings are portrayed as income," she says. "It's a ticking time bomb."!
GM, the world's largest automaker, also said it is cutting the projected rate of return on its pension fund to 9 percent annually from 10 percent in 2002.General Motors Corp. (NYSE:GM - News) said on Thursday that its pension costs before taxes will triple to $3 billion this year because assets in the fund, hammered by three years of stock market declines, fell short of anticipated obligations by $19.3 billion in 2002.
The move by GM, which has the largest private U.S. pension fund, had been expected. Analysts have said many companies are likely to follow suit, depressing earnings across corporate America this year. ....
What U.S. papers say about Bush plan
Faced with a sluggish economy that bears some similarities to the one that helped sink his fathers 1992 re-election bid, President Bush proposed a stimulus plan Tuesday that would return $674 billion to taxpayers over 10 years.
President George W. Bush's $670 billion tax-cut proposal is his biggest risk on the economy because he can't guarantee success, analysts said. Democrats portrayed the plan as fueling deficits and benefiting the rich.
Wednesday January 8, 2003 11:14
On Saturday morning I opened the Business section and found an article by Richard Lambert headlined US dollar faces year of living dangerously. He makes many of the points I was planning to make, and makes them very well the overvalued and declining dollar, the enormous US deficit on current account, the sorry state of the worlds second and third biggest economies, Japan and Germany, the Asian central banks propping up the dollar, the sharp rise in the price of gold. ...
There have been a number of reasons for this strength of the gold price. Between 1997 and 2002 the US current deficit tripled from an annual rate of 1.5 per cent of gross domestic product to just below 5 per cent. Some projections expect it to reach 7 per cent by 2007; even the present deficit is unsustainable.
The Clinton presidency in the years of the boom between 1997 and 2000 lost control of the US current balance; despite an economic slowdown, the Bush Administration has not regained it, and does not even appear to have a proper understanding of the problem. Gold is priced in terms of dollars. Gold has risen by 36 per cent in terms of the dollar since 1999; that revaluation probably has further to go. Some American commentators now believe that gold has entered a ten-year bull market against the dollar
Feeling even more sick?
The stock market just wrapped up its third straight year of losses, but when it comes to pension accounting, that reality has yet to catch up with much of corporate America.
At the start of 2002, companies in the Standard & Poor's 500 on average estimated that their pension assets -- the bulk of which are invested in the stock market -- would post lofty returns of 9.3% for the year, according to a study by Merrill Lynch & Co. ...
That's because on average, pension portfolios have 60% of their assets tied up in the stock market while the rest usually are locked up in debt or other assets that generate even lower returns.
According to its annual report released in March 2002, Verizon Communications Inc., the nation's largest local phone company, had a strong year in 2001. In the opening pages of the report, the company announced an annual profit of $389 million.
Only those investors who dug into the small print at the back of the document learned that Verizon's reported earnings included $2.7 billion in gains from its pension fund investments -- profit that didn't really exist.
The company pension fund actually lost $3.1 billion in 2001, a footnote on page 58 of the 68-page report revealed.
Martin Weiss believes that the pension fund timebomb will be one of the factors driving the Dow down to 5000. GM owes $12.7 BILLION to its pension fund. The September issue of Safe Money Report lists some 50 companies that owe money to their pension funds; ranging from hundreds of millions to billions.
Microsoft will officially end technical support for Windows 98 and Windows NT 4 on June 30. The financial implications for Microsoft and others could be significant, even in the short term.
We realize this page is getting too big. But there is important stuff to add.
We have often been accused of inconsistency for criticising the Federal Reserves recent promises for radical monetisation to avert a Japanese-style deflation, whilst simultaneously arguing for exactly that sort of policy prescription in Japan. We have chosen to open the New Year by addressing this apparent contradiction. Although both countries share many comparable symptoms of a post-bubble distress, they are clearly at different stages of the economic cycle and concomitantly different phases of economic adjustment. We think the relative time frames of the respective economies highlight a crucial distinction. Yes, the vigorous production of Why Were Not Like Japan articles has some validity; but such analyses should not provide US investors with any degree of comfort. On virtually any current comparison, Japan is far more amenable to improvement given the savings swing from deficit to surplus in its private sector, than the US, which has spent all of its monetary bullets pre-empting, as opposed to accommodating, painful, but necessary, economic adjustment.
Wal-Mart is introducing basic financial services for US customers, using the same low-margin strategy that has turned it into the world's biggest retailer.
The entry of the discount superstore giant into financial services has always been feared by financial competitors worried that it could undercut their margins while facing a lighter regulatory burden.
The Walmart of software is advancing. Forth and BASIC-52, of course! And fun too!
Senior citizen problems
Blowing the whistle on Remote digitization and 2003 microcontroller shake-out.
Isn't it fun to be in an industry leading deflation? We simply love these inexpensive gigahertz PCs!
In that context, todays deflation risks are global in scope. Thats a key by-product of our baseline forecast of the world economy -- a 2.9% estimate for world GDP growth in 2003 following two years of gains averaging just 2.1%. Not only did the 2001-02 outcome fall a cumulative 3.0 percentage points below the global economys longer-term 3.6% growth trend, but our 2003 scenario adds another 0.7 percentage point to the global output gap -- perpetuating the deflationary tendencies of this subpar global growth cycle. Moreover, our first cut at 2004 -- a 3.9% increase -- is only 0.3 percentage point above trend, thereby barely making a dent in the outsize imbalance between global supply and demand. Consequently, notwithstanding the recent pop in commodity prices, traditional macro tells us that a persistently wide global output gap makes it difficult to envision the return of pricing leverage at any point in the foreseeable future. The risk of global deflation can hardly be ruled out. ....
Tech show attendance last year slumped 10% from the year before, twice as much as other shows. Key3Media, the producer of Comdex, the granddaddy of North American tech shows, says it may file for bankruptcy protection because of dwindling crowds. And Macworld Expo in Tokyo, a big Apple Computer-related show, was canceled this year because not enough major exhibitors signed up. Even the Consumer Electronics Show (CES), one of the industry's healthiest, expects only slightly more visitors this year than last. It starts Thursday in Las Vegas. ...
"Trade shows mirror the health of a sector, and the tech industry is in disarray," says Douglas Ducate, CEO of the Center for Exhibition Industry Research. He expects some tech shows to consolidate and more companies to copy eBay and Oracle, who do smaller conventions targeted at customers and developers. ...
Closures and mergers in the tech and telecom industries erased hundreds of companies and about 470,000 jobs last year. Many were habitual attendees of trade shows.
Lewis warned of the Crash of 2000.
Let's see what going to happen in 2003. A microcontroller shake-out?
Bill, Lewis's MS abd Ph.D thesis advisor, warned about professional birth control 29 years ago.
William H. Payne: Graduate Education: The Ph.D. Glut. CACM 16(3): 181-182 (1973)
2002 may be the year of technical occupational migration.
John Young forced bill into the internet business.
13 November 1998
William Payne promises that he and Arthur Morales will shortly have a Web site for offering information on their activities and those of others. A URL for that site will be available soon.
Enclosed is a Grand Jury subpoena requiring that Cryptome produce certain records.
Reality?
Only those investors who dug into the small print at the back of the document learned that Verizon's reported earnings included $2.7 billion in gains from its pension fund investments -- profit that didn't really exist.
The company pension fund actually lost $3.1 billion in 2001, a footnote on page 58 of the 68-page report revealed. ...
Csd does not advocate inventing new operating systems when there are acceptable operating systems - especially Forth and BASIC-52, of course!
Likewise there is no sense reinventing analog circuits.
While Bill is not into analog circuit design, he works with people who are. And naturally he got curious about what they are doing. Here's two neat books if you want to learn the rudiments of analog circuits.
Hey, analog circuit designers borrow circuits of others. No sense reinventing!
Bill was told that Forrest Mims lives and Albuquerque and works for the Air Force Weapons Laboratory. Companies like Analog Devices, Linear Tech, Burr-Brown ... offer seminars to teach about their products. And frequently publish these informative manuals. If you come-up with a hardware design, these companies also have engineers who will look at your design and point out what you are possibly doing wrong and a fix! Free! |
Don't count on higher earnings. So far, during this 'recovery' stage, profits have been falling - an unprecedented experience. And there is not much reason to think they will get much better soon. Businesses have already cut expenses. Expense cuts produce quick increases in profit margins for an individual company. But one companies expense is another's income...so the net effect throughout the economy is negative. What produces profits is capital investment...of which there has been very little. Companies typically build new factories, hire new workers, and sell new products at a profit - that is what gives them earnings to distribute to their shareholders.
But policy makers have encouraged consumer spending, hoping to hold off a worse recession. This consumer spending by Americans has done wonders for the Chinese economy - currently expanding at an 8% rate. But it merely deprives the U.S. economy of the savings and capital investment it needs to produce profits.
The amount of retail PCs sold over the 2002 holidays dipped from an already dismal 2001, with desktop computers sliding 15 percent to 20 percent, according to preliminary estimates from research firm NPDTechworld. ...
There were bright spots. Notebook computers continued to gain popularity as their prices fell, with retail sales units rising 10 percent to 15 percent, according to NPDTechworld's preliminary figures. ...
With those lower prices, of course, comes shrinking profit margins. Meanwhile, businesses and consumers remain spooked by the threat of higher gas prices, war and an unstable stock market.
Fundraising in the international capital markets fell sharply in 2002 as issuance was hit by a weak global economy, falling stock markets, rising corporate bankruptcies and political instability.
Securities underwriting volume totalled $3,900bn for the year, down from $4,110bn the previous year, according to new data from Thomson Financial.
Guys, Tice at Pudentbear are publishing negative economy articles because they are selling short.
A rising number of U.S. mutual fund managers, chastened by steep stock declines, are using a tactic once reserved mainly for the riskier world of hedge funds -- they're selling stocks short on a hunch they will fall.
ADVERTISEMENT "After really tough times, people are realizing something we believed was prudent all along," said Paul McEntire, who manages the Marketocracy Technology Plus Fund and uses the strategy. "It's not a bad idea to do some hedging in adverse markets. It does allow you to play both sides of the street."
Deflation?
Pioneer is struggling mightily to keep its Tijuana factory complex open in the face of fierce competition from China, which has undercut Mexico as a cut-rate producer. The Japanese electronics company already has slashed 900 jobs at its three plants here, moving the manufacturing of smaller speakers to Shanghai, where they can be churned out for as little as $3 each -- half of what they cost to make in Mexico.
"China's cost competitiveness is very strong," said Hirokazu Tsujimoto, another Pioneer executive who makes his home in nearby San Diego. ...
Since 2000, the Mexican government estimates, the maquiladora industry has lost nearly 250,000 jobs and seen its roster of plants shrink to about 3,200 from a high of more than 3,700. Among those that have shuttered operations are Hasbro Inc., Sanyo Electric Co. and Canon Inc. Landing another good position "will be very difficult," said 50-year-old Victor Hernandez, an engineer from Monterrey who was a manager at Canon's printer plant here.
Saturday January 4, 2003 20:21
Heads up, Wall Street. If the market doesn't rally this month, stocks may be on track to post their fourth consecutive losing year.
The month of January is an incredible forecaster of things to come for the market. It may predict whether the year will be good or bad.
On Thursday, the first trading day of 2003, the S&P 500 index jumped 3.3 percent, its best start to a new year since 1988, according to MarketHistory.com. Both the Nasdaq and the Dow Jones industrial average ended up more than 3 percent, with the Dow scoring its third-biggest opening-day percentage gain since the average was created in 1896.
In our view, risk remains to the downside for consensus 2003 estimates [for technology companies] while tech valuations remain at lofty levels relative to the S&P 500, Goldman's analysts said. ...
It found that the outlook for spending in 2003 had deteriorated to a 1 per cent decline compared with estimates of 2-3 per cent growth in a previous survey conducted just two months ago.
We are surprised by themagnitude of the decline, Goldman said.
The lack of game changing technologies and little pent-up demand indicates a bleak beginning to 2003 for technologycompanies. It also dashes hopes that IT spending in 2002 hadstabilised and would offer moderate growth in 2003.
This has led to a price war in some IT sectors as tech companies have been forced to compete by discounting prices. The Goldman survey indicates that the discounting of IT products is likely to continue, with 56 per cent of respondents saying that discounting is on the rise.
The technology shift caused by USB 2.0 and gigahertz PCs, Visual Basic and C++ 6.0 coupled with super fast 80C32 SOCs is doing to do bad things to powerful peripheral microcontrollers where Java is one software solution. So it will be interesting to see is Java is going to do about as far as its underlying technology, Forth.
Look what Mentor graphics is doing with USB 2.0 and 8051s!
The Java platform was introduced by Sun in 1995 as a way for programs to run on all computers regardless of the operating system.
But Java, which has been criticized for being slow, faces stiff competition not only from Microsoft but also Macromedia Inc.'s Flash, which is less robust but considerably easier to program than Java.
The weak economy isn't just putting the squeeze on people on the unemployment line - it's also hurting recruiters and consultants who used to command big salaries to fill jobs that have dried up.
Analysts say the stalled economy has sapped the recruiting industry, replacing double-digit revenues its enjoyed through much of the 1990s with double-digit losses, layoffs and lost firms. Overall, analysts estimate the industry has lost $3 billion in revenue and as many as one out of three firms in the past two years.
The bankruptcy business is booming in Chicago, even as the city's hometown airlines struggles to emerge from the biggest such filing in aviation history.
Delaware long has been considered a prime spot to file, partly because many companies are incorporated there and because attorneys say judges there were easy to work with.
However, ``Delaware as a venue became unfavorable because of its backlog,'' said David Newby, a Chicago attorney for property owners leasing to Kmart. ``Ironically, it became too popular.''
Thursday January 2, 2003 11:41
Others point out that many companies whose share prices plunged have reported lower profits, and that their stocks still trade at higher multiples of earnings than in pre-bubble years. That is particularly true of large technology stocks
Respondents to the Goldman poll now see 2003 IT spending falling 1% instead of the 2%-3% average gain they previously saw. As a result, Goldman argued that current multiples among technology companies remain "lofty" on a relative basis, particularly with no obvious buying catalysts on the horizon.
Two-thirds of the managers surveyed think reduced IT spending is more likely, while 56% expect discounting to rise and 15% expect spending to pick up in the second half of 2003, the survey found.
Bill is not an analog engineer. But he worked with analog engineers at Sandia lab.
One of his job assignments was to evaluate a/d converters for Sandia labs' seismic program.
Reason was that he could get the software written in Forth for a 8051 microcontroller controlling both the a/d and d/a converters at a fraction of the time using other technologies!
So will pass along information gleaned from a/d experts on how to get the right, or at least consistent, answers and avoid random number components contaminating sensor values.
With the 80C32 USB and mixed signal SOCs coming to market we are entering a very exciting time in collection and processing, on gigahertz PCs of course - not weird software-starved peripheral microcontrollers, of sensor data.
Pricing power is possible with a technically superior product.
Wednesday January 1, 2003 20:51
The anticipated peripheral micrcontroller core shake-out this year could worsen the outlook for high tech.
After years of being the epicenter of hot initial public stock offerings, Silicon Valley was an IPO ghost town in 2002.
With investors shunning new stocks amid a third year of painful market losses, only six companies headquartered in Silicon Valley went public in 2002, raising $1.19 billion. That's a fraction of the $8.5 billion raised by 78 local companies in 2000, when the IPO frenzy peaked.
ARC could be a victim. However, ARC is apparently selling USB 2.0 IP!
Synopsys is apparently also selling USB 2.0 IP!
Ford Motor Corp. (NYSE:F - News), for instance, said in early December that it expects its pension fund to be underfunded by about $6.2 billion at the end of 2002.
At least nine times out of ten, the prediction du jour is growth, growth, rapid growth of some new-fangled technology, bullish predictions built on five year timelines....
Five year growth projections are bunk If it didn't work for the Commies in the '50s and '60s, why should it work for capitalists today? ...
Challenge the herd. Too many pundits and analysts end up following a single lead. Why? Because everyone else is at the same place. It's nice and warm to be in a crowd of like-thinkers. The dot.com boom to bust has numerous examples of one allegedly credible voice staking out a position, followed by a stampede of others. Don't be intimidated by an alphabet soup of letters. Some of the most wrong statements I've heard emerged out of the mouth of a Harvard MBAs
"Workers have less leverage at a point where the unemployment numbers are growing and the economy is weak," but health care cost disputes may still lead to strikes, said Harley Shaiken, a professor specializing in labor issues at the University of California at Berkeley. "From the point of view of many unions, what they see is an unraveling of what it has taken many decades to put together, which is the employer paying most of the costs of benefits."
Companies across the board are facing double-digit inflation in health care costs. The cost of health insurance rose 12.7 percent between the spring of 2001 and the spring of 2002, a survey by the Kaiser Family Foundation and the Health Research and Educational Trust found. That was the highest rate of growth since 1990, according to the survey, which predicted that the increases would continue.
In his 2002 Year-End Report on the Federal Judiciary, Rehnquist said no new bankruptcy judgeships had been created since 1992 although the number of cases filed has increased by more than 570,000 since then.
He said each bankruptcy judge now handles an average of 4,777 cases, compared to an average of 2,998 in 1992.
Rehnquist said the number of filings in bankruptcy courts grew 8 percent in the year to an all-time high of 1,547,669 cases filed. He said bankruptcy filings have risen 72.5 percent since 1993.
Tuesday December 31, 2002 15:18
For the year, the blue-chip Dow fell 16.7 percent, while the Nasdaq plunged 31 percent - the biggest loser out of all the major averages. The S&P 500 sank 23.3 percent, while the broader Wilshire 5000 tumbled around 22 percent. The Russell 2000 of small-cap stocks lost 21 percent.
Donald Selkin, director of equity research at Joseph Stevens, called 2002 an unmitigated disaster. "By all statistical accounts, it was a historically awful year," he said, noting that this was worst decline for the S&P since 1974 and the Dow since 1977.
The Morris Township, New Jersey-headquartered company has contributed $800 million in cash and stock to its pension plans in 2002 to improve the funded status of those plans, said Honeywell in a statement. Honeywell said in a November filing with the Securities and Exchange Commission that its employee pension plans would end 2002 with a $1.7 billion deficit without new funds.
Many pension plans that have defined benefits to be paid to retirees have developed deficits in recent years as the value of equities fell since 2000. Federal law requires companies to protect the solvency of pension plans, which prompted many companies to increase funding to cover future retiree benefits.
We're going to try to reduce the size of this page and at the end of 2002 and try to kill, one way or the other, improper digitization.
We're thinking how to do this the most succinct way!
Various people had their various methods for destroying vampires as well. In some Slavic nations a spike made of ash wood, if driven through the chest, was believed to kill a vampire -- this is everyone's favorite method, a stake through the heart. In different lands, however, the wood used sometimes needed to be from a certain tree. For example, oak wood did the job in Silesia, while hawthorn wood was required in Serbia.
The federal agency responsible for failed pension plans is discontinuing a supplemental monthly benefit for many retirees of GST Steel Co., which closed last year.
The Pension Benefit Guaranty Corp., which took over GST's pension plans in August, informed several hundred GST retirees that the agency does not guarantee supplemental benefits payments and thus will be discontinuing them as of January. Normal monthly payments will not be reduced, according to a letter the agency sent to one GST retiree.
An official with the United Steelworkers of America said he was still trying to get some clarification and additional information from the agency on what benefits are being terminated. "Our retirees received letters saying their pension supplements will be stopped in part or in whole for some people," said John Wiseman, international staff representative for the Steelworkers union. "It's my understanding that some of our people are over the guaranteed maximum (from the agency) because they retired at an earlier age."
After the unbridled gusto of the dot-com era, the ensuing downturn has been a throbbing headache. Workers in high-tech went from hoping to be millionaires to running out of unemployment benefits.
Monday December 30, 2002 20:30
U.S. public companies shattered bankruptcy records for a second straight year in 2002 as accounting fraud and the previous decade's spending spree felled a number of corporate heavyweights -- and experts are bracing for more filings in the coming years.
A total of 186 public companies with a staggering $368 billion in assets filed for bankruptcy in 2002, according to tracking service BankruptcyData.com. That is the largest asset total ever, sweeping past last year's record $259 billion.
The wreckage included five of the 10 largest bankruptcies ever, led by phone company WorldCom Inc., with $104 billion in assets. Filings by Conseco Inc., Global Crossing Ltd., Adelphia Communications Corp. and UAL Corp. also were among the top 10. Accounting scandals figured in the failure of all of those but UAL, which was hurt by the Sept. 11 attacks.
Just The Beginning
-- December 30, 2002We have been issuing warnings about the looming bankruptcy crisis for months now. Years of exaggerated earnings and ever-increasing piles of debt are beginning to take their toll -- crushing companies and forcing them into bankruptcy. But it's just the beginning. We expect several large companies and hundreds of smaller ones to declare bankruptcy in the coming months.
And it's not just the failed companies and their shareholders who take the hit when bankruptcy is declared. The ripple effects are vast and devastating, reaching the companies' creditors and suppliers.
As this bankruptcy avalanche continues into 2003, it will continue to strike the earnings of good and bad companies alike. Bottom line: Nothing can kill earnings more than the aftershocks of large and multiple bankruptcies.
Secondly, I'm afraid that deflation is going to be the big story coming up. Deflation in the face of the massive amounts of debt that has been built into this country can be an horrendous problem. Remember, high levels of debt are difficult enough to deal with during slightly inflationary times. But in deflationary times, high levels of debt = bankruptcies.
All told, 186 public companies with a staggering $368 billion in debt filed for bankruptcy in 2002, according to tracking service BankruptcyData.com. That is the largest asset total ever, sweeping past last year's record $259 billion.
The wreckage included five of the 10 largest bankruptcies ever, led by phone company WorldCom Inc., with $104 billion in assets. Filings by Conseco Inc., Global Crossing Ltd., Adelphia Communications Corp. and UAL Corp. also were among the top 10. Accounting scandals figured in the failure of all of those but UAL.
Bankruptcy experts are bracing for a new crop of failures by companies that depended on companies that went bust.
Deflation?
Austin American Statesman Friday December 27, 2002
DEFLATION VIGILANTES
By James Grant
Bulls believe, as we do not, that the page has been turned on the Age of Inflation. They interpret the persistence of deflationary symptoms - 99-cent Whoppers, free cell phones and $500 PCs - as evidence of a new phase of our price history. According to the historian David Hackett Fischer, grand cycles of inflation and deflation carry everything before them. They are more powerful even than central banks.
Stocks are still preposterously high...with the S&P trading at near 30 times earnings. The late, great bull market was the biggest ever...shouldn't the following bear market be equally impressive?
Profit margins are the thinnest since the Great Depression. And debts are the fattest they've been in many decades...perhaps ever. In the manufacturing sector, interest expense has risen from just 23% of profits in 1997 to almost 100% today. ....
"2002 was an amazing year and 2003 promises to be even more exciting!" Geech!
At this stage of the game in the 145th week of its devastating Great Bear bust the NASDAQ is down about 73% since its bubble top, sickening by any standard. At the same week 145 the Nikkei was only down 56% but the infamous Dow 30 of 1929 had plunged by a catastrophic 88%. The raw magnitude of these enormous losses almost defies belief. If investors havent done their homework and happen to get trapped in one of these supercycle busts, their capital doesnt recover for decades after, if ever. ...
But, if you are one of the tens of millions who has already painfully rode the kamikaze NASDAQ down 73% in the past 145 weeks since its March 2000 top, you have an unenviable Herculean task ahead of you. To recover your capital, just to get back to square one, from a 73% loss requires a monumental 270% gain. 270% gains dont just grow on trees and are extremely challenging to track down and bag. ....
A supercycle Great Bear bust has a single mission in life. It exists solely to destroy the manic euphoria rampant in general stock valuations during a bubble and maul stocks back down towards undervalued levels so the whole great cycle can begin anew like a phoenix from the ashes. While equity valuations on average run about 14x earnings over decades and centuries, they are bid up ridiculously high in a bubble when everyone lusts for stocks and they plunge to silly lows at a supercycle bear bottom when virtually everyone hates stocks with a passion.
Venture capitalists who were optimistic last year about an improving landscape in 2002 have long since recanted and are singing a much darker tune as they look ahead to the coming year. Paul Ferri of Matrix Partners sums up the current mood with his characteristic frankness: ''This is not going to be a great period for anybody in our business.'' ...
Look for some high-profile tech infrastructure crashes in 2003, Beach says.
According to its annual report released in March 2002, Verizon Communications Inc., the largest U.S. local phone company, had a strong year in 2001. In the opening pages of the report, the company announced an annual profit of $389 million.
Only those investors who dug into the small print at the back of the document learned that Verizon's reported earnings included $2.7 billion in gains from its pension fund investments -- profit that didn't really exist.
The company pension fund actually lost $3.1 billion in 2001, a footnote on page 58 of the 68-page report revealed.
In reporting gains it hadn't made, Verizon didn't violate any rules. Like other U.S. companies, Verizon was following accounting practices as written in 1985 by the Financial Accounting Standards Board, which sets U.S. accounting standards.
The FASB rules say that in preparing income statements, companies should include estimated gains -- not actual gains or losses -- from pension fund investments. ....
As the stock market plunged during the past three years, the pension funds of companies in the Standard & Poor's 500 Index lost more than $200 billion in value, according to studies by actuaries and several investment banks, including Credit Suisse First Boston and UBS Warburg LLC.
Because of FASB accounting rules, many of those losses weren't reported on balance sheets.
If pension liabilities had been counted in financial statements, aggregate earnings for the S&P 500 would have been 69 percent lower than the companies reported for 2001, or $68.7 billion rather than $219 billion, the CSFB study found.
"We expect a big, quick shift to cash-balance pension plans when that happens. There's enormous pent-up demand among employers who have been holding back because of a lack of clarity."
While pension advocacy groups continue to study the new regulations, many say they've already concluded that the proposals fall short of offering adequate protection to older workers.
Sunday December 29, 2002 20:21
The cutoff Saturday came because Congress failed last month to reconcile differences between a Democratic plan for a $5 billion extension of benefits and a Republican plan for a $900 million extension. Bush declined to take a position on an extension of benefits until two weeks ago, when it was announced that November's unemployment rate had reached 6 percent, the highest in more than eight years
"Even in such a time of madness as the late twenties, a great many man in Wall Street remained quite sane But they also remained very quiet. The sense of responsibility in the financial community for the community as a whole is not small. It is nearly nil. Perhaps this is inherent. In a community where the primary concern is making money, one of the necessary rules is to live and let live. To speak out against madness may be to ruin those who have succumbed to it. So the wise in Wall Street are nearly always silent. The foolish thus have the field to themselves. None rebukes them." ...
As noted, all this might logically be expected. It may not come to pass. This is not because the instinct for self-preservation in Wall Street is poorly developed. On the contrary, it is probably normally and may be above. But now, as throughout history, financial capacity and political perspicacity are inversely correlated. Long-run salvation by men of business has never been highly regarded if it means disturbance of orderly life and convenience in the present. So inaction will be advocated in the present even though it means deep trouble in the future. Here, at least equally with communism, lies the threat to capitalism. It is what causes men who know that things are going quite wrong to say that things are fundamentally sound."
Let's hope men of reason arise to the occasion.
Here are some ideas to guide you through 2003.
Do what you need to stay employable. Fight hard to make your work maximally relevant to your company and customers. Attack problems and avoid complaining.
Be prudent with your expenditures, and teach your children the same habits. Contribute, don't complain.
I just got my hands on this year's version. Here's a choice piece: "[Associate Editor Marcia] Vickers predicts that the Standard & Poor's 500-stock index will end 2003 with a percentage increase in the mid to high single digits."
She's an editor at a renowned business publication, so she must have some insight, right? Well, on a lark, I went back a few years to see what BusinessWeek said in past issues.
December 1999:
"... the 51 analysts we polled this year expect limited gains in 2000. The average prediction: an 8.3% gain, to 12,154, for the Dow; a 10% rise, to 1559, for the Standard & Poor's 500-stock index; a 5.1% increase, to 3805, for the Nasdaq;"My review: The Dow closed 2000 at 10,788; the S&P 500 at 1320; the Nasdaq at 2470 -- losses of 6%, 10%, and 40%, respectively.
December 2000:
"According to BusinessWeek's survey of 40 investment strategists, 2001 will be a year of controlled growth. The average prediction: a 12% rise, to 12,015, for the Dow Jones industrial average; a 14% rise, to 1,558, for the Standard & Poor's 500-stock index, and a 23% jump for the Nasdaq Composite Index."My Review: The Dow closed 2001 at 10,021; the S&P 500 at 1148; the Nasdaq at 1950 -- losses of 7%, 13%, and 21%, respectively.
December 2001:
"Indeed, the seers' predictions for 2002 are downright moderate. They are, on average, looking for a 13% rise, to 11,090, for the Dow Jones industrial average; a 15% increase, to 1292, for the Standard & Poor's 500-stock index; and a 14.5% jump, to 2236, for the Nasdaq Composite Index."My review: As of the close on Dec. 26, 2002, the Dow closed at 8432; the S&P 500 at 889; the Nasdaq at 1367 -- losses of 15%, 22%, and 30%, respectively.
There's no reason to take the assumptions of BusinessWeek's experts as anything but wild guesses, and not particularly good ones. I won't fault BW for being optimistic -- to a point. The long-term trend of the stock market is up, and at some point, the market will get better. It's the fact that they still have the temerity to think they can predict when this will happen that I find galling.
A lot of companies were able to generate better earnings by cutting prices to attract customers while offsetting the negative drag from lower sales prices by firing workers. How long can they pull it off? The answer is probably not much longer before their strategies come back to bite them on their bottom lines.
In a recent note to clients, Paul McCulley, managing director at the bond mutual fund company Pimco and one of the nation's most savvy Fed watchers, wrote that the central bank's new focus on deflation amounts to nothing less than a regime change.
Sunday December 22, 2002 07:09
Economists at leading investment banks believe growth in the euro area slowed to a standstill during the final three months of 2002, and will contract early next year, the Business newspaper reported. ...
"We cannot even exclude an outright recession, two or three consecutive quarters of negative GDP growth."
Saturday December 21, 2002 20:39
Alan Greenspan's speeches are economic poetry. Each word is carefully chosen.
So it was exceedingly interesting that the chairman of the Federal Reserve Board used the first section of his speech to the Economic Club of New York Thursday night to declare "the United States is nowhere close to sliding into a pernicious deflation."
That he even mentioned deflation is telling. The possibility of a generalized price decrease is beginning to concern many economists, so the Fed chairman had to talk about the subject.
But parse the words of poet Greenspan. What he said is that deflation may show up, but if it does it will not be "pernicious," that is, it will not be fatal or deadly.
Later, Greenspan reassured his audience, "If deflation were to develop, options for an aggressive monetary policy response are available." In other words, don't worry, if deflation comes, we can handle it.
Pierre Belec strikes again!
But so do we at pro se fights.
Don't hold your breath waiting for the stock market to hand you a pot of gold next year. Deflation -- an earnings-eating monster that robs companies of the chance to raise prices -- may hold a gun to corporate America's head. ...
In today's intensely competitive environment, chief executives have lost their pricing power. The pricing issue has meant earnings problems for a slew of companies. And, if profits can't grow, how can the stock market recover? ...
"There is historical precedent that the economy exhibits a lack of pricing flexibility following those once-in-a-generational changes in the economy," Engelke said, referring to the boom of the late 1990s that was followed by the bursting of the speculative bubble in 2000. ...
The economy is still not out of the woods yet, and at this stage of the game business spending will decide the direction of the economy. The problem is businesses have become much more dependent on the stock market to finance their growth. ...
So the prospect for 2003 doesn't appear to be good. Investors who have been anticipating a return of the good times, may be in for a rude awakening if their rosy outlook does not materialize. We're talking stock portfolio survival after three straight years of awesome losses.
So here we are, 160,000 unhappy Direct TV DSL customers left hanging. Other than being told we have thirty days the company has said nothing else or provided any guidance to its soon to be former customers. Considering that Direct TV is going to continue operating, at least that is what they have told us. You would think that they would have planed a less painful transition as not to sour relations with current and potential satellite TV customers. I know that if I had satellite TV with them I would have canceled it because of the shoddy way they are handling this.
I understand that companies can't keep operating money-losing operations, but I think they should have the decency of letting customers know they are going to discontinue them so they have time to find adequate replacements instead of just shutting the doors out of the blue like Direct TV has done.
High-tech manufacturing jobs have been leaving the United States for years in search of lower-cost labor abroad. Computer companies can no longer compete on price without relying on inexpensive labor offered by countries like China.
As Silicon Valley endures a protracted economic slowdown and waits for the ``next big thing'' to propel it forward, intellectual work may now be following manufacturing jobs in search of low-wage workers. ...
It's likely that Silicon Valley will continue to be an expensive place over the next several years, prompting businesses to continue to move jobs and expand overseas. Will young engineers, visionaries and entrepreneurs come here, as the last generation routinely did in the '80s and '90s? Or will they find that they can adopt the valley's essential values and culture without being physically present here? Will technology and an open global economy allow future entrepreneurs to create virtual global companies that have no particular stake in Silicon Valley?
In other words, will Silicon Valley continue to be relevant as a place?
Bill got a letter from Social Security today.
1 Your Social Security benefits will increase by 1.3% in 2003.
What If I Work? If you are "full" retirement age (age 65 in 2002) or older, you may keep all of your benefits no matter how much you earn. But if you were under age 65 all year, there is a limit to how much you can earn before your benefits are reduced.
We paid you benefits in 2002 based on the amount of money you estimated you would make. When your employer(s) reports your actual earnings for 2002 to us, we will adjust your benefits if necessary The earnings your employer reports are the amount that will be on the W-2(s) you will receive. If the earnings on your W-2(s) for 2002 include money you earned in another year, you should contact us before April 15 to let us know. We'll also ask you to estimate your earnings for 2003, so we can pay you correctly What Are The Earnings Limits In 2003?. The earnings limit for workers who are under "full" retirement age (65 and 2 months in 2003) is $11,520. The earnings limit for people turning 65 and 2 months in 2003 is $30,720. If you earn more than the earnings limit, your benefits will be reduced. |
Debt and pension funds may do bad things to corporate profits in 2003. On the other hand, some senior citizens might be in big trouble!
Lawyers may target their pension funds. But we're working on the lawyer industry problem!
Mr Greenspan, speaking to the Economic Club of New York, said growth could accelerate next year with little risk of inflation, provided global tensions ease and corporate profits improve. However, he said the US economy was still working through a soft patch.
All in all, we believe that the rally off the October low has gone about as far as it can go and is showing clear signs of petering out. In our view the next significant move will be down.
Computers, fortunately, are leading in deflation. We buy computers with our pension checks!
Chairman Greenspan went on to say "Moreover, a major objective of the recent heightened level of scrutiny is to ensure that any latent deflationary pressures are appropriately addressed well before they become a problem." This statement confirms the statements of Governor Bernanke that the Federal Reserve intends to use the tools at hand that have historically (1930 - 1934) been used to stimulate economic activity when decreasing interest rates fail to push business activity forward as is possible, if not probable, now.
Now the Chairman says, " Although the US economy has largely escaped any deflation since World War II, there are some well-founded reasons to presume that deflation is more of a threat to economic growth than is inflation." This confirms to me that the Federal Reserve and the Bush administration will move to whatever is required to whatever degree is required in order to stave off the political implications of deflation. I have said before that deflation would not be entertained. The Federal Reserve and the Bush administration will burn the barn down before accepting the political implications of deflation. I have defined the barn as the dollar. I am now more than ever convinced that I was and am correct in this assumption. The US dollar on the USDX is headed, IMO, to between .73 and .80 as I see it.
Credit cards are the most obvious weak point. Hell-bent on growth, issuers pumped up solicitations from 1 billion in 1991 to 5 billion last year. With 37% of credit card loans now subprime, defaults have doubled since 1991.
Greenspan last night did make the comment, Cash borrowed in the process of mortgage refinancing is bound to contract at some point. Well, we will add that the degree of extreme Credit excess we are experiencing currently in mortgage finance is indeed bound to contract one day. Wild inflationary (speculative) Bubbles invariable do collapse. And as Greenspan stated correctly, History indicates that bubbles tend to deflate not gradually and linearly but suddenly, unpredictably, and often violently. One of these days there will be a dramatic shift in the financial and economic landscape. In this regard, think in terms of what Governor Davis faces now that inflated state receipts have collapsed along with the tech Bubble. That the economy and financial system are today struggling in the face of unprecedented mortgage Bubble excess and extreme Credit market speculation does not bode well for the day when the Next Shoe Drops.
Friday December 20, 2002 07:14
What we were's going to see soon is a single chip [an 80C32!!!] on a board surrounded by some capacitors with a USB connector and a second connector going to custom hardware.
Most of the work will be done in software. But in PC not in the 80C32.
The peripheral microcontroller companies have not solved their software problem. And didn't anticipate the damage that USB 2.0 will do to them.
USB 2.0, super fast USB 2.0 80C32s, other mixed signal 80C32s, Visual Basic and C++ 6.0, and Microsoft office is simply going to drive most peripheral microcontroller companies out of business - likely in 2003.
Internet is helping to bring to investors attention the high tech peripheral microcontroller scam.
So it is probable that investors will not invest in harebrained peripheral microcontroller companies thus speeding their demise.
Prior to Internet, trade magazine controlled technical opinion and advice. And could promote bad advice and technology for business reasons. Advertising.
No longer!
Speculative manias typically commence with a displacement which excites speculative interest. The displacement may come either from an entirely new object of investment or from an increased profitability of established investments. It is followed by positive feedback as rising share prices induce inexperienced investors to enter the stock market, and results in euphoriaa sign that investors rationality is weakened. During the course of the manias, speculation becomes more diffuse and spreads to different classes of assets. New companies are floated to take advantage of the euphoria, investors leverage their gains using either financial derivatives or stock loans, credit becomes overextended, scandal and fraud proliferate, and the economy enters a period of financial distress which is the prelude to the onset of a crisis although no two speculative manias are identical they develop along similar lines It is often said that speculation never changes because human nature remains the same the early stock markets were moved by hopes and fears as much as their later counterparts. These emotions are released during moments of speculative euphoria. They follow the path of least resistance, molding each mania, regardless of its historical context into common form. This explains why all great speculative events seem top repeat themselves
to avoid being shaken out.
Many who, for a brief season, had emerged from the humbler walks of life, were cast back to their original obscurity. Substantial merchants were reduced almost to beggary, and many a representative of a noble line saw the fortunes of his house ruined beyond redemption. - Charles Mackay
And get some neat products ready for 2004.
"We don't see the economic environment for material earnings growth having developed yet," said the portfolio manager and chief investment officer of the Signature Funds at Toronto-based C.I. Fund Management Inc. "As we get closer to 2003, we have had further negative . . . GDP growth revisions and further layoffs and this generally supports our view that we are not going to see the earnings recovery that is widely anticipated."
With an estimated $10-billion in net outflows through November, stock funds are poised to suffer the first negative flow year since 1988, when skittish investors cashed out following the October '87 crash, fund tracker Lipper said Thursday.
Lipper estimates investors pumped $10 billion into stock funds in November, snapping a five-month streak of outflows. And it's not likely December will reverse the decline.
"December is typically slow seasonally for fund buying, and stocks have declined a bit since Dec. 2," Lipper senior analyst Donald Cassidy said. "It seems unlikely the net inflow this month can make up ... the year-to-date shortfall."
The key factor to remember in assessing the economic and stock market outlook is that this is not a typical post-war bear market, but the unwinding of the mania that occurred in the late 1990s. The bursting bubble has left the economy with a number of structural imbalances that can be corrected only with further severe damage to the market, which remains excessively overvalued. The repeated occurrence of manias and their consequences is well recognized in economic history, and a number of economists have studied them in great depth. In his book, Manias, Panics and Crashes, Charles P. Kindleberger, discusses the Fisher-Minsky model, which attached great importance to the role of debt structures in causing financial difficulties, and especially debt contracted to leverage the acquisition of speculative assets subject to resale. Edward Chancellor, in his book, The Devil Take the Hindmost, summarizes the Minsky model as follows.
With war looming perhaps just a few weeks away ... soaring oil prices (over $30 a barrel and headed higher) ... and the soaring mountain of personal debt -- at a record $1.8 TRILLION now -- you can easily see the U.S. economy is in a heap of trouble.
"The Great Crash: 1929" by John Kenneth Galbraith is simply fantastic.
In a speech to the Economic Club of New York, Mr. Greenspan repeated past assertions that the United States was "nowhere near" falling into deflation. But he dwelt on the subject at considerable length in a somber speech that made it clear he is worried about the possibility. ...
Thursday December 19, 2002 11:54
The "semiconductor recovery" has to be one of the most enduring myths on Wall Street. About once a year, the myth emerges from hibernation, mauls a few investors and then returns to its den until the following year.
Back in the early fall, many Wall Street analysts predicted a recovery in the semiconductor sector. But, as usual, no recovery has materialized. Quite the contrary, as Apogee Research recently observed: "Novellus Systems Inc. (NVLS) is hearing increasingly glum news from the semiconductor industry, but you'd hardly know it from watching NVLS shares trade. Last week, the Semiconductor Equipment and Materials trade group (SEMI) said that total equipment sales this year will plummet 32.4% below last year's level of $28 billion.
The group forecasts a 'recovery' in demand during 2003, but the so-called recovery is for industry sales of $21.8 billion, which would still be 22% below the 2001 level. In fact, expectations for $26.4 billion of sales in 2004 and $27.5 billion in 2005 still wouldn't match the 2001 level. Yet NVLS trades at over five times trailing 12-month sales and at 3.3 times its peak fiscal-year 2001 sales. How can that be, we wonder, when the industry is not expected to revisit such levels for at least three more years?
The money did not evaporate. The money was stolen using high tech as the bait.
Rarely has one small company touched so many and left such a wake of ashes.
Two more once-promising Arizona companies have dot.bombed.
DevelopOnline Corp., the first Arizona spin-off from Intel Corp., sold its technology last month to a Massachusetts competitor and closed down. And Opnix, a start-up that raised $8 million, is soliciting offers for its software that speeds Internet traffic.
Both companies were founded in mid-2000 when the fervor for new Internet technologies was high.
India will introduce in two months a super computer capable of 1 trillion calculations a second at about half the price of such machines made by companies in the U.S. and Japan.
The super computer, Param Padma, has been developed at a cost of about 500 million rupees ($10 million), R.K. Arora, executive director, Center for Development of Advanced Computing, said in an interview. It plans to sell the new computer at about $5 million each.
``We are clearly cost-competitive, and we can sell these machines at the half the price of a similar machine available in the market,'' Arora said.
India's engineers are turning their lower salaries into a cost advantage in areas such as computing and aerospace. Indian software companies are winning business from customers such as General Motors Corp. because of their lower costs as code writers are paid less than their counterparts overseas.
And for those who think this is simply an extremely pessimistic and baseless viewpoint, consider this sobering thought offered to me this past week by my friend Dr. John Whitney. Wall Street may be happy enough to see companies it bankrolls move to a low cost environment like China, but from where will a lasting recovery come from when core manufacturing jobs jobs which actually produce wealth for a nation no longer exist in America? Wall Street has loved the strong dollar because it was a major factor in pumping up the bubble. But now we face the destructive aftermath. With aggregate demand plummeting, it is my contention that the forces of deflation will overwhelm the forces of inflation. Thus I remain a deflationist.
Conseco's bankruptcy filing on Tuesday reinforces what we've been telling you for quite some time now: The Bankruptcy Avalanche is underway. Enron, Kmart, Global Crossing, Adelphia, WorldCom, US Airways, United Airlines, and hundreds of other public companies have filed for bankruptcy in the past two years, according to BankruptcyData.com. In fact, 11 of the 20 largest bankruptcies of U.S. companies occurred during the last two years -- that's $395.9 BILLION in assets affected, based on court documents and corporate filings.
But the Bankruptcy Avalanche isn't over yet. We expect several more major companies -- and hundreds more small companies that you may not have even heard of -- to declare bankruptcy in 2003. We are currently monitoring these companies closely because it's only a matter of time before many of them go belly up. When they do, their shareholders will be the big losers. Plus, it will only add to investor concerns about the health of the economy and the safety of their own holdings. And when investors are uncertain, they invariably sell everything they can. The stock market sell-off we've been telling you about is already underway. Still, there's much further to go before we hit bottom.
Wednesday December 18, 2002 21:19
The stock market is headed for a serious wreck and only those with the knowledge to see the light for what it is will get off the track before its too late. Lets look at several issues we are facing. Wall Streeters (including brokers, money managers, analysts and commentators), will never mention these facts to you the investor. Many dont know themselves, because the spin-doctors feed them what they want them to know and then, most wont read additional information to educate the customer.
First, S&P announced on October 24, 2002 their study of earnings for the S&P 500 for the year ending June 2002. S&P calculated the reported earnings, instead of the operating earnings we are given. Instead of the operating earnings of $44.93 per share that Thompson First Call reported, the reported earnings were $26.74 a share. That is until the new accounting standards, which include the expense of options and pension liabilities. These new standards will be enacted next year by the accounting industry. Subtracting the option and pension liability expenses reduces the reported earnings to $18.48 a share. This is a Price to Earnings (P/E) ratio of 48.6 based on a close of 897.65 on the S&P 500. This is real important. I repeat this is real important. The 50-year average P/E ratio is 15.9. No bull market has ever started from such high valuation levels. Also, the market is trending back towards the mean. Imagine the drop in the market valuation that will occur to get to 15.9. That is a 67% drop in the market from where we are now. Imagine that train wreck! ...
Lets look at pension liabilities. There are over 350 companies in the S&P 500 with defined benefit pension plans. (If you dont have a 401K, but have a pension plan, it is a defined benefit plan.) These companies have guaranteed their retired employees a defined income during their retirement years (hopefully).
Goldman, Sachs & Co. estimates that contributions are grossly under funded. They estimate that companies need to add $80 billion to these funds this year. Credit Suisse First Boston estimates the under funding to be $243 billion this year. Morgan Stanley estimates it at $300 billion. Whichever, estimate you use, the numbers are staggering. Federal regulations require corporate pension plans to be at least 90% funded. If a plan meets this requirement, the additional payments can be spread over a 30-year period. But if the funding drops below 90% of what the actuaries say is needed, companies may be required to make the additional payments within three to five years. These additional payments will be expenses against income in that year. How will earnings grow enough to absorb this expense and still net a 6%-12% increase? So why hasnt the Securities and Exchange Commission required
Last year, companies reported a gain of $104 billion from their pension funds, when the funds actually lost $90 billion. Makes you wonder how do they turn a lost into a gain. The secret is a company will hire a pension fund consultant to estimate its earnings in the future years, and if the estimated earnings are more than the estimated pension liabilities in the future, that positive difference is added as earnings. If you want more earnings, you get a higher estimate. Do you see the train coming?
Senior citizens become very concerned with the economy. Pension checks might stop coming. Or inflation might set in.
There are definitely reasons to worry!
The scent of deflation continues to waft through the economy along with still-meek signs of recovery. ...
The nation's manufacturing is using just 75.6 percent of its capacity, about 6 percentage points below its average for the past three decades, said economist Andy Kish of Economy.com. ...
While the economy overall seems nowhere near deflation --- a broad decrease in prices --- the measure of core goods prices has slid for 11 consecutive months. And results have been mixed, according to Mark Vitner, senior economist for Wachovia Securities. ...
"For many goods producers, the current economic environment is already devastating profit margins," he said. "Businesses have not been able to pass on their higher costs to consumers for nearly three years now."
And because the service side of the economy is not deflating, businesses must also find ways to swallow --- or avoid --- added costs of health care and insurance, Vitner said.
See what happens with good products which cost a reasonable amount.
On Tuesday rival Toyota Motor also predicted it would have a record year in 2003, with a 5% rise to 6.5 million vehicles.
Then there are the other cases.
Looking for a robust global economic recovery? You may need to look past 2003 - and think 2004.
After a year in which U.S. corporate scandals, talk of a war in Iraq and a global plunge in stock prices torpedoed hopes for a strong rebound from recession, pundits are pessimistic about the coming year.
The global economy will speed up only slightly from the current becalmed levels. Growth rates may not return to the average levels of the last half-century until at least 2004.
"Economists are always predicting a recovery. We are condemned to expect a recovery," said Eric Chaney, an economist at Morgan Stanley in London. "But the coming year could still be difficult."
A recovery of any sort, of course, would be better than the alternative: a renewed slump into recession that would heighten fears of deflation. That condition - a sustained drop in prices that depresses business and consumer activity alike - is something Japan has lived with for more than a decade. Economists speak about the possibility of it spreading in guarded tones, for fear of turning it into a self-fulfilling prophecy.
Though global deflation still seems to be an unlikely, worst-case scenario, economic growth remains persistently sluggish in all of the world's biggest economies.
The Dollar's bear market is probably going to extend for many years, but as is always the case during any long-term trend there will, from time to time, be substantial counter-trend moves. In other words, if we are right that the US$ has embarked on a long-term bear market there will still be periods, possibly quite lengthy periods (6-12 months or longer), during which the Dollar will trend higher and those investments that benefit from a weaker Dollar will trend lower.
Here's what O'Neill wrote on the Treasury Department Web site: "Accrual based financial reporting is critical to gaining a comprehensive understanding of the U.S. Government's operations. For fiscal 2001, our results were an accrual-based deficit of $515 billion in contrast to a $127 billion surplus reported last Fall."
Let me put it more simply. If the government kept its books the same way companies are supposed to account for profits and losses, the U.S. would be in the hole last year for more than half a trillion dollars.
I'll say it again. That's a one-year deficit of more than half a trillion dollars.
Tuesday December 17, 2002 19:44
Guys, we have to keep system software costs low. And have some fun too using the ideas of others!
NIH is a disaster!
Conseco Inc. may file for bankruptcy protection as early as tonight to stem mounting losses and restructure more than $6 billion of debt incurred while the insurance and finance holding company acquired smaller rivals, people familiar with the matter said.
Tokyo's Nikkei hits one-month low on Micron's loss
Brokerage houses including Nikko Cordial Corp (Tokyo:8603.T - News) were also hit by selling as the benchmark Nikkei average (^N225 - News), down 1.63 percent or 139.01 points at a one-month low of 8,371.72, was poised to finish lower after having just snapped a nine-day skid.
Micron knocked U.S. tech stocks back in after-hours trade. Traders are betting the Nasdaq will take a tumble later today," said Ken Masuda, senior dealer at Shinko Securities.
Guys, others are saying this too.
But as important for USB 2.0 80C32 SOCs running BASIC-52 and Forth, hardware and software maintenance is possible over Internet using Netmeeting!
'Wage Collapse', reads a headline in Newsday.com. The articles details the stories of three people in the New York area, who've joined the growing ranks educated workers, formerly employed in well-heeled positions, "filling salt shakers, selling sweaters or cell phones, making telemarketing calls, imputing medical billing information - even passing out fliers on street corners."
"People who have jobs had better be glad they have jobs," says Charlene Ruiz, a former consultant with US Trust Co. now working for $7.25 as a cashier in Manhattan cafe. "If I continue to work at this wage," says Ruiz, "I'll never get out of debt." She owes, says the article, $20,000 to credit card companies...and another $4,000 to the IRS. The Bureau of Labor Statistics reports that country the number of people working part-time jobs who would ordinarily be in high-paying salary positions has lept up to 18.3% of all part-time workers - or...4.2 million Americans.
"This is just fall out from the boom," says our own Dan Denning, editor of Strategic Investment. Its the beginning of the long bust, where some Americans will see their standards of living fall, their incomes stagnate, and their prospects for anything like retirement dwindle. The labor market is telling us something: America is just beginning to pay for its over-consumptive sins. Americans might not like it. And who knows what they might do about it...
Patty's father was born in 1912 and, too, has some wild stories of the Great Depression.
He was in the CCC fighting forest fires in Idaho.He told bill last March if a depression were to occur today he thinks there would be major riots!
Kay said the danger is that PC prices always drop. ``If people know there is cheaper stuff coming, they might just wait,'' he said.
"...resources have been misallocated because of the cheapness of credit in both stock and credit markets. So, you're not going to solve the problem by making money cheaper again." - Al Friedberg, March 23, 2001
The net loss increased to $315.9 million, or 52 cents a share, in the quarter ended Nov. 28, from $265.9 million, or 44 cents, a year earlier, the company said in a statement. Sales rose 62 percent to $685.1 million from $423.9 million. Last quarter's loss was the company's eighth in a row.
Micron's selling price last quarter fell an average of 12 percent from the fourth period, mainly because of sagging prices for dynamic random-access memory chips, the company said. Competition from rivals such as Samsung Electronics Co. and slowing growth in sales of PCs using DRAM are creating a glut for the chip, an analyst said.
The spot price for a benchmark 256-megabit double data rate memory chip was $8.30 in March, fell to $4.78 in June and is now about $6.45. Micron bases sales on the average price per megabit, or million bits of capacity in a chip.
Tracking the high tech economic meltdown, while reading Axelson and Hyde USB books, is nearly as interesting as litigation!
The ability of companies to make too much has grown into a problem with potentially profound and unpredictable effects on the world economy. The glut of world production is challenging old assumptions about trade relationships and raising questions about whether the United States can continue to be the economic powerhouse it has been for generations while so much of its manufacturing capacity, especially high-tech production, is increasingly shifted overseas.
I know there are millions of others in my leaky boat: Last year alone, the average 401(k) account balance slumped from almost $41,000 to $36,390, according to one research firm. Wilshire Associates, an investment advisory firm based in New York, says stocks have lost about $8 trillion in value since peaking in March 2000 -- an amount equal to the annual economic output of Japan, Germany and Canada combined.
We have to try to keep those 80C32 SOC systems software costs down.
Credit ratings agency Standard & Poor's said on Tuesday that the dollar amount of debt defaulting will vault to a record $157.3 billion this year. But the agency said default rates are declining globally after peeking this summer. Telecoms and European companies have been hit hardest during 2002, but prospects look better for 2003 as default rates are expected to decline gradually throughout next year. "The U.S. and EU switched places this year compared with 2001. During 2001, 6.92 percent of speculatively (or "junk") rated companies defaulted in the EU compared with 9.72 percent in the U.S.," said Brooks Brady, Associate Director for Standard & Poor's Risk Solutions
Robert Half Technology's 2003 Salary Guide (www.roberthalftechnology.com) predicts that information technology salaries will decline 1.3 percent on average in 2003.
Next year's U.S. gross domestic product growth in this so-far jobless recovery will be roughly the same as this year's 2.5%. But deflation tops the list of his concerns -- "If you don't see deflation risks, my advice is to go see an optometrist."
That's not his only worry. The strong housing market that has propped up the U.S. consumer's spending is showing signs of fatigue and the U.S. dollar, which has stayed relatively strong, could come under greater pressure that would be felt in stock and bond markets.
But not now. Schembry, 58, joined hundreds of other Lucent workers this year in taking a buyout at the company, one of many telecommunications companies to face a sudden slowdown after extensive overbuilding in the industry in the 1990s
The future looks daunting. Her husband, 61, is disabled. The $58,000 buyout package won't go far without a job. Her 401(k), once valued at $30,000, has dwindled to $3,000 because of a plunging stock market.
"At my age, where am I going to find a job?" she asked.
Workers age 50 and over are putting early retirement plans on hold because of the losses they've suffered on Wall Street, according to a new poll by AARP, the lobbyist organization for retired persons
Seventy-seven percent of those surveyed said they had lost money in the stock market since 2000. While 54 percent said they had losses of less than 25 percent of the money in their accounts, 9 percent said their investments had fallen by more than 50 percent.
The survey of 1,013 households with individual retirement accounts, 401(k), pension and stock investments also found that 77 percent had lost money in the stock market over the past two years. Thirty-four percent said they would take fewer vacations and 30 percent had postponed a major purchase in the past year. The survey, released today, also found that one-third of those who retired before 2000 had returned to full or part-time jobs to compensate for their stock losses. Among workers 50 to 70 years old who said they lost money in the stock market in the last two years, about one in five said they had postponed retirement because of the losses, according to a survey released today.
Monday December 16, 2002 20:17
We have a good practical problem to solve in 2003.
USB 2.0 80C32 SOCs running BASIC-52 and Forth connected to Visual Basic and C++ 6.0, Microsoft Office apps!
And lots of fun too! A positive contribution to computing.
Jon Johansen, known in Norway as "DVD Jon," is charged with having unlocked a code and distributed a computer program enabling unauthorized copying of DVD movies, angering Hollywood studios who fear mass piracy and loss of vast revenues.
The crypto problem was solved in 1917. But that didn't stop the retards.
The one-time pad is a series of random numers/letters that can be used once to encode a message. It was first described by Mauborne and Vernam in 1917. Each new message requires a new pad, otherwise it will be possible to break the code. Also, each one-time pad must be truly random series of digits/letters (a computer generated random series is only pseudo random and can be decrypted eventually). Given now a one time pad (where, say, the symbols are digits 0-9) and an encoded message (also using symbols 0-9), a simple table that links every message symbol and one-time pad symbol to a cypthertext symbol can be used to encrypt the message (this table is a more sophisticated alternative to the binary XOR operation that is normally used for bit-stream data and one-time pads):
Swiss banks which accept electronic payment orders use a similar mechanism to verify the identity of the issuer of an order. When a client authorises the bank to accept electronic orders, the bank delivers, by registered mail, a list of individual session passwords (usually numbers of 4 to 6 digits). The customer agrees to keep the list of passwords physically secure, and to not hold the bank liable if the customer allows the list to fall into unauthorised hands. Each time an order is given, in addition to the regular user identity and password, the next key from the list must be entered, and then crossed off by the user. The bank verifies the key against a copy of the list stored in their own secure computer, and only if the key matches is the order accepted. Multiple incorrect entries block electronic access to the customer's account until re-enabled by the customer providing suitable verification that an unauthorised access attempt did not, in fact, occur.
Patty in the Hauptbahnhof in Zurich.
Bill gave a talk in Zurich on the dangers of high tech espionage. Like getting caught! 1 2
ATLAST (Autodesk Threaded Language Application System Toolkit) is a (very) FORTH-like language kernel designed to be embedded into applications, rendering them extensible to a degree far beyond normal macro languages. Indeed, using ATLAST, it is often possible, in the spirit of FORTH, to "factor out" much of the control structure from an application, reducing it to a set of data-driven services whose interaction can be modified by the user with ATLAST. ATLAST is written entirely in portable C, and has been tested on MS-DOS, Windows, a variety of Unix machines, and the Macintosh. Complete documentation is included in PostScript, Microsoft Word (PC/Mac RTF format), and LaTeX source code form. ATLAST has been placed in the public domain and may be used without restriction or compensation.
Postcript is a variation of Forth.
Think USB 2.0 80C32 SOCs running BASIC-52 and Forth connected to Visual Basic and C++ 6.0, Microsoft Office apps!
There appears to be shake-out of unparalleled proportions underway in high tech in 2003.
USB 2.0 80C32 SOCs running BASIC-52 and Forth connected to Visual Basic and C++ 6.0, Microsoft Office apps!
As if a lousy business climate wasn't bad enough, semiconductor makers have more to worry about.
"The simple fact is that we won't be needing so many chip vendors in the future," Dataquest analyst Jim Tully said last week. "Many firms that are now fabless chip houses should seriously consider becoming IP [intellectual-property] vendors instead."
As for that next "killer app" for which the industry is holding its breath, Tully says forget about it.
"We are always asking, 'What will be the next killer application?' "he said. "The truth is there will be no more killer apps. There will just be a growing network of hardware resources onto which increasingly compelling software applications can be loaded."
We have to watch the system software bucks.
Atlanta gained possibly more high-tech jobs than any other city in the nation.
Across the South, the economic landscape has turned upside down. Atlanta, where all arrows once pointed up, is hurting most: 61,800 jobs were lost in the 12 months that ended in October more than in any other city in the country, according to the federal Department of Labor. And the job losses are continuing.
Even North Carolina's powerhouse area, Raleigh-Durham, which is normally all but immune to hard times thanks to a largely tenured work force of university and state government employees, has seen its unemployment rate double this year. In Hickory, N.C., and the surrounding area, entire fiber-optic cable factories have been mothballed in recent weeks, and 3,200 jobs have been cut since spring 2001.
But plummeting corporate profits one of the hallmarks of the national slowdown resulted in an average shortfall of 14 percent in corporate income tax collections, the study found.
Still, the most poignant illustrations of the downturn are the roughly 20,000 highly skilled but suddenly unemployed workers who were laid off from telecom and technology companies with headquarters or major operations in Atlanta. The companies include Cingular Wireless, WorldCom and Internet Security Systems, and even national consulting concerns like Accenture.
EEPROMs and Flash may get serious competition this year.
ferroelectric memory was chosen over EEPROM because it can offer high-speed rewriting.
Get rid of the old ideas and adopt the new good ones, of course.
USB 2.0 80C32 SOCs running BASIC-52 and Forth connected to Visual Basic and C++ 6.0, Microsoft Office apps!
After spending some $80 billion since 1996 to upgrade its systems, the U.S. cable television industry is just about finished with an overhaul designed to offer high-speed, digital broadband services to nearly every U.S. customer. The good news is that a range of new offerings from video-on-demand to telephone service will now become more widely available. The bad news is that basic cable television rates have been rising much faster than prices for other consumer goods and services and theyre going up again next year....
Meanwhile, despite the heavy investment, new subscriber growth has stalled largely due to an onslaught by the direct broadcast satellite industry. Some 76.5 million U.S. households now get basic cable service, compared to some 19.7 million satellite subscribers. But the satellite universe is expected to grow to 21.8 million subscribers next year, according to Kagan World Media, while the cable industry is forecasting little or no growth. Some cable companies, including AT&T Broadband (which was recently bought by Comcast) and Charter Communications, have even seen their subscriber bases shrink.
Is cable going to become an obsolete technology?
At a time when Americans are living longer, more than half the paid workers ages 25 to 64 don't own retirement savings accounts of any kind. About a third work for employers who don't offer retirement benefits, a recent study shows. ...
"Our savings rates here are a disaster, and the study continues to beat home the crisis in savings overall," said Derrick Max, executive director of the Alliance for Worker Retirement Security, which represents business groups and favors overhauling Social Security to add personal retirement accounts. ...
Max said the report bolsters support for Social Security private accounts, which President Bush supports but Democrats oppose. Social Security is projected to start paying out more in benefits than it takes in from payroll taxes by 2016 because the large baby boom generation begins retiring. The workforce size will be smaller, and demand for benefits will grow.
A Web Of Fallacies About Consumption And Investment
"...A central bank should have no concern with the level of stock valuations, but it ought to have an indirect concern when rising stock values disturb the equilibrium between investment and saving. In the U.S. case, the decline of saving after 1997 was so dramatic and so spectacular that it ought to have alerted Mr. Greenspan and American policymakers in general. And net investment is close to historical lows..."
What the U.S. economy truly experienced in the past few years was the wildest and most reckless consumption boom ever in history, associated with collapsing saving and an extremely low level of net investment. But we think that low net investment is only part of the problem. Another part is a grossly unbalanced investment structure. Protracted, gross underinvestment in manufacturing has contrasted with protracted, heavy overinvestment in the new information technology, retail trade and finance. ...
The truth, rather, is that post-bubble adjustment has not even started.
Sunday December 15, 2002 21:23
Imports from Hynix Semiconductor Inc. and other South Korean chipmakers threaten to hurt Micron Technology Inc, the US International Trade Commission said in a ruling that may lead to import tariffs.
Micron, the world's second-largest maker of computer memory chips and the last one based in the US, complained that its Korean rivals were illegally subsidized by the Korean government.
A final ruling isn't due until the middle of next year. "[Saturday's] decision sends a strong message that the US will not tolerate unfair trade practices," said Sean Mahoney, a spokesman for the Boise, Idaho, company.
2The Nikkei <.N225> was down 0.81 percent or 69.27 points at 8,446.80, poised to post a nine-day losing streak for the first time since 1991.
The broader TOPIX index <.TOPX> of all first section issues fell 0.90 percent to 828.27.
In this new role, Mr. Desmond, 44-years old, will oversee the transfer of much of Time Inc.'s magazine Web content to America Online, part of a broader revamp of the online service announced earlier this month. Mr. Desmond previously was president of Business 2.0, a Time Inc. business and technology magazine. An announcement is expected as early as today.
Google may help all of us in this legal fight!
Pro se fights is number 3
Results 1 - 10 of about 4,860. Search took 0.34 seconds.
< Pro Se Fights ... Antonin Scalia Associate Justice United States Supreme Court One First Street, NE ... state lawsuit and that my rights to resprent myself pro se have violated ... www.geocities.com/CapitolHill/Congress/8327/ - 101k - Cached - Similar pages
Preserving our rights guaranteed under the constitution for right by trial by jury, especially for two prima facie cases, and to represent ourselves pro se against crooked incompetent lawyers is perhaps one of the most important fights of our lives.
But you may have guessed this. 2
Dee Vance Benson may be an honest judge? We'll see.
And maybe even Scalia too?
Dig this
What we are doing with gigahertz PCs, super fast 80C32 SOCs, and USB 2.0, Visual Basic and C++ 6.0, Microsoft Office, ... hopefully 80C32 SOCs running BASIC-52 and Forth is much more important than
5 T. G. Lewis , W. H. Payne, Generalized Feedback Shift Register Pseudorandom Number Algorithm, Journal of the ACM (JACM), v.20 n.3, p.456-468, July 1973
And what happened here and its resolution is more imporant than about anything.
Get this settled!
Atmel builds an 80C52. So we looked at the financials. Geech!
We are into a technology shift.
gigahertz PCs, super fast 80C32 SOCs, and USB 2.0.
Most important, software! Visual Basic and C++ 6.0, Microsoft Office, ... and hopefully 80C32 SOCs running BASIC-52 and Forth!
Hey, the peripheral microcontroller vendors know they don't have the software. Smell a hardware vendor scam? The pension funds again!
And, of course, Keil is handingly the software problem, they will say. TI, Cygnal, Cypress, ....
Those who get to market with 80C32 USB 2.0 SOCS, like TI, will do just fine when we come out of this high tech mess - maybe in 2004.
Then there are the others. Old high tech goes goodbye quick. Into the ewaste heap.
Keep in mind when USB connects to a perihperal 80C32 SOC, then RAM is The Way, not eeprom, flash, OTP ROM ....
Cypress saw this! No ROM with their design!
"Today's economy was probably experiencing a once-in-a-century acceleration in innovation, but people may well conclude that a good deal of what we are currently experiencing was just one of the many euphoric speculative bubbles that have dotted human history." - Alan Greenspan, January 2000
As a result of the re-ranking, the companies' stocks to be removed are Abgenix Inc. (NasdaqNM:ABGX - News) , Andrx Group (NasdaqNM:ADRX - News) , Applied Micro Circuits Corp. (NasdaqNM:AMCC - News) , Atmel Corp. (NasdaqNM:ATML - News) , Charter Communications Inc. (NasdaqNM:CHTR - News) , Conexant Corp. (CNXT), Cytyc Corp. (NasdaqNM:CYTC - News) , Integrated Device Technology Inc. (NasdaqNM:IDTI - News) , ImClone Systems Inc. (NasdaqNM:IMCL - News) , i2 Technologies Inc. (ITWO), Protein Design Labs Inc. (NasdaqNM:PDLI - News) , PMC-Sierra Inc. (NasdaqNM:PMCS - News) , Rational Software Corp. (NasdaqNM:RATL - News) , Sepracor Inc. (NasdaqNM:SEPR - News) and Vitesse Semiconductor Corp. (NasdaqNM:VTSS - News) .
Bill has been defamed too. Fortunately in writing! This is libel!
The libel broke both federal and New Mexico state criminal laws in writing!
The decision was another timely reminder that cyberspace gave no special immunity when it came to defamation.
Collecting for a prima facie case of libel is a bit more complicated in practice than in theory.
But this involves lawyers who make money, our taxpayer dollars, attempting to protect the crooks.
Pension fund money will, of course, be used to pay lawyers to protect the crooks who stole it!
The real world again.
It's a sign of bearish times: As the Nasdaq Stock Market did its annual reshuffling of the Nasdaq-100 Index Friday, analysts expected more than a dozen technology issues to be booted in favor of safer consumer and industrial companies. ...
"Literally, in the index, the weight of technology has declined," said Nicholas Gulden, an analyst with Salomon Smith Barney.
The Nasdaq-100 closed Wednesday at 1,039.95, down more than 75 percent from its 2000 high, largely due to the protracted slide in tech. ...
Gulden predicted chipmakers PMC-Sierra Inc. and Vitesse Semiconductor Corp. would be removed from the Nasdaq-100. He also expected troubled biotech ImClone Systems Inc. to be deleted.
2003 may prove to be a very interesting year, unfortunately.
COULD THIS DECADE BE THE NEXT 1930S?
A REVIEW OF WORLD STOCK MARKETS IN THE 1920S
But are looking good!
"The length and severity of depressions depend partly on the magnitude of the 'real' maladjustments, which developed during the preceding boom and partly on the aggravating monetary and credit conditions." - Gottfried Haberler, Prosperity and Depression, 1937 1
The credit bubble and pension fund woes may do bad things in 2003. Along with China and those neat Harbor Freight sales too!
So we must be conservative and not blow our money on expensive antiquated 80C32 development systems! Which aren't fun either!
Bill's father bought stock during the depression. He had some money. The army was paying him. By 1957 he was doing well in retirement. But not before that!
Bill's father in law said the economy really didn't get good until the 1960s.
Bill is scanning Axelson's second edition book. It is outstanding!
Saturday December 14, 2002 18:06
Here's some more economic BS.
Companies will, or should - which is not certain, send checks to their pensioners.
Investing in more speculative ventures should not,
but may be, in the future.
Beijing, Dec. 14 (Bloomberg) -- China sold a record $106 billion of government bonds this year to pay for roads, dams and other projects that fueled the fastest growth of any major economy.
Bond sales to domestic investors such as banks, insurers and individuals by the Finance Ministry and the state-owned China Development Bank and Export-Import Bank of China rose 16 percent to 877.4 billion yuan from 758.8 billion yuan last year.
China is boosting spending on railroads, airports and other projects, helping the economy to grow 7.9 percent in the first three quarters of this year from a year earlier. The government needs growth of more than 7 percent to absorb 8 million new urban job-seekers each year. With debt levels a fraction of those in Japan and the U.S., China can afford to borrow, analysts said.
Dig the underline.
Some are going to eat the Big Budungee Bird this year, aka bankruptcy.
Those who have too much debt. And didn't limit their software costs!
And, of couse, we're not going to have to spend too much bucks on 80C32 system software if all works out as planned.
Then there is the fun too. Over Internet!
Who is funding most of the Chinese student education in America?
Their parents.
Their parents will give up about everything to make sure their children do better than they do.
Mr Lam is a bill ms student. And hopefully doing well!
WASHINGTON -- In naming Stephen Friedman to a key post Thursday, President Bush finished a clean sweep of his bumbling economic team. The question now is: Will it matter?
The early signals are unpromising, and that's bad for the administration, the economy and the country.
Conservative activists gloat that they exacted a pound of flesh for Friedman's appointment, as well as for Bush's choice of rail executive John W. Snow to replace Paul H. O'Neill as Treasury secretary.
They say such administration heavies as political advisor Karl Rove were forced to assure them that both men, but especially Friedman, will put aside past worries about budget deficits and fiscal probity in favor of the White House's tax-cutting, "pro-growth, pro-jobs" agenda.
"They told us [Friedman] will be a solid salesman for supply-side policies," said Stephen Moore, who heads the conservative political action committee Club for Growth and spearheaded an anti-Friedman drive.
Bill's father was born in 1888.
He had some very interesting stories of the Great Depression. About bank failures in North Dakota.
He was in the Army during the depression supervising CCC workers.
Patty's father was born in 1912 and, too, has some wild stories of the Great Depression. He was in the CCC fightin g forest fires in Idaho.
He told bill last March if a depression were to occur today he thinks there would be major riots!
Patty's father, is going to be 90 on December, lives in Florida, of course,
He is buying a laptop!
There are old people, then there are the really old people in Florida. Bill feels like a teeny bopper when he visits Florida!
Then, ensuring that I wouldnt get invited to any Wall Street Christmas parties in 2001, I wrote a controversial essay published May 11th, 2001 called Equity Bulls in Denial with the audacity to brazenly claim that the bear market in the US is just beginning, NOT ending as most bulls and bullish-bears assert today. This current NASDAQ, DJIA, and S&P action looks and smells like a bear market rally, as there is no fundamental foundation for the spectacular gains since early April.
U.S. corporations have been struggling against a slumping economy and the bear market for about three years. But NOW, they're about to face an even tougher opponent -- deflation. In November, producer prices plunged 0.4% overall, and 0.3% if you exclude food and energy. Specifically, the wholesale prices of items such as cars, sporting goods, and computers fell sharply. And falling prices destroy pricing pressure -- businesses aren't able to raise the price of goods -- which then annihilates profit margins.
US satellite TV operator Hughes Electronics is to shut down its high-speed internet service. The loss-making internet service, DirecTV Broadband, has 160,000 subscribers and is linked to the DirecTV satellite broadcaster.
Friday December 13, 2002 21:35
China's economy has the highest rate of growth of productivity in tradables of any economy in the world today. At a constant remnimbi-dollar exchange rate, Chinese competitiveness has also improved by leaps and bounds. That means that the swing that we are now witnessing in Asian trade balances toward greater and greater surpluses will largely come at the expense of U.S. trade. For all of the current discussion about deflation, and China's alleged role in exporting it, a longer-term perspective would see such deflationary pressures as a direct consequence of an ill-conceived effort to sustain a dollar exchange rate well above what long-term fundamentals would have otherwise justified. In perversely pursuing this strong dollar policy since the days of Robert Rubin's tenure as Treasury Secretary, the U.S. has unwittingly catalyzed its own economic demise and hastened the rise of Asia's ascendancy, notably China. ...
Above all, China has proved resistant to Western ways. Its relatively closed borders to volatile Western leveraged capital (notwithstanding the pleas of Wall Street) enabled it to emerge from the regional crisis of 1997/98 relatively unscathed. As the recent data suggests, it is now well prepared for growth on its own terms (which is why one should not expect imminent capital account liberalization and a corresponding revaluation of the currency). It will be the economic leader in the global recovery. Asian development strategies will continue to deepen further as China's regional economic dominance expands. China is not the source of the world's current problems; it is not, as is commonly argued, "exporting deflation." Rather, it has been the persistent refusal of the American government to conduct economic policy with an eye toward preventing a loss of U.S. competitiveness, and a corresponding rise in huge external imbalances, that has caused the relative shift in economic fortunes in regard to America and Asia. The source of potential U.S. deflation, and today's quandary for American monetary and financial officials, is very much home-grown.
Bill wrote
Payne. W. H., Graduate Education: The Ph.D. Glut, Communications of the ACM, 16 (March, 1973).
when he was a member of the graduate studies committee at Washington State University.
Bill looked at Douglas L Jones graduate student lists.
Jones is only one of hundreds or thousands of college profs turning out MS and PhD degrees in computing.
There is a reckoning for this overproduction coming!
Graduate educating mainland China may have a downside. But we sure get great deals at Harbor Frieght!
Hyde shows lots of hardware test equipment.
Some of us can't afford all of that expense. And don't need to!
The stock market was in double-digit rally mode for more than a third of the calendar days in question. Any one of those rally days could have been called "the low." Some probably were. The overall period was the Dow Industrials from April 1930 to July 1932, and it devastated investors. The Dow lost 86% of its value, rally days notwithstanding.
These rallies were the moments when the bear was catching his breath.
"The Great Crash: 1929" by John Kenneth Galbraith is simply fantastic.
Csd is taking a fast tour through the Hyde book and accompanying cd.
Hyde documentation is in html with web links.
Csd usually develops software on a machine off line and with lots of disks since we occasionally have some horrible windows crashes.
Dig this warning from a Hyde 8051 software example.
In 80C32 Forth assembler and high level Forth source code is assembled and compiled on the 8051 so binaries are usually not distributed!
Only source code is usually distributed. But you can metacompile in forth assembler which would be in binary form.
Loading binaries on any machine is very dangerous.
So let's try to get Hyde thinking about what to do. john.hyde@intel.com
The Hyde book, other than using backward [Keil] 8051 software tools, looks to be real valuable!
We're tracking a technology shift bought on by gighertz PCs and USB 2.0. And the accompanying economic havoc!
Here's a very interesting comparison of RISC and CISC processors which shows how DSP functions have been implemented in a PC with MMX 1, SSE, SSE2, 3DNow extensions.
The computer industry is leading in deflation.
Washington, Dec. 12 (Bloomberg) -- Federal Reserve policy makers raised a warning about deflation at their Nov. 6 meeting and said the economy was at risk of further weakness rather than strength, minutes of the gathering show. ....
Deflation is a general fall in the prices of goods and services in an economy. Profits collapse because the prices of finished goods can drop faster than the costs to make them; debt levels become more burdensome as interest rates rise in inflation- adjusted terms. The U.S. economy isn't experiencing deflation currently. It is passing through a period of disinflation, which means that inflation is accelerating at a slower rate than previously. ...
Before the November meeting, Fed officials learned that the economy lost 5,000 jobs in October and that the unemployment rate rose to 5.7 percent, following a decline of 13,000 jobs in September. Both numbers were subsequently revised by the Labor Department, which now says the economy gained 2,000 jobs over the two months. The economy lost 40,000 jobs in November, and the unemployment rate now matches April's eight-year high of 6 percent.
Initial scanning of both the Hyde and Axelson second editions indicate that both books should be very helpful.
The Hyde book comes with a cd. The Axelson book does not. This is no doubt attributed to Internet. Why include cds when you can download software?
Csd recently bough a copy of Nero cd rom burning software. Version 5.5 of the cd did not work. But when the v. 5.5.9.17 was downloaded over Internet, it worked more or less okay.
Shift left click would not select multiple files to drag and drop.
Nero express, supplied with a cd rom burner, appears to do about everything Nero does but is simpler.
In addition, for those who are exhibiting a high degree of impatience waiting for the market to decline, lets remember that at the October low, the S&P 500 was down 51% from its March 2000 peak, making this the steepest bear market since 1929 to 1932, and one of the steepest in U.S. history. In the past 33 months the S&P 500 has spent only about four weeks above its 200-day moving average, and is still below that today. Moreover the 200-day average is still declining as well.
Other major technical indicators are negative as well. The percentage of bears in the Investors Intelligence Survey has dropped to 24%, a level more associated with market peaks than bottoms. At every other major market trough the percentage of bears has amounted to 60% or higher. Strangely enough, new bull markets begin in an atmosphere of deep pessimism, not the excessive bullishness characteristic of bear rally tops. In addition, equity mutual fund cash is at only 5.1% of assets, compared to 10-to-12% at previous bottoms.
Patty and bill had dinner with John Gilmore and his girl friend in Albuquerque several years ago.
Gilmore was one of the founders of Sun computers.
Gilmore told patty and bill the story of how Forth 1 got started at Sun.
Owner of Firmworks , mitch bradly, orignally wrote the Sun boot software in Forth.
But as we know, Forthians do very poorly in the business world.
Most old Forthians are now Javans.
Gigahertz PCs coupled with USB 2.0 and super fast 80C32 SOCs may cause a shakeout in the peripheral microcontroller industry in 2003.
So the Javans [old forthians] lose again.
But this is why we want to use BASIC-52 and Forth technology for 80C32 SOCs and USB produccts!
The BASIC-52 and Forth technology is inexpensive and is more reliable than other system software technologies!
Besides most of the work has already been done and is in the open literature!
Thursday December 12, 2002 15:46
Hunker down in 2003! And prepare for 2004 with some neat USB 80C32 SOC products.
T-Online International, Europe's biggest Internet service provider, expects 2003 to be "a lot harder than 2002," its chief executive said in remarks published Wednesday, sparking a sell-off in the company's shares just nine days after the stock went public.
"I do not see light at the end of the tunnel before mid-2004," Thomas Holtrop was quoted as saying in an interview with the German edition of the Financial Times daily
But is this new vision alone going to save AOL and allow it to remain the powerhouse of the online world? Probably not. To the degree that its existing customers switch from narrowband to the new broadband option, AOL's revenues will decline-by about $9 per customer each month. AOL will have to grow its membership base aggressively and also find ways to sell more services to its existing members just to keep revenues from shrinking. ...
I suspect not. Consumers will gravitate to online services that give them superior ways of working, socializing, manipulating media, and organizing their lives online. I don't hear nearly enough talk about that from AOL. Those are software challenges, and they take money and commitment to develop. AOL TW spends a negligible amount on R&D, especially in software, in contrast to competitor Microsoft which spends $5 billion a year. AOL TW executives say there's just no money for it. That's a tragedy, because there are many ways that the service could be pioneering new functionality for its members.
Bill once counted 11 Chinese restaurants in the Pullman, WA - Moscow, ID area. All are very good.
Graduate schools are compelled to give graduate degrees - even if they don't have qualified graduate students. Or faculty for that matter.
But this doesn't stop any of them from giving, or perhaps counterfeiting, graduate degrees.
No graduate students, no money from the state.
Graduate schools were running out of students in science and engineering.
But then they discovered a source of lots of students. China.
Foreign graduate students comprised more than 94% of computer science and electrical engineering at Oregon State University, Lewis told Payne more than 10 years ago.
The gap fell from a revised record $127.6 billion in the previous three months, the Commerce Department said. It was the first improvement in a year.
The U.S. needs to attract almost $1.4 billion a day in foreign investment to fund the shortfall and keep the value of the dollar stable. While foreign investors still have faith in the world's biggest economy, some say that may not be the case indefinitely.
``Gold has rallied due to dollar weakness. With continued risks of Iraqi and terrorist tension, we believe that a short position in gold is too risky for most players and this prevents the gold market from coming under too much pressure,'' said John Reade, analyst at UBS Warburg.
The dollar fell to its lowest for a month against the euro, sliding in thin trade ahead of U.S. retail sales and other data expected to paint an anemic picture of U.S. economic recovery.
Traders said the U.S. currency was also under pressure after a U.S. newspaper said Washington had a credible report that Islamic groups linked to al Qaeda received a chemical weapon in Iraq last month or late October. A weaker dollar makes bullion more attractive to holders of non-dollar currencies.
Economists had expected a third-quarter deficit of $132 billion after a previously reported shortfall of $130 billion from April to June.
Over the past 12 months, the deficit equaled 4.5 percent of gross domestic product, up from 3.9 percent in 2001.
An imbalance between the goods and services the U.S. imports and those it exports accounts for about 85 percent of the gap in the current account. The nation imported $110.9 billion more than it exported in the third quarter compared with a $109.3 billion trade gap the prior three months.
The trade portion of the deficit is likely to worsen in coming months as imports continue to grow faster than exports and U.S. economic growth outpaces growth abroad.
And look at what some of these likely US-educated students are doing now!
Csd spends lots of money at Harbor Frieght! And gets great deals!
Digital calipers used to cost on the order of $79!
Csd purchased an exquisite auto-range digital multimeter which measures, the addition to voltage, resistance, amps, diode test, hFE, capacitance and frequency!
In a yellow rubberized case!
For about $28!
Csd waits for sales!
It works great!
Economists, like
lawyers, are BS artists.
Dig this paragraph as
another example of muddled thinking.
Members commented that the potential costs of a policy easing action that later proved not to have been needed were quite limited in that there was little risk that such a move would foster inflationary pressures under likely economic conditions over the next several quarters. Moreover, the policy easing could readily be unwound without significant effects on financial markets if the reversal appeared to be warranted by growing pressures on resources in a strengthening economy. In contrast, a failure to take an action that was needed because of a faltering economic performance would increase the odds of a cumulatively weakening economy and possibly even attendant deflation. An effort to offset such a development, should it appear to be materializing, would present difficult policy implementation problems.
Today Hyde's second edition arrive from Amazon.
Two days ago Axelson's second edition arrived from Softpro.
Here's a just taken jpg. There is the Cypress EZ-USB-FX both Princeton and Harvard architecture 80C32 development board on top of the email from Russia and CAVE algorithm description.
Csd keep its driver development disks outside the pc for easy access.
Now that the legal project is well in control, we're switching focus to USB and THE FUTURE.
We have to keep 80C32 USB system software costs down in a possibly deflating economy.
Greenberg's outstanding article on the Lucent pension funds has been spiffied up.
Bill has never been interested in administrative jobs. You quickly become technically obsolete in those jobs.
And when the pension checks possibly stop coming you are deep kim chee!
And we are project oriented. Both technical and legal.
Debt, Deflation, Default, Deficits, Demographics...
For example, the "deciduous dollar declined" yesterday against the yen in anticipation of a report, due out today, which is expected to show that the U.S. current account deficit widened to a whopping $132 billion in the third quarter.
At that rate, the U.S. must lure in $1.3 billion a day in unsuspecting foreign capital just to keep the consumer binge alive. The dollar has dropped 1.3% since Treasury man O'Neill, the administration's formerly vocal proponent of a strong dollar policy, was shown the door on Friday. Is there more where that came from? Probably, we suspect.
The surplus countries Kohn must have in mind are countries like Japan and China," says Denning, attempting to unpack the Fed governor's comments. "China alone accounts for 21% of the U.S. trade deficit, even more than Japan. And that's after China raised its U.S. dollar reserves $62.4 billion in the first 10 months of the year. Japanese dollar reserves are up by over $60 billion in the same time. What does this mean?"
Second edition of Axelson's USB book arrived yesterday. So maybe csd can get back on to tech stuff now we have met a deadline in our legal project.
From Silicon Valley to Boston's Route 128, technology companies are waiting for a rebound in business spending to pull them out of their slump. But when the Bay Area and other regional hotbeds emerge from the deep hole left by the technology crash, they'll find that the landscape looks a lot different. Many of the jobs lost will not be back.
The global tech bust has forced a dramatic restructuring of the industry as firms struggle to return to profitability. Companies have shed thousands of jobs in the United States and Europe while shifting production to lower-cost countries, mostly in Asia.
AnswerWe share your concerns and believe that you are correct in highlighting these problems. Unfortunately, however, mainstream economists are ignoring the credit bubble, and think that the banking system is in good shape. In fact, their solution to the economic malaise is for consumers and corporations to go into more debt so that they can spend the economy out of the doldrums. We dont see any of the items you mentioned being discussed as a problem within the Administration or Congress. As you probably know from his speeches and testimony, Greenspan has repeatedly praised derivatives as having reduced risk throughout the economy. As is true throughout financial history, the problems wont be addressed until they reach crisis proportions. We might add that the problems now being addressedEnron, WorldCom, corporate governance, Wall Street research, investment banking, and the anemic economyonly became key issues after they began making headlines in the media. The attitude of the government and the media is that if a problem has not become apparent to all, then its just not a problem.
The ever-growing federal deficit will weigh on interest rates even more than investors realize. That's because the deficit is already even BIGGER than what the government reports. The government is projecting a deficit of more than $200 billion in fiscal 2003. But if you strip out all of the smoke and mirrors and look at what the government actually owes, the government is running a deficit at the annual rate of more than $940 billion -- and that's according to the Federal Reserve's own numbers. If the U.S. needs to pay for a war or implements tax cuts to try to stimulate the economy, the deficit will grow even more!
Result: The government will likely flood the bond market with more long-term bonds -- and will need to increase interest rates in order to bring in investors. That's bad news if you're hanging onto long-term bonds right now. When reams of long-term bonds are dumped on the market in the months ahead, prices will plunge.
Csd is anxious to spend more time on the USB project and less time on the legal project. But lots of thinking on the technical project has been done while doing legal work.
Wednesday December 11, 2002 19:54
What is going to happen when senior citizens don't receive their monthly pension checks in the mail?
And their medical benefits stop.
They may go to a lawyer.
But we have bad news.
The lawyers will steal their remaining money.
There is a reason for this website and what what Morales and Payne are trying to do at pro se fights.
VIEWPOINT editorial Barry Greenberger, Managing Editor Beneath the dry industry statistics are people who have been seriously hurt EBN December 9, 2002 |
Wednesday December 11, 2002 10:44
Grove also addressed the diminished likelihood of an upturn in the chip industry in the near future. "Over the course of the past year (the industry) has been bounding along on the bottom," he said, but he warned that the threat of a "war" on Iraq doesnt bode well for the future employment rate in the US and a may spark a consequent "meltdown" in some South American economies.
Grove also later warned that the trend of migrating chip manufacturing to far eastern fabs could shift the balance eastwards. "It is easy to project," he said, "that the interdependence becomes more one-sided, with an adverse impact on our educational system because so much of the university funding comes from industry. There is a spiral there in the wrong direction."
The CPU upgrade rat race is over. The commodities war is on
WHILE INTEL AND a select number of other companies bet tens of billions of dollars on the next generation of chip fabrication technology, a good chunk of the industrialized world is happy with the machines they bought last year or two years ago. Today, accountants to video heads have enough horsepower at their fingertips to keep themselves sated. Why shell out $1500 in next year's technology for a measly 3-5% performance bump on the most bleeding-edge applications? It is buy and keep for consumers and businesses alike, not buy and dump every other year in the climb from the 8088 to the Pentium.
Is Carpenter trying to tell us that 53% of companies selling stock have a P/E = 0? Or maybe even negative? Like Palm!
Where's The Mean?
by Lynn CarpenterIs the market too high? Wall Street says it's cheap. But many of the folks on Main Street think it's still too expensive. Which street is right? The short answer is, "both."If the market really is too high, then we have no business buying stocks at all.
The average P/E ratio for the Dow Industrials is 20 these days, not too bad. But that's a pretty stodgy group of stocks. For the S&P 500, the average P/E is 33... and to us value investors, 33 is generally too high. By the way, the S&P Industrial Index sports a P/E of 48, pretty stiff for a box of 15 industrial stocks chosen for their high dividends. Worse, the yield on the S&P Industrials is now just 1.5% compared to the Dow Industrials at 2.2%! ...
In fact, even if we stretch the average a bit, allowing for a range of 1416, the P/E ratio has been at its historic average for only moments in time. It happened more frequently from 1880 to 1912. The S&P P/E ratio hit the average nine times in those 32 years. One of those times it stayed there for nearly two years. Every other time it landed briefly, for six months or less.
The next time the P/E ratio was "average" on a regular basis was... oh boy, you don't want to hear this... 1928 to 1941 roughly bracketing the Great Crash, Great Depression and Second Great War. If that's average, I am sincerely hoping we never see average again. ...
The U.S. stock market has 9,175 stocks including those with very small caps and infrequent trading. So let's talk about those 9,000 stocks. Of those stocks with earnings, only 16% (1,444) are currently selling for P/E ratios over 20. However, an awesome 53% of the available stocks have no P/E ratio at all. (Yikes! That's not good.) They might as well have a P/E of 300, which some of them would if they had a quarter-penny's earnings.
LYNN CARPENTER EXPOSES WALL STREET 'S FRIGHTENING NEW "DARK SIDE!"
"Back When The Stock Market Went Mad, I Warned You,' STAY AWAY' From Enron..."I Begged You, 'DON'T TOUCH' Tyco or Global Crossing...
"Some of my readers even took my advice and bought puts on Enron early... on THREE separate occasions... and one investor wrote me recently to say my warning helped him pocket over 200%!
"But please don't think the worst of the Wall Street crisis is isolated to a few companies or corrupt financial firms. FAR FROM IT...
"These recent wipeouts are nothing compared to the TIDAL WAVE OF BANKRUPTCIES ahead. Disney, Ford and General Electric have already sent up emergency flares. And they're not alone. Not at all...
"If you do anything in the next 5 minutes, make sure you check out my shocking 'WALL STREET BLACKLIST' where I name 26 stocks you shouldn't touch with a 10-foot pole! You'll find the list inside, please read on..."
We have to hold down software costs while producing reliable easy to maintain apps. And have fun too, of course.
How close is the US to deflation? Fed chairman Alan Greenspan told anxious members of Congress in October that "we are not close to the deflationary cliff". But how good are his forecasting skills? The Fed, by its own admission, utterly failed to foresee Japan's slide into deflation seven years ago.
And a solid core of serious analysts now believes it to be a very real risk. The Fed's own behaviour betrays a distinct unease. A month ago it cut the key official interest rate by an aggressively large increment - from 1.75 per cent, already the lowest in over 40 years, to 1.25 per cent.
Consider the evidence of deflation so far.
The broadest measure of price pressures in the US economy - the gross domestic product deflator - is barely above zero. In the year to the end of September, it was up by 0.83 per cent, its feeblest in half a century.
"The essence of deflation is that business leaders know they do not have pricing power," the ability to raise the price of their products, points out Wayne Angell, a former governor of the Fed and chief economist for the Wall Street investment bank Bear Stearns.
And the prices received by US firms overall have declined in each of the past five quarters, the longest run in more than 50 years. For the latest quarter, they were down 1.3 per cent.
"They expect that pricing power will return with recovery." With deflation, it doesn't. "This is precisely what is happening in the US economy today.""They expect that pricing power will return with recovery." With deflation, it doesn't. "This is precisely what is happening in the US economy today."
Software is a breeze on the PC side with Visual Basic, C++ 6.0, exquisite in-line assembler, and Microsoft Office.
Software development is easy too on the 80C32 side with BASIC-52 and Forth ... with a bit of interactive incremental assembler too!
Both software and hardware maintenance is easy on the 80C32 peripheral side using Internet. The Brazilians in Santa Caterina are on to this!
Let's look at bit more carefully where the Brazilians are that are doing the really neat BASIC-52 and Forth work. Here's a map of Santa Caterina. Look at Florianopolis on the island and Blumenau.
Then let's look at the German influence.
Germans migrated, or escaped, to south America at the end of the second world war.
Picture of the kubelwagen might bring tears to the eyes of some of the FURB and Federal University of Santa Catarina students' grandfathers.
The practical constraints are that we don't want to spend lots of money of system software.
Consumers are maxed out. Personal bankruptcies surged 15% from 2001 to 2002, credit card write-offs recently rose by more than 35% to their highest level in decades, and mortgage foreclosures have hit their highest level in history. On top of that, the unemployment rate just jumped to 6% in November, and 40,000 net jobs disappeared from the economy. Millions of consumers are buried under mountains of debt and are worried about the prospect of long-term unemployment if this economy doesn't turn around quickly.
And you're thinking about sending several K$ to Keil?
Think 80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth, instead!
The high growth experience of the 1990s has unfortunately trained people to expect to be able to achieve high levels of growth under any circumstances. This is a fallacy. The three years of the new century have seen but three years of declines. The speculative Nasdaq has fallen over 75% from its ultimate Y2K peak to its low so far. The broader S&P 500 fell 7% in 2000, 15% in 2001 and approximately 22% year to date in 2002. ...
Nystrom states
What many people thought of as "money" in their brokerage accounts has simply evaporated with those declines.
Some people got the money. The money did not evaporate. Much of the pension fund money ended up in silicon valley!
And look what the tech creeps are doing now!
According to Thomson Financial, insider selling skyrocketed 125 percent in November to $2,6 billion from $1.2 billion in October. Tech insiders were especially busy as the level of selling increased to $861 million from $244 million, a whopping 250 percent rise. And insider buying? That was up just 5 percent, to $193 million from $184 million. ...
Insiders at semiconductor company Maxim Integrated Products also sold a sizable amount of shares last month. The stock is up more than 65 percent in the past two months. Nine Maxim insiders, including CEO John Gifford, sold more than 440,000 shares for about $14.8 million. This was the largest group sale by Maxim insiders since March, according to Thomson.
Investors are wising-up to high tech.
The year's biggest technology initial public offering got off to a bumpy start on Tuesday, raising nearly $230m less than its more optimistic backers had been expecting.
Shares in Seagate, the world's biggest maker of computer hard disc drives, were priced late on Tuesday at $12, below a range of $13-$15 that some of the deal's promoters had thought might even have been exceeded.
The transaction raised $870m, rather than the $1.1bn that would have come from the more optimistic pricing - still almost as much as all the other tech IPOs this year combined. ...
"It is hard for buyers to feel good about a deal where the sellers are going to make a ton of money," said Ben Holmes, president of Morningnotes.com. "There is not a lot of upside for buyers. That's why there is no expectation that this deal will trade up."
Tuesday
December 10, 2002 18:30
Try this video link if you want some deadly high tech stuff
We are fed information by organized media to try to make us do things. Like buy or sell stock. Or change our opinion. Or tell us about new microcontroller hardware without the complete hardware software picture.
Media suppresses things the paying establishment doesn't want us to know about.
But with internet things have changed.
Organized media and those who pay them don't have as much control over what we are told.
Also the establishment has no advantage in the technology over others. And it costs the establishment lots of money to use this new technology!
Without personal computers, lots of disk space, scanners, ink jet printers and, of course, internet Morales and Payne would be in real trouble. But this time the establishment may be in some real trouble.
We need to switch to USB.
Bill hasn't done USB before. But we all must arise to the occassion.
Bill hasn't done this
before either!
Let's hope!
After an 8 percent decline for the Nasdaq in the past five trading days -- and a near four-percent rout on Monday -- here's some more bad news for tech: Earnings estimates may be way too high. ...
The S&P 500 is now trading at about 17.3 times 2003 earnings estimates. That's not overly expensive by historical measures. But considering that the index was trading at 15.6 times 2003 earnings estimates on Oct. 1, when the earnings outlook was brighter, it's a cause for concern. Even if fourth-quarter earnings for this year are better than expected, stocks might not head higher on the news.
"There is a huge disconnect between the magnitude of what is likely to happen with earnings and what stocks have already discounted," says Richard Williams, strategist for Summit Analytic Partners, an independent research firm.
Valuations for technology companies in particular have climbed to lofty heights. Cisco Systems, for example, has a P/E of 25 times calendar 2003 earnings estimates. And multiples for smaller technology companies are even higher. Cisco competitors Extreme Networks and Foundry Networks trade at 34 times and 37 times 2003 earnings estimates, respectively.
This charts shows how much money 50 insiders reaped since from selling stock in companies that ultimately saw their stock values plunge.
The main page got too big again.
So Belec and other fraud articles are archived at Hi tech fraud.
Beginning in 1942, however, U.S. stocks finally shook off the lingering effects of the Great Depression and began the longest secular bull market in history. The advance started with four consecutive up years, pushing the S&P index higher by 100% over that period. In fact, the S&P closed lower over a calendar year only five times between 1942 and 1961. ...
While history may suggest that stocks are due for a meaningful surge, investors shouldn't expect anything close to the bull run that erupted after the attack on Pearl Harbor 61 years ago today.
Back then, equity valuations had fallen to extremely attractive levels amid deflation, unemployment, economic stagnation and the complete absence of Main Street participation in the stock market.
Today, valuations are reasonable at best, and borderline ludicrous at worst. The difference in interpretation stems from the fact that creative accounting practices have obfuscated the true worth of many U.S. businesses, even as the economic outlook remains unusually murky. ...
Since 1896, the S&P 500 has fallen for four consecutive years on only one occasion. That period was between 1929 and 1932, when the index dropped by 70% on a calendar-year basis, but by almost 90% when measured from top to bottom. ...
Of course, the similarities to the current environment are at least a little scary.
Most of the microcontroller hardware or IP vendors have obscured system software costs. Perhaps this is part of another scam? To steal pension fund money.
And we know there is a technology shift under way. Gigahertz PCs running Microsoft Windows, USB 2.0, and super fast 80C32 SOCs which may simply clobber peripheral side microcontrollers in the marketplace.
Consider clockspeed, a measure of how fast computer chips process data. While many PC users now work on computers with clockspeeds of around 400 to 600 megahertz, Kay expects hardware makers to introduce chips that are 10 times as fast within the next year or so, at about 4 gigahertz. Intel recently rolled out a 3-GHz processor.
Three new Athlons--the 2600+, 2700+ and 2800+--were announced in November and run at speeds between 2.13GHz and 2.25GHz, according to PC makers. The model rating system compares the new Athlons' relative performance to a previous version of the chip. But it also serves as a rough comparison to Intel's Pentium 4. The Athlon XP 2800+, AMD contends, performs the same or possibly better than a 2.8GHz Pentium 4.
ARM, ARC, MIPS, Tensilica, PMC-Sierra, UBICOM, ... and lots of other microcontrollers even manufactured by big semiconductor companies such as TI and Motorola may flop.
Fred Hickey, who writes the monthly High-Tech Strategist, has been dead right on the group for the past three years, and titles his latest report, Another Return to Stupidville. He points out that any sub-sector strength we see sporadically is based on inventory filling, and that virtually all the end markets are dead in the water. Most companies are reporting disappointing results and are continuing to slash capital expenditures and layoff their employees. Demand for PCs, wireless phones and telecom equipment is still very weak and this is feeding back into the semiconductor manufacturers and the equipment companies.
TI, of course, sees this and looks to be giving the big push toward 80C32 USB SOCs!
Csd's business interests are in wdm drivers and 80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth, of course! And having fun too!
Monday December 9, 2002 20:50
Consider all expenses. Especially 80C32 operating system expenses, of course!
Carrie Wolford never sat down and did the math to see how the cost of her commute from Paulding County to Buckhead compared with her mortgage.
"I didn't even think about," said Wolford, 34. "We just didn't care. We wanted to be able to get a nice house, and we just couldn't do it intown."
Wolford and her husband, David, moved a long way from Atlanta for a bigger, less-expensive home.
They found --- and love --- their four-bedroom, two-story traditional at the Menlow Station subdivision near Hiram. They carpool together and drive nearly 30 miles each way to work.
Csd bought 3 books last week as an investment for the future.
Csd doesn't plan to buy a Keil 8051 development system, however!
Which 80C32 hardware looks good? Csd has a Cypress board EZ USB FX.
But this is not SOC. In fact, there are about 11 chips on the development board.
Csd expects too see in 2003 a single chip 80C32 USB 1 or 2 on a board with only external capacitors, on USB connector and possibly one more connector!
This will dramatically change board manufacturing. And shift emphasis toward minimal peripheral software. The majority of computation will be done inside a PC.
Will Cypress be a winner?
What about Warren Buffett, probably the most lauded investor still trading today? He doesn't look at PE, or O'Neil's PEG, or price-to-sales or yield, other measures you'd expect a traditional value investor to use. Instead, Buffett really scours a company's debts, and buys shares only in firms that can pay off their long-term debt with two years of net earnings... last year's earnings. His ratio? Long-term debt must be less than twice last year's earnings.
Cypress's stock is down. And the Cypress debt to earning ratio is marginal according to Buffett.
Dig the Cypress interest payments.
TI is poised to come on very strong in the 80C32 USB 2.0 SOC marketplace. And with TI's acquisition of Burr-Brown, also using 80C32 SOCs with it mixed signal products, TI will be in a position to furnish a several chip solution connected to a PC over USB 2.0.
Csd believes that a USB 80C32 will nearly alway communicate with a non USB 80C32 mixed signal chip. Serial port communications with a PC, used in Cygnal and Cypress, was not a good idea. The 80C32 serial ports should be used for a serial expansion bus or 8051 networking communications.
Besides, a USB 80C32 can be viewed as merely a replacement for a rs-232 line driver chip!
Csd has made a revokable decision on how to build the windows wdm rs-232 driver to replaced the DOS Forth driver! The easiest way of course.
Java makes sense on high powered peripheral microcontrollers. And so does a Forth operating system. But these may not sell.
USB 2.0, GIGHERTZ PCs, Microsoft Visual Basic and C++ 6.0 and 80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth, of course, may do the above in.
But in the long run, his contribution may be greater. Browsers and servers may come and go, but successful new languages are extremely rare. Java, or something like it, is the key to a truly interactive Internet and a fully hollowed-out computer.
But we could be wrong too! Here's what csd is doing to prepare for 2003.
Csd is doing successful old languages and operating systems. BASIC and it's weird associate, Forth. Most of the work has already been done and we are building on the ideas of other software superstars.
Weird is correct for Forth. But very powerful running on a 80C32 and a real money-maker. For the app seller. Not for the system programmer. Damn.
Csd was given the below Oney flier this morning.
Csd ordered the second edition of Hyde's book from Amazon. It apparently is least expensive source from those given on Hyde's web site.
Csd has Oney's driver book and accompanying cd disk.
Oney's software coupled with Cypress's EZ-USB-FX software caused one of the worst Windows 2000 crashes csd has ever experienced!
The Oney approach is one way write wdm drivers.
Numega Driverworks is another.
Competition is good!
Numega has a very interesting USB tutorial.
Csd wants to see what Hyde wrote and how he handles software on the peripheral side.
We, of course, want to go the 80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth.
But there is a general agreement that Windows and USB is where one wants to be!
The Java and C++ people seem to have a hatred of each other.
Java is a variation on Forth. Sun computers boot in Forth.
Microsoft Visual Basics prior to 6.0 compiled into Forth-like threaded code. Now, of course, 6.0 compiles into C++.
Csd is not going the Java route on PCs. There are lots more advantages to Visual Basic and C++ 6.0 on a PC.
Java is slow!
True compilers like Visual Basic and C++ 6.0 produce code which runs much faster than a Java program.
One reason is both Forth and Java are slow is all of the pushing and popping of arguments from the data stack. But 90% of the work in all programs is done in about 10% of the code. So you recode the 10% in assembler.
For an 80C32 real-time operating system Forth and BASIC-52 are very attractive choices. For sound financial reasons.
Controlling software costs with 80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth is a csd business goal.
Above is serial port driver for Windows 98se. A vxd, not a wdm, driver.
Csd was hoping to have a common solution for 98se and 2000.
Sunday December
1, 2002 09:44
Csd saw
an error message in the dump regarding the path. All worked okay though since
system path included a path to link.
Depending on the path setting in Windows 2000 you might have to add an explicit path to link.
Here's a slightly modified dump.bat with an explicit path
To find Windows 2000 path, open a DOS window and type path.
This sure beats typing a long dumpbin or link line interactively ... and usually making a mistake.
80C32 USB SOCs hopefully running BASIC-52++ and 80C32 Forth, of course! Connected to gigaHertz PC with one or more gigabytes of main memory. Gobs of disk space too.
Running Windows 2000 or 98SE, Visual Basic 6.0, Visual C++ 6.0, and Microsoft Office.
Saturday
November 30, 2002 18:35
About the disassembler in link -dump.
When csd first tried this option it didn't work. A missing module.
Our son installed the program Deer Hunter on our Internet machine.
Then the disassembler worked!
The real world again.
About the money.
If we get our wages paid for doing this work plus some consulting in the future, then is this all we can hope for.
We are dealing with the ideas and work of others. Who will hopefully use our work in their consulting businesses to make money too.
Numega DriverWorks and VtoolsD, which is obsolete except for some outstanding examples and should not be lost, like 8051 Forth and BASIC-52 is all source code.
Megabytes in the case of DriverWorks.
Several hundred screens in the case of 8051 Forth. And lot of assember listings in the case of BASIC-52.
So let's try to get our wages to do this systems work paid for then move on to some profitable app work!
And have fun too!
If anyone is stupid enough to think they are going to take this code and make millions of dollars, then think again.
For such a stupid person would have violated Mushashi's first rule.
Do not think dishonestly
We are dealing with the works of others which we hope to add to.
Csd must post some info on the classical Japanese company.
One which all of us would like to work for.
But may have been lost in the tech rush to economic disaster.
Stay tuned.
Guys, we're getting serious about the best way to replace a DOS Forth interface through the serial port to a windows serial port interface.
There are lot of way to do this. But we're looking to for best.
8051 Forth multiplexes Terminal Input Buffer and disk buffer data going from a PC to the 8051 using one of the control lines as a flag.
Screen display data and returning disk buffer data are also multiplexed using one of the control lines as a flag.
Can serial.sys be easily used for the conversion from a DOS window to a windows window?
We don't know yet.
But if not, the Microsoft serial.sys serial port driver profile can be disabled.
Numega DriverWorks can be used to generate the skeleton for a custom serial port driver.
The serial port will only be used to get usb working. At that time serial port communications will be discarded.
All you seen the panel below is being done on Windows 2000.
We keep our software tool costs in line with what we can expect to make from them.
Seniors, the pension funds have been ripped off.
High tech is largely to blame.
Shakeout in micrcontollers, and hopefully obsolete 80C32 developement tools too, is likely.
BASIC-52++ and 80C32 Forth will win!
Some of the articles will be archived. Some will not.
Csd is thinkinig about the strategy.
Java is a variation on Forth. And may eventually go just about as far a Forth went.
BASIC, on the other hand, is compatible with what people already know. This makes BASICs a lot easier to sell.
We must keep in mind that many BASICs work similar to Forth.
Microsoft's BASICs compiled into Forth-like code until the advent of Visual Basic 6.0 which complies into C and C++!
Thursday
November 28, 2002 10:46
Keil has of course attracted attention for wanting to sell us a very expensive and antiquated non-operating system software for 80C32 hardware/software development.
There are better solutions. BASIC-52++ and 80C32 Forth.
Wednesday
November 27, 2002 19:07
Csd is generating both ISA bus for RS-232 and USB skeletons. And we
see what we're doing wrong!
We think. But we, of course, live in the real world where things work a bit different.
Windows continues to get more picky making you do it Microsoft's way!
Neither of the DriverWorks 2.0 skeletons run with the console app test programs. But it's fun to try to find the problems. Eventually things work after study and experimentation.
Software is expensive and, of course, the most insidious part is that only about 20-25% of the eventual cost measured over the software's useful lifetime has been invested once one thinks the software is working!
But Internet software maintenance should help improve this.
Monitor Masters in Albuquerque replies
Hi Bill, Don't worry, I believe I know what you are seeing. It's not really a failure and won't get any worse. These monitors have a "moiré" adjustment. I think this is what you are seeing. I don't have another monitor like yours to make sure these instructions are precise. But the "second function" of some of your normal adjustments is for "Moire". You should be able to see some little symbols around the adjustments buttons, trying to indicate that if you push and hold one button (I believe it is the brightness button, but without a monitor I can't be sure) and one of the other buttons you'll be given "Vertical and Horizontal Moire" adjustments. I think the Vertical shows up more. So when the monitor is cold, try playing with these adjustments. Dan says if this is off some it could appear more when the monitor is cold. We believe this is what you are seeing. If this doesn't help, then I guess you will have to bring it back in sometime next week and we will try to get it to "misbehave" for us. (We're of course closing for the Holiday weekend. :) Hope this works, |
Csd just retired a slot 1 Pentium 333 MHz.
And an Acer ScanPrisa scanner - which did a great job, after one warranty replacement.
Two Acer cd/rw drives, which also were replaced under warranty, and one cd rom reader.
They are in a plastic bag in the garage.
Throw these in the garbage?
No.
Environmental pollution is not a good idea. We have to think toward the future.
That's why they are in plastic bag in the garage.
But what about the ever-increasing dead batteries for the digital camera?
How do we properly dispose of these?
We're increasingly concerned about this.
Hey, getting rid of obsolete software is easy and non-polluting.
Uninstall and the recycle bin.
We have problems which must be solved. In a non-environmental polluting way.
Fewer micrcontroller types [the shake-out], most of may end up at the dump polluting.
Old 80C32s should be properly recycled into newer and faster parts.
Better PC side software. Visual Basic 6.0, C/C++, in-line assembler.
Less power consumption. The 80C32 power saving modeS, of course.
Fewer batteries. Solar recharging.
Csd and others have the ticket.
Power
efficient 80C32 SOCs, running BASIC-52++ and 80C32 Forth, of
course!
System software should never assume vast proportions. Especially if it was built on the ideas of others. Like 80C32 Forth or BASIC-52.
We can't afford this and users can't make money with this concept. We have to be economically efficient bulding our apps.
All users should have system software source code available which is complied by a metacompiler. We need this to understand what's happening on the system end.
Business travel should be mostly unnecessary. We can do this work over internet.
An absolutely brilliant solution means building on the ideas of the super bright. And spending the time to figure-out what they did.
Do not think dishonestly
The Way is in training
Become acquainted with every art
Know the ways of all professions
Distinguish between gain and loss in worldly matters
Develop intuitive judgment and understanding for everythingMiyamoto Musashi
So onward to get the PC side of Embedded Controller Forth for the 8051 family fixed.
And have some fun too learning what the Microsoft VB 6.0, C/C++, and Numega DriverWorks 2.0, not 2.6 unfortunately, programmers have given us.
Monday November
25, 2002 08:43
Csd is in the process of replacing
the DOS Window Forth PC side of Embedded Controller Forth for the 8051 family
software with a Microsoft Windows MDI Visual Basic 6.0, C/C++, in-line
assembler wdm driver.
Productivity increase using these tools, compared to using Forth nearly 20 years ago, is on order of magnitudes!
Forth usually not an appropriate tool to be used on a PC.
However, you can still do IN and OUT in a DOS window on Windows 2000!
Doing IN and OUT from Windows 2000 ring 3 app, causes the app to be immediately terminated ... with no error message.
Sunday November
24, 2002 20:09
Google has likely made more
contribution on acquisition of knowledge than anything before in
history.
All of us should be concerned about debt load and how to avoid it.
Like using BASIC-52++ and 80C32 Forth instead of paying big bucks to Keil for an antiquated technology.
We need to get some intelligent SOC hardware vendors to pay for porting of existing BASIC-52++ and 80C32 Forth operating systems so that we can all move on to development of peripheral and PC side apps quickly, inexpensively, with super reliability.
Intel paid for BASIC-52. Absolutely brilliant decision on Intel's part.
Oh, hardware vendors are counting on entrepreneurial software vendors spending money on hopes of making big bucks supplying software for their hardware.
The entrepreneurial software vendors will lose. They have to invest too much money to get their products to market.
Reasons is that they don't use the ideas of others. Like BASIC-52++ and 80C32 Forth.
And this software and documentation must, of course, be posted on Internet!
Join the program. It's fun. And you meet some really neat and super-intelligent people through their written works.
Writing apps is lots more profitable than trying to sell system software!
We figured this out. But you may have to learn some new and weird, but fun, stuff.
And adopt a personal code of conduct you must believe in and follow.
Do not think dishonestly
The Way is in training
Become acquainted with every art
Know the ways of all professions
Distinguish between gain and loss in worldly matters
Develop intuitive judgment and understanding for everythingMiyamoto Musashi
Because we are using the ideas of others who gave them to us in source code.
We attempt to stand on the shoulders of giants. Hey, these are Newton's words.
That's what we were up to when we distributed all of our system software on BBSs and Internet.
Csd is tracking a possible microcontoller shake-out brought on by
Friday November 22, 2002 12:47
Technology has changed to favor gigahertz PCs, Microsoft Visual Basic 6.0, C/C++, USB 2.0 and the fast 80C32 USB microcontrollers.
Shake-out time.
Thursday November 21, 2002 10:46
Standard Microsystems Corporation has a series of three ads in the November 18, 2002 Electronic Engineering Times.
One is on its Fast 8051 USB 2.0 product. SMCS states
Provisions for external Flash Memory up to 64K bytes for program storage is provided.
12K bytes of scratchpad SRAM and 768 Bytes of program SRAM are also provided.
So it looks like some external memory may be required.
It appears from SMCS development literature that no interactive operating systems hosting incremental compilers and assembler are available for their "development" boards.
SMCS too could use BASIC-52++ and 80C32 Forth working on its part.
TI looks to give Cypress and Cygnal some real competition in the 80C32 SOC marketplace.
Let's hope TI sees the merits of getting BASIC-52++ and 80C32 Forth working on it's 80C32 parts.
From the ROM and RAM sizes mentioned below, this does not appear to be a problem.
There is some very interesting and fun work to be done here. And not too much of it.
Face it. No one is going to make a fortune off this software.
But some can make big bucks getting a super reliable app out the door quickly. Which can be easily maintained and enhanced using Internet!
Look at what the boys from Brazil did. And what Hanson and bill did using Netmeeting.
Lowest noise, precision data-acquisition
system-on-a-chip
MSC121O Achieves New Level of High Performance The MSC121O family allows the user to optimize the memory size to the application. The MSC121O is a new, low-power (4 mW) family of mixed-signal devices incorporating a 24-bit analog-to-digital converter (ADC), enhanced 8051 core with Flash memory and a variety of on-chip peripherals including a 32-bit accumulator and precision voltage reference. The M8C1210's microcontroller core is an enhanced 8051 core which executes up to three times faster than the standard 8051 core, given the same clock source. This makes it possible to run the device at a lower external clock frequency and achieve the same performance at lower power consumption than the standard 8051 core. Available with four different sizes of Flash memory, the MSC121O family allows the user to optimize the memory size to the appbcation. In addition, it provides an easy migration path to large memory sizes without having to add the external memory often required hy other devices. The Flash memory has an endurance of lOOK E/W cycles and is programmable over the entire operating range of the device: 2.7 V to 5.25 V. The MSC121O family offers superior performance throughout. Linearity error is less than ±0.0015 percent of FS. Effective resolution is 22 bits at low data rates for PGA = 1. For PGA = 128, noise is as low as 75 nV, which makes it the lowest-noise ADC integrated with a microcontroller on the market. Also, the on-chip VREF offers excellent accuracy of 0.2 percent and drift of5ppmP C. Digital features of the MSC121O family include 1.2 Khytes of on-chip SRAM, four I/O ports, external memory capability, a 32-bit accumalator! shifter, dual UARTs, three timers/counters, SPI interface with direct memory access and a watchdog timer. Versatile MSC1 210 Family Res. Sampling No. of Input Program SRAM Price Upgrade legacy serial peripherals to USB with a single chip
Avoid redesign by adding TI's TUSB341O to bridge an RS232 communication stream to USB. The TUSB34JO controller bridges between a USB port and an enhanced UART serial port using all the necessary logic to communicate with the host computer through the USB bus. The device is ideal for building a communication port between a legacy serial peripheral device and a PC with USB ports. All device functions including USB command decoding and UART setup and error reporting, are managed by internal microcontroller firmware under the auspices of the PC host. This USB-to-serial port controller contains an industry-standard 8052 MCU with 16 Kbytes of RAM. The default code in the chip's 10 Kbytes of ROM contalns an 12C boot loader with the option to load TI's UART bridging code or application-specdic code from an onboard EEPROM or directly from the PC over USB. USB-to-Serial Single-Channel Controller The TUSB341O can support a total of three input and three output interrupt, bulk) endpoints. Product Features Enhanced UART with Typical Applications Legacy RS232-to-USB converter cables USB 2.0, high-speed ATA/ATAPI bridge solution PREVIEW* The TUSB625O programmable USB 2 0-to-ATA/ATAPI bridge features a unique architecture that offers the designer superior throughput performance - 33-percent improvement in sustained throughput over competitive bridge devices. This higher rate of throughput can enhance the read/write performance of peripheral hard disk drives. Programming flexibility allows developers to use a wider variety of off-the- shelf ATA/ATAPI devices than would be possible using alternative bridge chips. The TUSB625O is designed to use both the high performance of a state machine and the programmability and flexibility of an embedded microcontroller and firmware. With its embedded 30-MIPS microcontroller, eight configurable endpoints, and up to 40 Kbytes of configurable code and data buffer SRAM, the TUSB6250 provides the best bridge solution to meet both the performance and flexibility requirements of next-generation external storage devices. Contact TI for samples. * Available by end of November '02 USB to ATAPI Bridge TUSB6250 can support a total of 4 input and 4 output endpoints configurable as bulk (double-buffered), interrupt or control. USB o Fully compliant with the USB 2.0 mass storage
specification, including the host/device data transfer matrix subset (the 13
cases) o No hang conditionso Tolerant of ATAPI drive operational abnormalities, including random data request lengths and random pause/terminate/requests o Low-power architecture makes bus-powered Typical Applications
|
The d word.
Think BASIC-52 and Forth.
And making some money with some fun!
Sunday November 17, 2002 19:00
System software costs have to be kept low. Otherwise the app costs skyrocket.
And source code provided over Internet for application developers. For the same reason.
It takes app code implementers too long to figure out what the systems programmer have done.
Source code for wmd drivers, C dlls, and VB apps need to be provided for the same reason. Productivity.
Expensive development of new operating systems is not the answer.
And finally, good peripheral operating systems on the 80C32 end must be provided. This is where the greatest need is.
Answer is to use software technologies already developed. BASIC-52 and Forth.
Productivity increases from the Microsoft software development platform, USB 2.0, and intelligent super fast 80C32 SOCs coupled with interactive operating systems with incremental assemblers and compilers, like BASIC-52++ and 80C32 Forth, and, of course, Internet is and will cause major financial perturbations in high tech.
There are some very good articles contained on these pages about the market for high end peripheral microcontroller.
Csd is mainly into interface software between the 80C32 microcontroller and Windows. Csd simply doesn't see how the high end peripheral microcontroller companies can spend the bucks or con software companies to produce the required peripheral system software. This is too expensive considering the return on investment.
So we advocate going with super fast 80C32s SOC running BASIC-52++ and 80C32 Forth where the bucks have already been spent on development and doing the computations in a gigahertz PC connected to a 80C32s SOC over UBS 1/2.
Csd exercises the Cypress EZ-USB-FX machine by running memtest for long periods.
You can click on hold which stops the program.
Clicking on Run starts it running again. Some of the time. Other times it crashes and has to be reloaded. The real world again.
Why?
Vectoring into a BASIC-52 or Forth warm start might allow discovering the source of the problem.
Skeleton USB drivers and test code generated from DriverStudio 2.0 don't work. But, alas, this is the real world.
Numega DriverStudio is at about version 2.6 so this is not surprising.
But it's interesting and sort of fun to try to figure out what's going wrong.
Csd has to do both rs 232 and USB. At first! Then good bye rs 232.
This means modifying a ISA bus driver by adding USB code.
Csd experienced the worst Windows 2000 crash ever while experimenting with the software supplied on cd with Walter Oney's wdm driver book.
Some of the Oney software completely clobbered Cypress's EZ-USB-FX system.
But on Friday the Cypress system is working again.
Here's a jpg just taken.
Note the stupid Keil required rs 232 connector next to the usb connector.
We've got to get 80C32 BASIC-52++ and Forth running in a Windows Window coded in Visual Basic 6.0, C/C++ and inline assembler!
Think productivity and fun too!
Thursday November 14, 2002 20:24
Look for good deals in software where others have spent to develop. BASIC-52 and Forth, of course.
Recessions result when there is a shift in technology.
Visual Basic 6.0, Microsoft Office, USB 2.0, and super fast USB 80C32s hopefully running BASIC-52++ [tee hee, damn lawyers] or Forth are the new wave.
Magazines like Embedded Systems Programming catered to technologies which placed ads.
Not to technologies like Forth and BASIC-52 which did not or really cannot afford advertising.
But with the shift to Internet technology, you can now learn the true story of micrcontroller operating systems. This is the way it should be.
The contemplation of things as they are,
without error or confusion,
without substitution or imposture,
is in itself a nobler thing
than a whole harvest of invention.
Francis Bacon
The 80C32 Forth and BASIC-52 projects are very little invested, lots of people benefit, some companies who are asking lots of bucks for antiquated software get clobbered, and we all have lots of fun on Internet!
And you get to learn some weird Forth and BASIC-52 internals.
So let's get software working inexpensively and well. Over Internet, of course.
Silicon Valley got blindsided not only by Windows 2000 reliability but probably even more by Visual Basic 6.0.
This is particularly true when coupled with USB 2.0 and super fast 80C32 SOCs.
Who wants to spend money programming some archane microcontroller at the peripheral side with antiquated software? When you can do the job easier in Visual Basic 6.0 or C/C++.
Programmer productivity using Visual Basic 6.0 is an order of magnitude more than C/C++.
Sheriff's book not only explains in writing how VB 6.0 is integrated into Microsoft Office but Sheriff has a web site with code examples.
Csd wants to integrate the Microsoft software environment into peripheral Forth and BASIC-52++ [If you want to write lawyer letters for about 8 years. Or, even worse, pay a shyster to try to do it for you] through Windows.
We did the DOS interface about twenty years ago.
Wednesday November 13, 2002 10:27
Oops, csd mistakenly used PC/ASSEMBLER when PC/ASSEMBLER® is appropriate ... after 8 years with about one lawyer letter a year.
Here is PC/ASSEMBLER® at the patent and tradmark website.
Word Mark PC/ASSEMBLER
Goods and Services IC 016. US 002 005 022 023 029 037 038 050. G & S: computer programs and program manuals packaged as a unit for use in incremental interactive assembler programming on personal computers. FIRST USE: 19850101. FIRST USE IN COMMERCE: 19850401
Mark Drawing Code (1) TYPED DRAWING
Serial Number 73619885
Filing Date September 15, 1986
Supplemental Register Date August 7, 1987 Registration Number 2183073
Registration Date August 18, 1998
Owner (REGISTRANT) PAYNE, PATRICIA DBA COMPUTER SYSTEMS DOCUMENTATION INDIVIDUAL UNITED STATES P.O. BOX 5478 ALBUQUERQUE NEW MEXICO 87115
Type of Mark TRADEMARK
Register SUPPLEMENTAL
Live/Dead Indicator LIVE
Now to try to make some money with this. And have some fun too!
An Israeli was also tying to trademark PC/ASSEMBLER for a printed circuit board assembler.
So we had to defend our claim while the trademark office slowly over 8 years thought about our arguments.
Perhaps we should post the letter to see how frustrating dealing with Patents and Trademarks is?
You use while you are writing lawyer letters and hopefully waiting for ®.
Containing software development costs is very important, especially in a possibly deflating economy.
Visual Basic 6.0, Visual C/C++ are real bargains along with 80C32 Forth and BASIC-52 too, of course.
So getting Forth and BASIC-52++[tm, tee hee] is a top priority task for the success of selling 80C32 SOCs.
It took csd 8 years to get the official government trademark PC/ASSEMBLER® with a fair number of lawyer letters!
Users can't afford antiquated non-operating system development environments. You know, Keil.
Users need to get a small amount of software running quickly in high level and some assembler on their 80C32 SOC.
Users need an interactive development environment supporting incremental assembler and compiler which runs on the 80C32 SOC.
Users need a interactive operating system on a SOC so they can troubleshoot problems interactively over Internet rather than have to make service calls.
SOC, other than 80C32s, have a huge, and potentially ruinous, software problem.
The expense of development of operating systems and compilers, other than Forth of course, will be great.
Besides with USB 2.0 who wants or needs to write lots of software on a peripheral microcontroller when it can be done easily inside a gigahertz PC?
Tuesday November 12, 2002 07:43
Jim Turley writes The Death of ASICs in the 64 page Embedded Systems Programming magazine
It's a good time to be a programmer. Writing software is cheaper than creating hard-wired logic, it's easier and quicker to correct mistakes, and, best of all, microprocessors are getting faster than other silicon chips.
This may be true, however, the question of where a programmer wants to write code currently is in question.
On a peripheral microcontroller or in a PC.
USB 2.0 should shift the 'where' to inside the PC because of the software rich environment compared to the software poor environment at the peripheral end.
Embedded Systems Programming continue to get thinner likely because of reduced advertisers. The 'where?'
Monday November 11, 2002 17:17
High tech is at the focus of a deflation.
The Gigahertz AMD XP chips we buy at less than $100.
1/2 gig of PC133 ram is selling for less than $50 in Albuquerque this weekend.
But there is good news too. Get the job done without spending lots of bucks and have fun!
Forth, BASIC-52 [++, course] and 80C32 USB SOCS still look attractive for advantages enumerated on this page.
Who's going to pay off that debt? Likely few.
Who's going to lose their investments. Likely us.
So csd better get some interesting, somewhat profitable, and, fun projects going just in case retirement checks get smaller.
Forth is relatively easy to port to a SOC because most of the work has been previous done.
BASIC-52 is a bit more complicated. One reason is that BASIC-52 isn't disk based.
BASIC-52 compiled program like
100 a = 2
200 b = 3
300 print
a+b
is loaded into the BASIC-52 chip as a fast typest would do. No file disk buffers.
Forth on the other hand Forth has both terminal and disk buffer as seen in the memory map below.
BASIC-52 works in both the compile and interpret states.
For example,
here's a BASIC-52 line entered in 80C32 USB system terminal emulator.
When Enter is keyed 5 appears on the line below the a.
So BASIC-52 needs a bit of spiffing up to be presentable with a 80C32 USB SOC.
Floating point arithmetic, of course, should be done on the PC side of the USB interface. And maybe some other work too?
Csd's first task is to get 8051 Forth working on all Windows operating system in a Windows window coded in Visual Basic 6.0, Visual C/C++, and inline assembler from its previous DOS version which was coded in Forth.
Sunday November 10, 2002 19:52
When technology changes old players have no advantage over new players.
USB 2.0 and super-fast 80C32 [not the others!] SOCs are new technology.
Psst! You want to do the computations inside a gigahertz PC using Microsoft Visual C/C++ and Visual Basic 6.0 communicating with an 80C32 intelligent sensor running either BASIC-52++ [TM, of course,] or Forth over USB 2.0. Not in a software-poor obscure microcontroller.
TM?
Come on. No one makes, or should make, money selling the ideas of others!
We will do this for the consulting bucks! And cheap book fees, now in HTML, course. Along with all source code.
The peripheral microcontroller people can't afford to develop the required software.
They didn't think of the total solution. Or perhaps concealed their problem to venture capitalists? For business reasons, of course.
And, of course, they will suffer the penalty. Chapter 7 or 11. The real world, again.
Then distribute this over Internet to customers.
And provide the PC side VB USB app and wdm driver skeletons.
Just think what these 80C32 SOC/BASIC-52++[ TM, of course]/Forth application customers are going to do in the marketplace.
Simply clobber unintelligent competition who went with no or unestablished 80C32 operating systems.
This isn't much work provided you use the absolute genius [brighter than absolutely brilliant] software ideas of Charles Moore [a software Newton].
But Moore has some pretty weird ideas [only one file, the screen file]. But this has to be considered historically.
So you have to watch Moore's ideas in real world situations. And the Forthians too.
Companies have gone bankrupt using Moore's ideas.
Likewise the Forthians have chapter 7ed, not 11, a number of companies.
Forthians, in general, are most interested in messing with Forth than getting the job done.
The absolutely brilliant ideas of Intel [Tom Kautausky, sp?], Numega Driverworks, and 80C32 SOC vendor peripheral hardware ideas should be used.
But you have to acknowledge that Moore, Kautausky, the ladies at Numega, and some of the hardware SOC builders probably have done better than you can do it. They did, trust us.
Getting an interactive incremental compiling and assembling operating system working on a peripheral microcontroller, especially an 80C32, is totally fun!
When the operating system is working, the 80C32 becomes totally friendly. Your buddy which will do anything you tell it to do, even compile and assembler code from a disk file on a PC, and will let you interactively look inside of it.
We learned this at Sandia National Laboratories. The fun way.
Now we can about access these intelligent 80C32 running BASIC-52++[TM, of course] over Internet to reduce the cost of hardware maintenance and software distribution in source code, of course. For those who wish to participate, of course.
So let's the vendors who spent lots of money on Keil antiquated, expensive, and non-established operating system software go bankrupt in this likely deflationary economic cycle.
And their customers, too, who did not see the merit of using an established operating system on an 80C32. And went the C route, or even stupider yet, assembler route.
While we make money and have a fun time!
We, the participants, of course.
Csd generated a wdm driver skeleton with 2 bulk input and output endpoints, 1 each of an interrupt and control endpoints. And no isochronous endpoints. The weather was rotten yesterday.
But before usb communications the DOS Forth interface to the 8051 must be converted to a C++ wdm windows interface.
Next weekend quail season opens in New Mexico. This may slow down progress a bit.
The BASIC-52++[tm], Forth, Windows, Internet project looks to be shaping up to be a fun, worthwhile, and even profitable project for those, of course, who participate.
Look at someother neat, creative, and positive New Mexico projects. Kubelwagen shop tour and Aschwanden!
Friday November 8, 2002 06:46
Csd begins writing the C interface code to and the C++ wdm driver now that the terminal emulator is successfully capturing and displaying a line from the keyboard.
Here's the emulator with a Forth ok sequence.
Look at QUIT below very carefully. There is a leading blank on the first ok. In this way you can how Forth got where it is.
Csd types a Forth line.
Both BASIC-52 and Forth are two state operating systems. Interpret or compile. : starts a subroutine compile, ; stops a subroutine compile.
add adds the top two 16 bit numbers from the data stack and returns the sum. 2 3, now forth is back in the interpret state are pushed on the stack. add is then executed and 5 is at the top of the stack. . prints the value 5 the line below the :.
Csd hopes you can see where we are going.
1 Forth operates on an 80C32 system, like Cypress's EZ-USB-FX or maybe even some other intellegent SOC vendor's 80C32.
2 USB support is developed [ported?] to Forth and Forth assembler.From software standpoint Cypress did about everything wrong on the 80C32 side.
Keil, assembler, and no operating system on the 80C32.
3 RS-232 communications are replaced by USB.
The most important!!!
4 Working Forth assembler USB code is ported to BASIC-52++ [tm].
++, of course, since we have usb support floating point math can be done on the PC side and we don't have to waste 80C32 ROM or RAM with unnecessary floating point extension code.
Develop assembler code on a 80C32 Forth machine, then port it to a BASIC-52 machine is an old trick.
Since you develop it on the target machine with an interactive incremental assembler, the developer can get the code working in short order with no development system.
Forth is okay in some settings. Forth is totally incompatible with anything you have ever learned before. Therefore the numbers of adopters will be small.
BASIC-52 is lots better from the standpoint of adopters.
BASIC-52 is compatible with what engineers and lots of Visual Basic 6.0 programmers already know. Therefore, BASIC-52 will win in the marketplace.
Jameco sells the 80C52-BASIC for $19.95. An 80C32 sells for $4.49.
The cost of software development for an BASIC part is on orders of magnitude less than for an 80C32.
Experienced people discovered BASIC-52 and Forth apps are lots more reliable than software produced with other technologies. Namely C.
The 80C32 SOC vendors have an opportunity to have their parts become 80C52-BASIC++ [tm] parts by merely flashing the software. What an opportunity!
Give csd a call if you want to participate in this project.
Thursday November 7, 2002 15:34
Using the single step and debug window of Visual Basic 6.0, the problem with 'l' was located and a work-around included.
For some currently unknown reason "l" increments ListIndex after the subroutine updating the keyboard string has been exited.
In this project PC Forth code is being ported from a DOS app to Visual C++ at ring 0 and Visual Basic at ring 3.
The really important point is that nothing is being reinvented.
The ultimate enhancement for this project is switching from rs 232 communications to USB.
The PC side will be made to work directly in Windows rather than out of a DOS window.
The correct jpg for interpret has been fixed.
Csd is working on the terminal input for both Forth and BASIC-52. There was almost an absolutely brilliant solution ... in about a half dozen lines of code.
But, alas, a problem has arisen.
Two jpgs show testing the keyboard input for the maximum number of lines and characters per line.
When focus is on the terminal screen, the user begins to type then the typed characters appears one the line down from the last response of Forth or BASIC-52.
An "ok"
Notice that the text is highlighted in dark blue. We can tolerate that.
However, when an "l" [EL not ONE] is entered the dark blue highlight scrolls down [circular]. No other characters cause this!
The plan for all software is that one should try to do the job in the highest level language first, then drop down to lower levels as needed ... which may include building some hardware.
Csd will fiddle with Visual Basic 6.0 to try to eliminate both the blue background and l-scroll. But not too much time will be spent since this is not an essential material problem [legal parlance].
The keyboard entry works, so the next task is to pass the line from the ring 0 VB app to a wdm driver at ring 0.
Csd has a plan to migrate from rs 232 to USB which includes using rs 232 historical Forth and BASIC-52 to debug the USB code.
Csd, of course, has a Cypress EZ-USB-FX board. Included is a largely useless, expensive, and antiquated incomplete Keil C and assembler development software system. You need to pay over $2k to get a fully functional version of this antiquated system! And, if memory serves correctly, you need to pay an annual maintenance fee.
But we're going to try to do something about this! And have fun while doing it too.
Keep in mind that one ultimate goal is to be able to log on to a 80C32 SOC over Internet for interactive troubleshooting. And also source code software maintenance and upgrades!
Both BASIC-52 and Forth running on an 80C32 have incremental compilers and destructor capability.
Plus Forth has an incremental interactive or batch assembler!
Wednesday November 6, 2002 06:32
Csd is using part of the cover of Embedded Controller Forth for its icon.
See!
Windows 2000, which you can see some of, is being used.
Visual C/C++ works lots better on 2000 than on 98! Tuesday November 5, 2002 12:59
We have at least on 80C32 operating system running. The USB 80C32 SOC.
However, we can have one or more non-USB SOC operating systems running communicating first with the USB SOC either parallel or serial. The USB SOC passes terminal and disk data to the PC of the additional non-USB SOCs over USB using additional endpoints.
Therefore there can be more than one 80C32 operating system windows running at the same time. For this reason all operating system sub windows will all be MDI childs with the main operating system window.
Today is the start of the Remote 80C32 intelligent sensor/Windows project. You know, actual coding! The design has been worked out mentally after lots of study and planning.
Goal is to be able to log on to an 80C32 intelligent sensor running either a Forth or BASIC52 operating system [C is not an operating system. It's an American disease!] over Internet. This is to be used for remote hardware diagnostics and source code software maintenance of intelligent sensors.
We are going to do this with the least amount of work possible [absolute brilliance requires total laziness] using Visual Basic 6.0, C/C++ 6.0, in-line assembler, MFC, and Microsoft Office.
We aspire to absolute brilliance. And not having an unteathered creative psychotic design episode at the keyboard.
Absolute brilliance means that you use the good ideas [opposed to bad ideas like the Keil expensive and antiquated development system] of others. In this case the ideas of Moore and Katausky at the 80C32 end and gigahertz PCs running Microsoft Windows at the PC ends.
Automobiles dealers learned that when the located their dealerships close together they all did lots better financially.
The same logic likely applies to 80C32 SOC vendors.
This software should be made to run on as many different 80C32 SOC vendor's machines as possible.
But communications with the PC is through USB. So those vendors who didn't implement USB will be required to have a USB equipped 80C32 between their part and a PC. Look at it this way, the 80C32 USB SOC simply replaces a rs-232 line driver/receiver chip.
The legal project has halted.
The lawyers have to figure out what they are going to do. If they don't do the right thing, like settle, its criminal complaint affidavit writing time again on November 14, 2002.
Read more about the adventures of white ford and grey rabbit.
He who get up with the bright.
His kite will fly right.
Benjamin Franklin?
This is senior citizen stuff!
The last time bill did an 8051/PC project, bill did the 8051 side of the software.
Joseph Flores did the PC side.
This time bill is doing the PC side using the most modern tools.
At the time of 8086 PCs were used. Forth was okay on a PC then. But not now. We're using Microsoft's outstanding tool.
BASIC-52 and Forth on the 80C32 side is still the ticket today as it was then.
Monday November 4, 2002 15:17
When picking anything one needs to have in mind a set of criteria that one is trying to satisfy. You know, the requirements.
In our search here, we a looking for the best solution for an operating system on an 80C32 SOC.
There are other real time operating systems for an 80C32 but when you start to match their capabilities, advantages and disadvantages with that of BASIC-52 and Forth they come out very poorly.
Body of knowledge of both BASIC-52 and Forth has been established over 10-20 years which is going to be hard to overcome by anyone who thinks they can do better.
And those likely self proclaimed geniuses who embark to design a new 80C32 operating system, rather than enhance BASIC-52 or Forth are likely to create a mukinfuss. And go broke in the process too.
A mad scientist crossed an abalone with a crocodile hoping to produce and abodile.
But, alas, things did go quite a planned.
What resulted was a crockabaloney.
Here's a fun Laissez Faire Electronic Times Internet article on cryptographic algorithm criteria which also tells the etymology [et's be intellectural] the word mukinfuss. Sandia National Labs again.
Csd is going to start designing and writing code to connect both BASIC-52 and Forth machines to Windows so that 80C32 SOC intelligent sensors can be interactively logged on to over Internet.
This will be written in Microsoft C/C++ 6.0, in line assembler, and Visual Basic 6.0.
Not Forth!!! We made a mistake here a long time ago. Now is the time to correct that mistake.
The Forth metacompiler should have been written in GW BASIC! Gee whiz.
Just think how easy that would have been to port of Visual Basic 6.0! Sunday November 3, 2002 09:11
The other project is progressing. We all have, or should, do different things. Ligation was forced on us by lawyers. And we are responding as best we can.
Look at all the interesting stuff happening in the south of Brazil. But we're doing some real interesting stuff here too.
High tech is real interesting and real scary too. Too much for the marketplace?
Let's look at bit more carefully where the Brazilians are that are doing the really neat BASIC-52 and Forth work. Here's a map of Santa Caterina. Look at Florianopolis on the island and Blumenau.
Then let's look at the German influence.
Germans migrated, or escaped, to south America at the end of the second world war.
Picture of the kubelwagen might bring tears to the eyes of some of the FURB and Federal University of Santa Catarina students' grandfathers.
This internet stuff is getting more and more fun!
And Morales and Payne caught two New Mexico supreme court employees in a Title 18 criminal violation of federal law. So we'll take some time to post here for a SOC vendor.
Dr Richard Hanson and and bill experimented with Microsoft Netmeeting. It is possible to log on to an 8051 Forth development system on Internet using Netmeeting. This must be characterized as absolutely brilliant because we didn't do any work. Others did the work.
Those 80C32 SOC vendors who use the ideas of other to make their product very useful quickly and inexpensively will be a financial success. Think BASIC-52 and 80C32 Forth. Keil C cross development system is a bad idea.
Csd received some email from a SOC vendor asking details of Forth memory layout. So to be absolutely brilliant we have good news.
Mitch Derick and Linda Baker have a great layout for you to look at. Their book contains 8085 code.
Sandia labs had Henry Neugass redo the Derick and Baker book for our 8051 Forth. This is posted at this website.
Now for the bad news. The Brazilians may have wasted their time if they didn't use Netmeeting to solve the problem.
But any group who uses BASIC-52 and Forth are, of course, brilliant. But not absolute brilliant.
Absolute brilliance requires total lazyness so you have time to enjoy the finer things in life. One of which is not sitting in front of a computer!
And those who use Keil C cross development system are, of course, absoletely stupid.
This website is about making money and not going broke before this happens by wasting lots of money on a C cross development system.
You may wish to read some of the business articles at the Prudent Bear. Tice and company have used their absolute brilliance directing you to the works of others.
Depressions, not recessions, are usually proceeded by a high tech shakeout. Bill was told by an engineer on the Alaska ferry Kennicott.
Improved steel production methods in 1896; Electric motors in 1929.
The engineer told bill his ship was also stopped and boarded by Iranians during the Iraq-Iran war. The Iranian boarding officer got his BS at Mississippi state, the engineer said!
Csd predicts that USB 2.0, fast 80C32 SOCs, gigahertz PCs, Microsoft exquisite Visual C/C++ and Visual BASIC are going to do the number on many peripheral hardware and software [Java] companies. Friday November 1, 2002 17:17
The BOOT-UP LITERALS and metacompiled nucleus and app part of the dictionary are in ROM on a ROMed system. The remainder of 8051 Forth is in RAM. Of course, its best to immediate copy the ROM portion of Forth to RAM and execute entirely out of RAM. |
Readers
Lot's of the hype about C is IT people trying to enter the embedded controller business. C has lots of disadvantages. Not an operating system. Memory hungry. Rotten memory management until fixed with constructors and destructors in C++ [Forth uses forget and BASIC-52 uses CLEAR for destructors].
C is an established business trying to establish itself in embedded controller software.
Time, we hope, to open your mind to others ideas. BASIC-52 and Forth. And make some money with your apps instead of spending it on Keil antiquated software.
And, of course, have fun while doing the software easily and interactively.
Others have done the work to allow us to do this! Thursday October 31, 2002 19:04
Bill sadly must return to his other project today. But this is a very important project too. The US legal system needs some fixing!
Hopefully this will give the 80C32 SOC vendors time to come up with the right decisions.
BASIC-52, Forth, USB 2.0, remote BASIC-52 and Forth
Zeitgeist again. And you thought you were first? Scooped again!
Look at what the Brazilians are doing. Remote Forth too. Scroll to the bottom of the left frame.
Absolute brilliance. We didn't have to do it. Thank goodness. They did it.
And, thanks to internet, we're in touch with some of the brilliant Brazilians!
Sent: Wednesday, October 30, 2002 7:03 PM
Subject: Re: Re: REXPIC
Very good this documentation!
Thanks!
How to get more about it?
How did you find me?
Miguel
Miguel, Google found you with a search on BASICON.
NIH - not invented here.
NIH is the mark of the truly uneducated and stupid person.
See that the Brazilians are funded by ATMEL who build an 80C52.
ATMEL also builds an ARM!
But as we know, USB 2.0 and super fast 80C32 SOCs running both BASIC-52 and Forth are going to do the number on ARM, MIPS, ARC, TENSILICA, ... etc.
The Sandia seismic data authenticator, seen below, used ATMEL eeproms.
Note the dates of 1997 and 1999. It always takes a lot longer than you think! But this is the real world.
So let's study what the brilliant Brazilians did. Carefully, slowly, and with great thought.
Think fewer maintenance trips! Wednesday October 30, 2002 20:11
BASICON bit the dust!
Csd's 8051 forth assembler was ported to BASICON and derivative boards running BASIC-52.
All readers please note these words carefully.
When we started with Forth at Sandia labs we did this so that we didn't have to purchase a development system. We had an onboard interactive incremental compiler and assembler operating system. We had source code.
And we could do most of the work in high level Forth as opposed to assembler.
Sandia labs had disasters with large assembler codes ... with US nuclear bomb projects!
What we did not realize is how much more reliable a Forth [or BASIC-52 too] operating system is.
Ray Duncan was the first, Bill remembers, to comment that when you work with Forth day in and day out you realize how reliable the technology is.
This has absolutely been csd's experience. And others who make money off from these super reliable systems. Even if they didn't realize it then. But they may now after trying alternate technology.
And we had a simply fun time doing the Forth work! And, of course, bill wrote a book about how we had all of this fun!
And bill thinks he knows one reason why both Forth and BASIC-52 are more reliable. Reason is in the forth source code.
Bill is running Forth in a DOS window under 98SE on machine while he is composing this web page.
Forth spends its time in an infinite loop BEGIN ... AGAIN. See
Notice the RP! This intitalizes the return stack pointer. The return stack starts out fresh every time through the loop.
A Microsoft employee said that if you start Windows and let it run, it will eventually crash. Microsoft does not know why?
One likely cause might be a stack problem.
There are only two stacks in a Forth operating system the data stack and the return stack.
Everyone must keep in mind
- C is a compiler not an operating system.
- Forth is a real time interactive operating system.
- BASIC-52 is a real time interactive operating system.
- C is another "Oh hum" compiler. Fortran ported to a PDP 11.
- C is an American disease!
- Charles Moore is lots brigher than Dennis Richie, if computer ideas are weighed!
Dennis M. RitchieRitchie came to Bell Labs in 1967 from Harvard University, where he had completed undergraduate work in physics and a doctorate in applied mathematics. Soon, he began helping Ken Thompson create Unix for minicomputers. He later transported Unix to the Interdata 8/32. The foundation for this portability mas a general-purpose language created in 1972 by Ritchie, who added data types and new syntax to Thompson's B language and renamed it C. In 1978, Ritchie co-authored The C Programming Language with Brian Kernighan, introducing this highly efficient language to the world. Since then, C has become the most widely used language in computers ot all sizes. The American National Standards Institute eventually established an ANSI standard for C.
ELECTRONIC design, October 21,2002
Electronic design has in the Hall of Fame 2002 Inductees John Backus for FORTAN, Grace Hopper for COBOL Gary Kildall for CMP, and John L von Nuemann for the Princeton architecture ... as opposed to the Harvard architecture.
No Charles Moore.
No Kemeny and Kurtz for BASIC.
No Bill Gates for BASIC.
You can send an email to David Bursky Editor in Chief of Electronic design if you want.- Memory vendors simply love C.
Below is a flow chart of interpret. Here's interpret
And where is this code coming from?
Out of the Sandia labs 8051 Forth operating system.
Who owns this code? You do taxpayer. You sent money to the feds with income tax, the fed sent some of your money to the National Security Agency, NSA sent some of that money to Sandia labs. And we spent your taxpayer money developing 8051 Forth. And had a simply great and fun time doing this.
But now it is time to put the code, wdm drivers, USB 2.0, pci code to work to make money for some of the 80C32 SOC vendors. And us, too, of course!
We are following Palm which ported its operating system from the Motorola dragonball to an ARM. The project looks like it is done.
But looks what's happened to Palm financially. A financial disaster.
80C32 SOCs need either a BASIC-52 variant or Forth operating system connected to a PC through USB 2.0. This is very inexpensive to do since the technology is known and source codes are available.
There is a problem with the SOC designers, however. They add too much FLASH code memory and not enough RAM.
BASIC-52 fits in 8k of code. But required external [additional chip] RAM.
Here's an excellent site on BASIC-52.
One goal of this proposed work is already in place. Zeitgeist again.
Try out the 8052-Basic chip on the web. This is an actual system that you can control remotely. Enter commands and enter and run your own programs. From the Remote Experiment Lab at the Universidade Federal de Santa Catarina in Brazil. (In Portuguese and English).
The goal is to troubleshoot an intelligent sensor having problems over internet! Fewer maintenance trips!!!
Sandia labs 8051 Forth which includes an ROMed interactive incremental 80C32 assembler and full screen text editor [Flores and Payne got a bit carried away] fits in about 18K bytes of memory.
So let's try to clue the 80C32 hardware SOC designers on what should be done with 80C32 memory architecture. They did an overkill on FLASH and underkill on RAM.
FLASH, csd was told is very expensive to put on a chip. Thursday October 31, 2002 09:57
Here's ideal 80C32 memory model A full 64k bytes of code and ram memory minus 2. 4. 8. 16. 32. 64. 128, or 256 bytes which decodes to external parallel i/o. In Embedded controller Forth for the 8051 family decoding was for the 256 bytes of i/o space. Also essential is that both the Harvard [separate code and data spaces] and Princeton [overlapped code and data spaces] architectures should be software selectable. This is implemented by anding -rd and -psen. At Sandia labs we started with 8kx8 byte rams and eproms. But at the end used 32kx8 eproms and rams. The Sandia seismic data authenticator used on 32kx8 eeprom and one 32kx8 ram. HOWEVER, if we had a 64kx8 ram available then we would of used it. Now listen up hardware engineers. The first then we would have done is to copy the 32kx8 [0 to 7fff, 0 32,767] eeprom into ram. This is done by copying code into the high 32kx8 memory to execute. Then we would have deselect eeprom and run entirely out of ram. But we had real estate problems and a power budget so we couldn't do this. Hardware engineers, the interrupt vector addresses are in low memory. Setting these into FLASH is a bad idea! Now if USB is connected software can be dowloaded from a PC, compiled, assembled, linked and loaded by either a BASIC-52 or Forth operating system. All of this is cheap to do some a software standpoint. Let's all hope that one or more of the 80C32 SOC builder see the merit of this proposal. Venture Investment in a Dive Rising Risk Scares Off the Capitalists BY MICHAEL LIEDTKE SAN FRANCISCO - With Losses from their high-rolling days still piling up, shell- shocked venture capitalists continued to shun new risks in this year's third quarter, dropping the industry's investment activity to a 41/2-year low, according to a report released Tuesday. Venture capitalists invested $4.48 billion in startups during the period ending Sept. 30, the weakest quarter since the first three months of 1998, according to a survey compiled by PricewaterhouseCoopers, Venture Economics and the National Venture Capital Association. This year's third quarter represented a 48 percent decrease from the same time last year, when venture capitalists poured $8.68 billion into startups, the report said. It also marked the ninth consecutive quarter in which venture capitalists curtailed their investments from the preceding three-month period. The reasons for the downturn have remained largely unchanged since the Internet gold rush turned into a financial bloodbath during the spring of 2000. As the stock market began to turn a cold shoulder to dot-coms and other high-tech businesses, venture capitalists found themselves stuck with unprofitable startups that no one else wanted. Meanwhile, even promising startups are finding it increasingly difficult to find customers interested in spending heavily on technology, further reducing their chances of survival and saddling venture capitalists with the worst losses in the industry's history. "We all have had a very cold shower," said Bob Grady, a venture capitalist with the Carlyle Group. Most venture capitalists and analysts believe the industry's investments will dwindle even more in the next few quarters. "We haven't seen the end of the decline," said Robert "Robin" Bellas, a general partner with Morganthaler Ventures. "My gut feeling is that this won't stop until we get down to $2.5 billion to $3 billion per quarter." Venture capitalists have responded to the adversity by shoveling more money into the best startups in their existing portfolios and investing less in new opportunities. The number of startups that received their first infusion of venture capital during the third quarter totaled 159, the lowest number in nearly eight years, according to the report. Back in the heyday of dot-coms in late 1999 and early 2000, nearly 1,000 startups per quarter were getting their first dose of venture capital. Albuquerque Journal Wednesday October 30, 2002
|
Csd has a Cypress EZ-USB-FX development board.
Cypress's design ideas for USB are very good, in csd's opinion. Both Harvard and Princeton architectures, no ALE, great manuals, ... But there may be several problems. Cypress's part is NOT [keep looking as we fix screw ups! But this is the real world] a SOC.
Before devoting lots of time to a chip, one should make fairly certain that the chip will be around for a while. There is some doubt of that in Cypress's case.
Messages report that Cypress will be forced to file for bankruptcy within several years.
TI continues to look better! And let's all hope that Cygnal is able to pull it off!
We must all keep in mind that we are in this for the money! And a good time while we make it.
Csd really needs to post some stuff on the classical Japanese company. For this is the way business should be. Not the messes in business we are currently into.
Csd has a box where this is stored too! In the garage. This was found while searching for the BASIC-52 material. Tuesday October 29, 2002 18:20
GAAP accounting KILLS the figures ... by: albea_2000 Long-Term Sentiment: Strong Sell 10/29/02 01:17 pm According to the rules of GAAP accounting, Cypress Semiconductor earned $223 million from operations in the last five years. But, as I recently pointed out in an open letter to CEO T.J. Rodgers, that figure doesn't account for 'one-time' charges that repeat every year. Nor does it factor in the consistently rising cash cost of capital expenditures in the semiconductor industry. And, most importantly, because investors pay the cost of stock options, not the company, it doesn't factor in one of the company's principle expenses: compensation. These expenses, whether or not T.J. Rodgers desires to include them in his 'pro forma' earnings report that he gives to Wall Street (and defends in the media), have an impact on his company's balance sheet...and an even bigger impact on yours. When you include all costs, rather than making $223 million over the last five years, the true value of Cypress's contribution to our economy was a loss of $503 million. If you doubt my figures, consider that in 1999 - at the height of the biggest boom of all time in the semiconductor industry - Cypress had to take out a $500 million loan. For the first time ever, the company would carry a significant amount of debt on its balance sheet. If the company was really earnings millions each year, why borrow money? And why that much? Today the firm's debt load stands at $517 million, and in the midst of an industry crash, this debt threatens to absorb the entire equity value of the company. Since I published my letter, shares have fallen from $12.00 to under $5.00, and Cypress has not rebounded strongly with the rest of the sector over the past two weeks. Frankly, its currency has been debased. |
Csd speculates that USB 2.0 and 80C32 SOCs may be a next high tech wave. There are lots of sound reasons to go this way.
Csd found its BASIC-52 box on Sunday.
The last BASIC-52 projects were done about 10 years ago. In one project 8051 Forth was used to write and debug assembler code which was later ported into BASIC-52.
The need for an efficient and inexpensive way to do interactive 80C32 peripheral side software coupled with Windows,USB 1/2, Visual C/C++ and Basic 6.0, and wdm drivers prompts a careful look at how all of these technologies can be used together to solve the problem.
A possibly fun retirement project!
The MSC BASIC-52 manual was given to Bill by Intel employee Mark Thompson along with the source code for BASIC52 on the date on the cover.
Thompson is dead.
Boutelle is dead. His ashes tossed into the sea at Half Moon bay.
But the ideas which both helped to develop and perpetuate are alive. Both Boutelle and Thompson live after death with their ideas.
Think 80C32 SOCs! And USB 2.0!
The rules have changed.
And we are about ready to take advantage of the new rules!
And, of course, have some real fun! Without too much work.
Other very brilliant people have done most of the work. I one cares to take the time to study what they did.
NIH [not invented here] is the kiss of financial death. Never reinvent.
Original meritiorous invention is fine. But this rarely happens.
Highly intelligent, especially in their minds, hyperactive, undereducated, unpublished and perhaps psychotic designers should be avoided. This is a recipe for disaster! For these purple designers do worse than merely create messes. They create mukinfusses!
Undereducated does not mean lack of formal education.
Jerry Boutelle only had a high school education and an honorable discharge from the Coast Guard. But Boutelle was extensively read, knew about all microcomputers in detail, knew C, assemblers, and wrote the absolutely brilliant Nautilus Forth metacompilerS.
What were are doing here is very un-university oriented.
Forth and BASIC knowledge is not pushed at Universities or maybe even understood, other than maybe at Dartmouth and University of Rochester.
We must keep in mind
College professors teach what they know. They don't teach what they don't know.
bill payne
What you are looking for in a designer is an absolutely brilliant, highly educated, published, rational, and totally lazy designer.
Absolutely brilliant people realize
1 there are a lot of really smart people out there
2 and they aren't going to do it all themselves.
So absolutely brilliant people
study the works of others long and hard, evaulate the merits of the
works of others in light of problems to be solved, then try to enhance and
extend previous work.
Zeitgeist! Others too, are doing this,
1
2
3
4
5.
There must be an alternative to Keil. Tuesday October 29, 2002 16:12
We are on the move.. And aspire to become absolutely brilliant! by getting BASIC-52, with some enhancements [this is MORE THAN SECRET, NSI, NOFORN, ORCON, ... since we don't want others like the mechanic to know what we are going to do] working on one or more of the 80C32 SOCs. Why do absolutely brilliant people need to write about the importance of porting a variation of BASIC-52 to an 80C32 SOC? They don't. Intel has done this for us!!!
So here's the stuff that has to be ported to a 80C32 flash based SOC.
BASIC-52 required a terminal emulator. Csd used the terminal emulator MIRROR for two BASIC-52 projects. This must run out of a DOS window now. And possibly won't run at all. This need to be converted to a windows
application. |
What TI is saying in its MSC1210 ad is so important, and hard to read below, that it is scanned in for all of us.
One must digitize at the sensor to avoid some really nasty real world problems.
Digital noise fouls low level analog signals. And what can be equally bad is that really high level analog noise can upset digital circuits which causes skipped Mux channels or even microcontroller software crashes.
TI/Burr-Brown likely knows this.
Let's hope TI is smart enough, like Intel, to get Forth and BASIC-52 working on their 80C32 parts. And to get their parts USB powered and connected to a PC via USB 2.0.
Cygnal, of course, is messing with the 500 pound gorilla in the market place.
TI/Burr-Brown advertises
24-bit ADC with no missing codes
We wonder if Cygnal can make the same claim?
Yesteday at about 16:00 csd got a phone call to participate in a survey.
The lady asked if we did microcontroller work. Then proceed to ask 8 bit?, 16 bit? 32 bit? 64 bit?
Csd undiplomatically explained to the lady that because of USB 2.0 and gigahertz PCs that those involved in 16, 32, and 64 bit microcontrollers are very likely to go bankrupt.
Because the intelligent decision is to use a super fast 80C32 to collect the data, then ship the data over USB 2.0 to a gigahertz PC, and do the computation in C/C++ or Visual Basic 6.0.
This, of course, will drag the Javans down the tubes with the unintelligent hardware vendors. But Forth software vendors have an unparalleled history of going broke!
Tuesday October 29, 2002 07:20
TI is the 500 pound gorilla. It sleeps where ever it
wants.
Intel charted the correct course in software for its 8052 part with MCS® BASIC-52.
80C52 Forth is also good. But, hey, we realize Forth is not as popular as BASIC.
And we have some SECRET INFORMATION to make BASIC52 every bit as powerful, if not more, than Forth!
We can make BASIC52 a lot more friendly to engineers ... with an interactive incremental integral 80C32 assembler! Which closely resembles Microsoft's exquisite C/C++ inline assembler. Which Microsoft appeared to copy from a Forth assembler!!!
What Intel did with the Katausky [sp? - Mark Thompson only prounced the name when he gave bill the book and source code] BASIC-52 operating system is simply brilliant!
Now the question is what is TI or Dallas going to do? Follow Intel's lead?
Or do something simply stupid in software like Cygnal or Cypress? Monday October 28, 2002 20:32
|
Those 80C32s SOCs without USB support maybe in fairly good shape if Cygnal does the right thing with its upcoming USB part with on-board 5v to 3.3 regulation.
Cygnal is going to on-board regulation to drop the USB 5 volts [100 to 500 ma] to 3.3 volts.
Let's all hope Cygnal has a 3.3v output pin on their USB part which can be used to power a second 80C32 SOC!
There are a number of ways, both parallel and serial, that one 80C32 can talk to another.
This way a non-usb 80C32 SOC can use the USB connected 80C32 to communicate over usb. There, of course, is somewhat of a speed penalty.
Q Can an 80C32 SOC do real time floating point digital filtering?
A Yes, provided its connected to a gigHertz PC through USB 2. And you do the filtering in an app running in ring 0.
Real time peripheral computing looks to become real fun! Friday October 25, 2002 10:23
The original cygnal development system has two Cygnal
parts. One 80C32 is U1, the fly speck chip, seen in the jpg. This communicates with a second 80C32 in a separate adapter box. Both are powered from a transformer which, of course, and communicating with an the antiquated Keil C development system over rs 232. These were disastrous blunders. The upcoming USB part development board hopefully will have 3.3v out of the Cygnal part and 5v available from the USB cable on a single development board. The prototype area can then contain a second 80C32 SOC connected either parallel or serial to the Cygnal USB part, if necessary for an application. This will provide a two, one fly speck sized, chip solution to many real-world problems. Having Forth and/or a modified BASIC52, each with an interactive incremental compiler and and assembler, working on both parts would be very helpful. Helpful because 1 Cost of the software development system is neglible. 2 Speed at which the 80C32 application can be written and debugged is on orders of magnitude faster than on a cross development system. 3 Forth and BASIC52 apps are super reliable in the field. 4 Most of the app development can be done in high level Forth or BASIC. 5 If additional speed is required, then Forth or BASIC subroutine are replace by assembler equivalents. A separate Window for the USB and non-USB 80C32 would be used to interact with each of the parts. The Windows 2000 8051 forth seen below at 1 and 2 is running in a DOS window running as a virtual 86 machine over rs 232. This needs to be changed to a Windows ring 3 app window communicating over USB 1/2.0. USB communications will solve the problem of a very fast PC communicating with an 8051 family microcontroller which has haunted the Sandia 8051 development system since it inception in the early 1980s. . |
Realization with USB 2.0 running at 480 Mbits/sec is that it is unnecessary to have an ARM, MIPS, or DSP expense at the periphiperal end when computations, can be done faster in a gigahertz PC.
And look at all of the mixed signal hardware in some of the 80C32 SOCs!
Psst ... it's fairly easy to place a 80C32 usb 1/2 processor back-to-back with another without usb support and have Forth run on both of them. Each in its own Window.
Maximum cost is two endpoints for the usb part and 2 endpoints for the non-usb part.
The gigahertz PCs can apparently outperform DSPs with its exquisite NDP [numeric data processor] floating point hardware! This is not lost on those companies, like TI, who are into DSPs.
The software may be done as a ring 0 app or perhaps a ring 3 dll or Windows app.
But if you still see the merit of using an ARM, csd alerts to to ARM Forth.
When we decided to use 8085 Forth at Sandia labs, we required a ROMed 8085 incremental assembler. The 8085 assembled used <BUILDS DOES>.
Jerry Boutelle's Nautilus 1 metacompiler did not do <BUILDS DOES>. But the Natuilus 2 metacompiler does!
Bill, Joseph Flores and Konrad Roeder [hardware technician and mensa member] visited Ray Duncan in Maria del Ray to pick up the software.
While Duncan was editing-in <BUILDS DOES> add-on software into Nautilus 1 metacompiler software Duncan just received from Microprocessor Engineering, we went swimming in the Pacific Ocean.
When done right, Forth and BASICs are fun!
Csd speculates that the superfast 80C32s with USB 2.0 are simply going to clobber the ARMS, MIPS, and other MPUs in the marketplace.
Reason is that peripheral hardware costs go way down, while the software is cheaper to develop in the PC using Microsoft's Visual C/C++ with its exquisite Forth like assembler with Forth like local labels, and Visual Basic apps.
But we have to get Forth and BASIC52 running intelligently on one or more 80C32 SOCs.
Current leader in USB 80C32 is Cypress
as Keil discovered.
Cypress got rid of ale, has usb 1/2 support, a dma engine, and, VERY IMPORTANT, supports both the Harvard and Princeton architectures.
But the Cypress parts aren't a SOC. Alas.
But watch out for Texas Instrument once it gets moving with its TI/Burr Brown 80C32 SOCs.
TI has the resources to roll over about any other semiconductor company if it chooses.
And TI has rightly chosen to go with the 80C32. But also does ARM for the Palm ... where the OS still doesn't seem to be working yet. Thursday October 24, 2002 13:01
IP giants face winter of market's discontent By Michael Santarini There are troubled time for micropocessor intellectual-properly vendors. Even giants like ARM and MIPS are retrenching as their stock prices falter. But dwindling royalties and licensing revenues era just half the story. A source close to MIPS Technologies said the company faces an impending rebellion among licensees. According to the source some are unhappy that the conceded the 32-bit MPU market focus instead on developing a 64-bit core, which make up a tiny fraction of the market for embedded processors. For the past eight years MIPS has been hemorrhaging market share to 32-bit embedded-core kingpin ARM Holdings and is now being threatened by the PowerPC for the No.2 position in RISC MPU shipments, according to a report by computer industry consultant Andrew Allison. Some MIPS licensees are said to be preparing an uprising that, if successful. could lead to a change in management and a new direction for the company, the source said. But the malcontents would almost certainly face a fierce resistance from MIPS and some of its largest backers, such as NEC and Toshiba, both of which work closely with MIPS on 64-bit architectures. MIPS could not comment last week, because it was in a mandatory "quiet period" leading up to the release this week of its quarterly earnings report. Judging by its stock price, MIPS certainly appears to be a potential acquisition target. The stock has been hovering between $1.5 and $2 per share for the past month. And for the third quarter, the company preannounced that it expects to report a net loss per share of 9 cents to 10 cents on revenue of approximately $9.4 million. The company is trying to turn things around; at the same time that it preannounced its revenue shortfall, MIPS also said that it was reorganizing its sales force and closing its services center in Denmark. Yet it remains to be seen whether MIPS, with a market cap of $64.4 million and rights to an MPU architecture worth arguably many times that cap, can turn its business around before it becomes a serious takeover target. ARM, which for some time seemed immune to the downturn, is having its share of troubles too. Last week the Cambridge, U.K., chip designer announced its first-ever fall in quarterly sales. It is also laying off about 10 percent of its work force, the first time ARM has had to reduce head count. Along with the announcement of its third-quarter results, the company said it would take a $3 million charge in the fourth quarter and lay off about 100 people, bringing the level back to the 700 it employed a year ago. ARM said it is taking a cautious view of the short to medium term and that revenue is likely to be "flattish" for the foreseeable future. ARM blamed the drop in sales on a dip in the number of new licensing contracts. The company recently said that Texas Instruments, Qualcomm and LSI Logic will use the ARM core in future products. But ARM admitted earlier this month that it sold just eight licenses during the second quarter, compared with 27 for the previous three months. Peter Clarke contributed to this story. Electronic Engineering Times October 21, 2002 |
The goal of the 80C32 USB SOCs, Forth, and BASIC52 is provide a useful product at a reasonable cost.
Holding down software costs is essential. Buying a $99 Cygnal development system, then having to pay about $2,000 to Keil is simply unreasonable.
Stock prices of Texas Instruments TXN plummeted. Cypress's stock prices are also in the tank CY.
So maybe it's time to curtail unnecessary 80C32 SOC development software costs? And make some money!
We, indeed, live in interesting times.
Csd has identified many 80C32 SOC players.
A USB SOC [The Cypress EZ-USB-FX/2 is not a SOC] should be a major player soon.
Who is going to solve both the hardware and software problems? We'll see.
Csd is using an Epson Perfection 1250 USB scanner using a AMD 1.6 gigaHz XP running Windows 98SE.
Scanner using software occasionally announces that someone else is using the scanner USB port. The system must be rebooted to make things work. But there's apparently a driver fix at the Epsons website.
USB still is not without problems but the advantages are so numerous [built in multiplexing, power through the cable, hardware flow control, ...] that csd believes this is the wave of the future! Especially if the peripheral host in an 80C32 SOC running a super reliable operating system like Forth or BASIC52! Wednesday October 23, 2002 14:30
VIEW FROM THE
TRENCHES A PLANET ANALOG GUEST COMMENTARY USB for handhelds demands a by Dave Bell, Vice President & General Manager The USB has finally achieved status that its name foretold-it has become truly universal. Since its inception less than a decade ago, more than a billion USB ports have shipped. And today, every desktop and notebook PC ships with at least one USB port. Several recent enhancements are adding to the original USB's capability. The need for higher bandwidth communication-partially fueled by competition from Firewire (IEEE-1394)-has been answered by USB version 2.0. This bandwidth allows images, MP3 ifies and the like to be transferred much faster than before. However, two other trends are equally important for handhelds. Cellular phones, PDAs, digital cameras, MP3 players and wireless communicators are now beginning to tap available USB power to recharge their batteries. USB power supplied by the host allows batteries to be recharged at the same time data transfer is occurring, and in some cases, eliminates the need for a bulky wall adapter. This power resource has existed since the inception of USB, but constructing a charger that accurately limits current drawn from the USB to either lOOmA or 500mA (as requIred by the USB specification) is difficult to implement using standard components. New power management ICs are about to simplil~r this task, automatically limiting current drawn from the USB and diverting excess peak power demand (to spin-up a disk, for example) to the battery. The new USB 2.0 specification defines another capability: USB On-The- Go." The USB architecture defines a master-slave relationship whereby the master supplies USB power to one or more slaves. With the original USB a slave device such as a digital camera cannot communicate directly with another slave (a printer, for example). USB On-The-Go allows a device to assume either a slave or a master personality so that handheld devices (traditionally slaves) can communicate directly with other peripherals without the need for a PC host. In order to become a USB master the handheld device must deliver regulated, fault-protected 5V power to the USB. Updated USB capabilities such as higher bandwidth, battery charging, and USB On-The-Go ensure the continued growth of this already popular protocol. At the same time, power IC manufacturers are beginning to deliver tiny application-specific ICs that enable designers of handheld products to exploit these capabilities with little additional effort. Between Bob Dobkin, the inventor of the three-terminal regulator; on one side of him at LTC, and the analog enthusiast Jim Williams on the other, it's a wonder how Dave Bell gets any attention. Yet he has become a steady contributor to EE Times' special sections, and - as this piece indicates-an authoritative voice on power management issues. Plant Analog EE TIMES MAGAZINE SUPPLEMENT, October 21, 2002 |
Let's think about How BASICs Work and porting BASIC52 or a variant to a 80C32 SOC equipped with a USB and DMA engines.
BASIC52 has floating point extension software.
No one in their right mind [retirement checks are simply great for freedom of expression] would install a floating point extension on a USB equipped 80C32 SOC.
Reason is that the Basic implementer would take two floating point numbers, package them in a USB packet, then send them to a gigahertz PC for processing. The PC would then return the answer.
On the PC side there are three places the floating point computations could be done.
1 In ring 0 wdm driver using either C or assembler. C would likely be best. At least at first.
2 In a ring 3 dll written in either C or Visual Basic.But the two numbers and the floating point operation code would have to be transferred, presumably though DeviceIoControl to ring 3 and the answer sent back from ring 3 to ring 0, then over USB to the 80C32 SOC.
3 The ring 3 dll could pass the two numbers and floating point operating code to a ring 3 Visual Basic 6.0 application which would do the floating point operation, then pass the result to the dll, which in turn would, pass the result to the ring 0 wdm driver, which would send the result to the 80C32 SOC via USB.
You might think that no one would be stupid enough to select option 3 when 1 is the obvious choice.
Wrong! But this is another sad story.
Csd preaches the importance of digitizing at the sensor. This was learned from magnetometer and seismometer builder experts.
Google, csd is convinced knows all.
Google picked up the Fun Stuff page which identified another screw up which likely resulted from not digitizing at the sensor.
About Dynamometer Numbers
The dynamometer (chassis or engine) can ... Friday August 30, 2002 08:25. Thanks for the wheel lecture yesterday. You saved me money and grief. You can do a headliner and rear panels yourself. ... www.geocities.com/computersystemsdocumentation/don/don1.htm - 88k - Cached -Similar pages
Cygnal is moving ahead with a 80C32 USB 1.1 SOC ... and hopefully not another development system blunder [No USB and expensive and antiquated Keil C cross development system].
But there is another Austin company working on a USB 2.0 SOC!!!
But this company might not have the good sense, as Cygnal did, to use a 80C32 SOC. But, on the other hand, they might, if alerted.
Let's hope that some of these companies sees the merits of intelligently porting FORTH or BASIC52 to their parts.
And if you want to see 80C32 FORTH run super fast, it's simple with USB.
Do -FIND and several other words on the PC instead of the 80C32!
Same is true for BASIC52. Wednesday October 23, 2002 07:39
Getting Forth and BASICs working on some of the 80C32 SOCs is becoming a top priority project.
Communication between the 80C32 and PC must be USB 1/2, of course, through a C++ wdm driver, connected to a C dll, which communicates with a Visual Basic 6.0 app! Tuesday October 22, 2002 12:58
Hey, we all make mistakes, but hopefully not essential material mistakes.
So learn, enjoy, and have fun doing operating systems ... which others have done much better in the past than any of us can likely do in the future.
So let's borrow and port the brilliant ideas of others. Monday October 21, 2002 20:19
The object is to not reinvent operating systems but to study existing ones with the thought of porting on or more to 80C32 USB SOCs.
Sandia National Labs paid to port FORTH from the 8085 to the 8051. The National Security Agency largely paid for that port.
At Sandia labs we studied operating systems before deciding on one for the 8085 family. FORTH won.
If any aspiring 8051 operating system designer and implementer thinks he or she can do a better job than Charles Moore of designing an interactive, memory efficient, super reliable, disk based operating system which hosts its own interactive incremental compiler and assembler, then they should think again.
The object at Sandia labs was get crypto hardware and software applications working fast and have fun while we were doing it.
Flores and Payne did only as much systems work as required to get Forth working quickly. And we hired FORTH experts Gary Fierbach, Ray Duncan and Jerry Boutelle to help us.
So perhaps it's best to port 8051 FORTH to one or more SOCs and get on to the job of writing super reliable apps without a hardware development system or the Keil C cross development system?
So appreciate what Loring Wirbel wrote in 1996 since it does relate to FORTH and Bill's other project. And to what's happening about Iraq.
80C32 SOCs with a USB engine on board are looking very promising. Other things are not looking so good.
Work has to be done in this slack business cycle to prepare for the next up time. Reading and understanding Manner's article is very important.
We have to get Forth and BASIC52 ported to some of the SOCs so that reliable peripheral software can be hosted by one or the other real time interactive operating systems.
And do other fun stuff too. Friday October 18, 2002 12:25
DATAPORT by
LORING WIRBEL For whom the bell tolls Two separate financial panels, one at the Communications Design Conference and one at the National Fiber Optic Engineers Conference, have convinced me we haven't touched bottom yet. Representatives from UBS Warburg and Morgenthaler warned of further OEM contraction in 2003, Jeremy Bunting of Thomas Weisel Partners bemoaned "negative enterprise value" of companies with stock selling below a dollar, and Greg Rossman of Pequot Ventures said research and in-network test-and-iteration cycles were losing a minimum of two years' shakeout time in this wretched downturn. On the one hand, it's dangerous to enter the "fear itself' mind-set that many out-of-work engineers around Silicon Valley seem to be immersed in these days. Yes, many people are in danger of losing their homes in early 2003, and that's no laughing matter. Still, unabashed pessimism could cause us to lose sight of several wireless and serial-interconnect markets where signs of rebound can be seen on the horizon. On the other hand, looking at the cover article on research-and-development trends in the September issue of IEEE Spectrum , I realize that many good companies have lost sight of Judy Estrin's famous warning to avoid eating the seed corn. Estrin's former employer Cisco Systems, is trying to keep R&D;funding at reasonable levels, as are JDS Uniphase, LSI Logic and Sun Microsystems. But the deep cuts in R&D at Lucent, Nortel, Agere and Broadcom represent a disturbing trend. Yes, these companies must cut burn rates on all fronts. But once the R&D and core technology platforms are sacrificed, can a company avoid a downward spiral? It is perhaps inevitable that hundreds of startups in both network equipment and semiconductor categories will go out of business during the fourth quarter of this year and the first half of next. Some large, longstanding and wonderful companies may head to oblivion along with them. Perhaps some of them won't be able to avoid the trap of trying to please shareholders by cutting further than basic, sustainable survival levels. But once the baseline of R&D and core technologies is slashed in the name of efficiency, it maybe impossible to cover. The bell is tolling for many of thee this fall. Electonics Engineering Times October 14, 2002 No man is an Iland, intire of it selfe; every man is a peece of the Continent, a part of the maine; if a Clod bee washed away by the Sea, Europe is the lesse; as well as if a Promontorie were, as well as if a Mannor of thy friends or of Mine owne were; any mans death diminishes me, because I am involved in Mankinde; And therefore never send to know for whom the bell tolls; it tolls for thee.
|
Cygnal must overcome several horrible blunders with its development kits.
Cygnal used RS-232 to connect its part through a second 8051 to the development board!
Cypress development systems, too, uses the RS-232 to communicate with the Keil development system when it has a USB 1 or 2 available. Ultimate stupidity?
Retirement checks, of course, means that you can more freely express yourself.
Cygnal development kits include an unnecessary transformer rather than use the power supplied over a USB cable!
Lastly Cygnal, and Cypress too, used the antiquated and expensive Keil cross C development system.
What is needed, of course, is a real time interactive disk based operating system hosting it own incremental compiler and assembler on the 80C32!
Forth or BASIC52 is needed.
Let's hope Cygnal gets it USB part out soon and one or more of the 80C32 SOC vendors see the merit of porting and existing operating system to their parts.
Cygnal is going to have the 5v to 3.3v regulator onboard its chip!
The 80C32 SOCs need to communicate over USB 1/2 to a Microsoft Visual C/C++ 6.0 wdm driver, communicating to a C dll, which communicates to Microsoft Visual Basic 6.0 terminal/keyboard/disk application. Thursday October 17, 2002 08:53
SUPPLIER CLOSE-UP
Cygnal tries to repeat history with MCU bid BY DARRELL DUNN Taking a page from a decade-old game plan executed successfully by Microchip Technology Inc., start-up Cygnal Integrated Products Inc. is trying to muscle its way into the ranks of 8-bit microcontroller suppliers by fusing le-end MCUs with high levels of analog integration. While the effort is winning Cygnal early recognition, some analysts say it is too early to draw comparisons with the world's No. 2 maker of 8-bit MCUs. Still, Derrell Coker, Cygnal's president and chief executive, Is unfixed and said his company Is committed to exploiting the same blueprint that propelled Microchip to early success. "We've been called Microchip on steroids," Coker said last week at the Silicon Hills Summit held here. "We've been able to go at this problem [of MCU and analog integration] with a clean sheet of paper and do it right the first time. I wouldn't say [other MCU vendors] don't have the talent, or don't have the wherewithal, but they have to work with their legacy products and legacy IP, and that makes it hard for them to do an effective job," he said. Founded in Austin, Texas, in 1999 by Coker, former Benchmark Mlcroeletronics Corp. excutive and with funding from Austin Ventures and others, Cygnal designs MCUs based on the industry venerable 8051 core. The company said its eventual claim to fame will be what it believes is unparalleled levels of on-chip analog and mixed-signal integration. Just as Microchip burst on the scene in the early 1990s with a new 8-bit MCU portfolio, Cygnal said it, too, expects to build Its business quickly. Last year, Cygnal had revenue of just $1.5 million, but Coker expects that to increase to about $6 million this year and to more than $50 million in 2004. Attacking the low end Similar to Microchip's early strategy, Cygnal is attacking the low end of the 8-bit MCU market and has been flooding potential customers with developer kits it said will help drive its revenue gains in coming years. Coker said the company has sold more than 7,000 kits, and has more than 1,000 customer designs in progress. The promise of integrating the 8051 core with data converters1 sensors, buses, flash memory, and other previously discrete components will earn the company additional design wins, he said. Some early engagements have included applications such as battery management, dense wavelength division multiplexing, headset displays, barcode readers, gas-leak detectors, and smart sensors. In addition to winning sockets from MCU suppliers like Microchip and Texas Instruments Inc., Coker said Cygnal is looking to steal the wind from analog suppliers such as Analog Devices, Linear Technology, and Maxim. Coker said he has identified more than $1 billion worth of potential business Cygnal could win that is currently being addressed by those five companies. "They do seem to have some products others can't match," said Tom Starnes, an analyst at Gartner Dataquest in Austin. "The ability to integrate analog with an MCU is something that's not easy to do, and it has given them a good head start. The question is whether they can keep it up, because you've got companies like Microchip and TI that are continuing their efforts in analog integration." Microchip rings in Ganesh Moorthy, vice president of Microchip's Advanced Microcontroller and Automotive Division, Chandler, Ariz., said the company has "recognized the need for incorporating cost-effective analog peripherals on-chip, and has been delivering PICmicro mixed-signal MCUs for many years. Cygnal addresses the needs of the 8051 installed base, similar to what many other semiconductor manufacturers have been trying to do." Coker said Microchip has gone full circle in its analog efforts, originally looking to provide integrated solutions, and more recently stepping back to provide a line of discrete analog components. He also said the low-end MCU market has been ignored by the start-up community over the past decade, with more attention placed on markets like DSP and 32-bit embedded processors. "This is a market that most companies decided to leave as status quo," he said. "But anyone who really knows this market, like Microchip, knows it continues to grow. We believe there are innovative things we can do within an established standard that other people aren't even thinking about." Cost problem ahead? Cygnal, a fabless supplier, uses Taiwan Semiconductor Manufacturing Co. Ltd. as its primary foundry partner. That represents a problematic distinction between the company and the manufacturing model employed by Microchip, Starns said, "Microchip in the early days was able to use Its older in-house fabs because they were good enough for this market," he said. "That enabled them keep their costs very low. Cygnal won't be able to do that. Their game plan makes sense, but saying you're going to be like Microchip in the 8-bit MCU space is a bit like all those [IP] licensing companies saying they are going to be like ARM or MIPS." Coker said Cygnal's opportunities will extend well beyond its initial foray into 8-bit MCUs, although no specific roadmap has been released. "We're looking at DSP, we're looking at other processor cores" he said. "We think we can repeat this same kind of concept on other base cores besides the 8051. But right now we believe there's a lot gas left in the 8051 platform." EBN October 14, 2002 |
Something very important below.
Forth is an operating system written in Forth and Forth assembler. And, if properly done, the source of the operating system is processed by a metacompiler which produces an executable binary.
When we started to look for an operating system at Sandia for an 8085, we asked Ray Duncan if we should decide to use Forth whether we should try to bring it up from assembler or use a metacompiler.
Duncan effectively told us we were fools if we tried to bring Forth up from assembler.
Eckert and Rowe make an inaccurate statement
Forth is a language
Forth is a disk based interactive operating system written in Forth and Forth assembler. And, if properly done, the source of the operating system is processed by a metacompiler which produces an executable binary.
Forth is one of the more memory efficient operating systems. BASICs being the other.
Also, applications written in Forth are on many orders of magnitude more reliable than others written in other technologies. Mainly C.
C is a memory hog. But who cares on a PC with its
infinite amount of RAM.
C stores local variable on the stack.
And, most important, C is not an operating
system.
You can write an operating system is C. But here's a problem. C is a memory hog. And this is a severe and probably fatal problem on an 80C32 where memory space is limited.
More astounding is that BASIC52 fits into 8k of internal memory.
There are better alternatives to the Keil C cross development system.
Think Forth or BASIC!!! Both are fun!!!! And you don't have to pay, and keep paying, for a Keil development system.
A full Forth operating system including compiler, linker, loader, incremental assembler, and full screen editor on the 80C32!!! takes about 18k bytes! Been there, done that.
However, we simply love Microsoft Visual C/C++ 6.0, with its exquisite inline Forth-like assembler, for windows dll and wdm driver. And love, even more, Visual BASIC 6.0 for apps!
And most important, Forth applications can be money makers. Little work to get the code running and super reliable in the field.
Here's bill's 1984 IEEE Software article on the merits of Forth. The statements are as true today as when they were written. And the technology may be more important in this technology bust cycle. Forth is essentially FREE.
And we are going to have even, perhaps, more fun porting 80C32 Forth to one or more SOCs as we are with ligitation and VWs!
Manner's article states
Will that skill be expertise in SOC?
SOCs with USB, csd is betting. And hopefully FORTH and BASIC will be available on some of the 80C52 SOCs. Monday October 7, 2002 18:20
Forth
by Brad Eckert and Don Rowe Forth is a niche programming language originally designed for real-time control of telescopes. As programmers from other fields discovered Forth, a grassroots effort emerged to mold it into an ANSI-standard language. Programmers whove used Forth describe the language as being like a room without walls. Some thrive on such freedom, while others are uncomfortable with it. Since Forth is a typeless language, the compiler can do little checking for you before you run your program. As a result, the most common failure scenario is a system crash. Forth is used mostly to test and debug hardware and bring up systems. Only about one in 50 embedded developers report using Forth regularly. Interestingly, some UNIX workstations boot a small Forth interpreter before the rest of the operating system. This environment provides some basic programming capabilities right out of ROM, and a small Forth bootloader stored there enables the operating system to be manually or automatically loaded from a disk drive or over a network and then run. The Forth estate Forth is a language with a simple syntax and many keywords. This is in contrast to Algol-style languages (such as Pascal and C/C++), which have a complex syntax and few keywords. If youre completely new to Forth, try to forget everything you know about programming languages as you read on. Forth programs are made of many small procedures. Forth is compiled, yet has no compiler in the traditional sense. Essentially, its a population of subroutines and an interpreter. The subroutines are called words. (In this article, words will appear in UPPER- CASE.) The dictionary is a data structure that associates the compiled words with their string names. The interpreter can invoke words that perform compilation actions, thereby extending the dictionary in the middle of a program. Figure 1 shows a flow chart of a Forth interpreter. The interpreter evaluates white space-delimited strings taken from an input stream, such as a console or file, usually in one pass. Word games You write a Forth program by defining new words, and run it by executing the top-level word. Forth manipulates data on a parameter stack that is separate from the call stack. (There are no registers.) Although static variables can be defined, words generally pop their parameters from the parameter stack, and push their results onto it. For example, the built-in word + pops the top two values, adds them, and pushes the sum back onto the stack. Bitwise AND operates similarly. The word < pops two values, compares them, and pushes the result (0 or -1). So a Forth programmer would code (2±3)*(4+5) as 23 + 45 + *, in reverse polish notation (RPN). The Forth standard specifies a boolean result as all 0s or all ls, which is 0 or -1 in twos complement arithmetic. This allows you to mix arithmetic and boolean operations, for example « 7 AND. The compiler allows any kind of type mixing, as Forth is typeless. With most data kept on the parameter stack, theres little need to track variable names or addresses, and temporary storage is automatic. Several built-in words manipulate the stack by rotating, removing, copying, or displaying items from various stack positions: SWAP swaps the top two items, DROP removes the top stack item, and OVER copies the second stack item to the top of the stack, thereby increasing the stack size by one. Words that manipulate character strings generally require a pointer as the second item on the stack and the string length on the top. Branching and looping words also use the stack. IF pops and tests the value on top of the stack. If the value is non-zero, the next word executes. Otherwise, control passes to the word following the ELSE, if present. BEGIN starts an indefinite loop and the corresponding END pops and tests the top stack value, looping back to BEGIN if its zero. DO/LOOP pairs repeat until an index (passed on the stack) increments to or beyond a limit (also passed on the stack). An important difference is that the index and limit are copied to the call stackto avoid cluttering the parameter stack. Listing 1 shows the definition and testing of a new word,
BIGGEST The word switches to compiler mode and begins the definition of the new
word. OVER OVER duplicates the values to be compared while maintaining their
order. The word < compares them, popping the two values copied by OVER OVER
and pushing the result of the comparison, which is subsequently popped and
tested by IF. If the comparison results in a-i, control passes to SWAP which
swaps the two values. Either way, the smaller value is now on top, ready to be
removed by DROP. The word ; terminates the definition and takes the interpreter
out of compile mode. Once defined, we can use a new word immediately. 5 9 BIGGEST pushes 5, then 9, then removes the smaller value. The word . prints the value on top of the stack (in this case 9), and the stack is again empty. The state of the stack as the program executes is shown below the code. Go Forth and prosper Forth has earned a reputation for being a write-only language. A typical Forth program defines and uses thousands of new words. In the absence of good naming conventions and comments, this can be a big maintenance headache. On the other hand, theres no reason Forth programs cant be useful and well documented. esp Brad Eckert holds a BS in physics from Shippensburg University. He has been a designer of both hardware and software for embedded systems for about 15 years. Brad wrote and maintains a free Forth-based framework for extensible firmware. His mail address is brad@tinyboot.com. Don Rowe is a consultant specializing in embedded controllers. He has over 25 years of experience with digital and analog design, software testing and reverse engineering Contact him at dws@canzonatech.com. Further Reading Embedded Systems Programming OCTOBER 2002 |
The Sharp literature arrived today.
Eleven of the twelve data sheet described ARM based SOCs. Only one data sheet described the 80C32 product.
HOWEVER, the 80C32 SOCs may outsell the ARM products, especially if they have USB 1-2 engines.
Where do you want to do serious computing?
On a software-starved peripheral SOC?
Or in an infinitely fast PC, with infinite RAM, and infinite disk space running Windows 2000 hosting Microsoft Visual Basic 6.0 and C/C++?
USB 2.0 will obviate need for a powerful peripheral processor is many applications. Besides customers want industrial data available to Microsoft office.
But one still needs a peripheral side micrconroller. And a super fast 80C32 with few cycles per instruction, dma and usb engines hopefully implementing both the Harvard and Princeton architecture [like the Cypress EZ-USB FX1 2 does] appears to be a winner.
Note that LZ87010 has 64k of FLASH! A FORTH or BASIC32 can be burned into flash. But only 4,096 bytes of external RAM. Darned.
Some company needs need to build a part with 64k of FLASH and 64K - 256 bytes of RAM with a fence register to mark the end of FLASH and the start of RAM. The 256 bytes should be allocated for peripheral decode.
Bill's other project continues to occupy much of his time but plans are being drawn on how to port existing operating systems to 80C32 SOCs. Tuesday October 1, 2002 09:37
csd got a nice email from David Lammers asking if Cygnal was going to make it.
Cygnal's first development kit flopped for the reason USB wasn't used, the on-board serial port was used, and the Keil C/assembler software development environment was used.
What is needed for these 80C52 SOCs is FORTH or BASIC which can be flashed on the SOC.
So let's all hope Cygnal corrects its development system board with release of its USB part this winter!
Sharp looks to have a neat product. Two cycles per instruction! Csd wonders if they implemented both the Harvard and Princeton architecture as Cypress wisely did. We'll find out!
There are peripheral USB 2.0 parts that can interface to a 80C52 SOC like Sharp. To be a success in the marketplace USB 1 or 2 should be used to connect the MCU to a PC.
Mode 0 80C52 synchronous communication expansion bus capability is to important to be wasted on asynchronous serial port development system communication!
Let's all hope that one of the 80C52 SOC companies sees the economic need to port 80c52 FORTH and BASIC52 to their product to speed and reduce the cost of software development. And have a more reliable application too. Thursday September 26, 2002 10:06
Sharp unveils 8-bit MCU to cut clock cycles to two CAMAS, WASH. An 8-bit, 8051- based microcontroller
from Sharp Microelectronics of the Americas can execute an instruction in two
clock cycles instead of the standard 12. Applications for the 40-MHz LZ87010
include motor control, test and measurement, instrumentation and
metering. Besides its enhanced 8051 core, the latest member of Sharps BlueStreak family of MCUs and systems-on-chip packs 64 kbytes of flash and 4.25 kbytes of SRAM. The flash can be programmed serially or in parallel, or can be self-programmed during run-time. Other features include eight channels for l2-bit A/D conversion, two independent D/A channels with 128-byte waveform generator RAMs and 16 high-current outputs. The device offers six 16-bit timer/counters with input capture, output compare and pulse-width modulation capability, plus a watchdog timer and real-time clock. The LZ87010 also has two UARTs and an 12C serial interface. Packaged in a 100-pin LQFP and priced at $8 in lots of 25,000, the LZ87010 Blue- Streak MCU is shipping to beta customers and will be available in production quantities in the first quarter of 2003. Call (800) 642-0261 Electronic Engineer Times September 23, 2002 |
Software costs for these parts can be reduced by getting a BASIC and FORTH operating system working on them.
Also experience has shown that the FORTH technology is many time more reliable than other technologies such as C and assembler. Tuesday September 0, 2002 07:33
The System on a Chip
Your company plans to launch a new line of ultrasonic sensors. The design calls for significant improvements over a previous design, which is based on an 8051 microprocessor and discrete components. You have a total of four models to develop: externally programmable, discrete output; externally programmable RS-232 output; fixed internal program, discrete output; fixed internal program, RS-232 output. Traditionally, this would require four new designs using discrete components, with four individual bills of materials, schematics, and PCB layouts. If you were going to sell 1 million units of each, you might be able to justify the engineering and tooling cost for each of the four designs. But what if the anticipated sales volume is only 1000 pieces per year for each unit? Suddenly the success of the entire product line depends on your ability to modify a single design and capitalize on every possible design efficiency. The issues here are universal throughout the electronics industry, and not just with low-volume products. The ever-decreasing cost of computers and the demand for more power means that more and more functions are taken off the PCB and put into the Pentium processor (e.g., power management, disk controllers, and cache). ICs that were previously external are now internal. You need a similar advantage in your ultrasonic sensor design. If there were a microprocessor that had a serial port, digital I/O pins, and an external programming interface (e.g., an 12C or USB port), you could use the processor for all four designsand probably the same PCB as well. The only difference might be the external connections that you make to the board (read: money saved on manufacturing, engineering, and tooling). Remember your old 8051 code? If the new processor were available with 8051 support, you could reuse and update your existing code. So the economies of software development are similarly improved. The good news is that you really can do this. The system on a chip (SoC) business is a segment of the semiconductor industry that solves these kinds of problems. Roll Your Own? In the past, high-volume manufacturers would design application- specific integrated circuits (ASICs) for themselves. Of course, some people saw an opportunity to develop quasi-custom ASIC solutions, buying intellectual property from chip developers, mixing/matching components, and offering them for sale to developers as off-the-shelf products. Only a short time ago, building a completely customized single- chip solution for an embedded application was a risky and expensive process. If there were a missed feature or a misjudgment of market requirements, it was more than just a headache. More than likely, the next project for the product development team would be new resumes. But what if you could customize a chip solution for yourself, in low volume? Configurable SoCs First, consider field programmable gate arrays (FPGAs) and programmable logic devices (PLDs), which are at the extreme end of configurable processors. While the nuances of these two technologies are beyond the scope of this article, pure FPCAS and PLDs are essentially blank slates. You can burn in custom logic and effectively simulate the operation of any active device you wish to create. These devices are used for implementing customized low-volume chip solutions, in somewhat the same way you might burn software on a CD-ROM and sell it in low volume instead of pressing CDs mass production style. Many SoC products combine multiple peripherals with a certain amount of blank-slate configurability, usually in the form of libraries predefined by the manufacturer. This allows a single piece of hardware to be programmed with a diverse array of turnkey feature options, speeding development and allowing extensive customization. Advantage over ASICs SoCs configurability is a huge benefit. The process of designing ASICs always required an incredible amount of foresight into the customers requirements and the target market. Re-spinning silicon after rollout meant lost market share and financial doom. And the process has not improved. As chip densities have soared, ASIC development has become even more cumbersome and expensive and is a greater gamble for the developer. Only the largest companies can afford the time, development costs, and risk for anything but the highest-volume applications. Silicon Intellectual Property As custom chip solutions have evolved, a market for chip dies has also grown. SoC and FPGA technologies have in turn given ASIC developers the ability to roll out designs faster and in lower volumes. So if theres a discrete component youd like included in your ASIC, you can probably buy the intellectual property on a royalty basis. This has increased the options available to SoC and FPCA customers. Lucky for product developers, configurable processor architectures now provide the flexibility, performance, low cost, fast time to market, and product differentiation needed by embedded solutions. Several types of solutions are available, including soft instruction, configurable, and combo processors (which implement aspects of both soft instruction processors and configurable processors). Soft Instruction Processors Soft instruction processors let you customize CPU architectures by specifying instructions supported, peripherals available, and number of registers. Some vendors provide mechanisms to add, delete, and create highly tailored instructions. Design packages for these architectures sometimes include performance tools with instant feedback on the performance, die size, and power requirements of a design. With the final architecture residing in silicon, these types of architectures are well suited for high-volume, low-cost applications, which formerly would have used ASICs. Configurable Processors Configurable processors, also known as souped-up 8032 MCU platform in its ES family. The ES marries standard 8051 MCU technology to configurable system logic (CSL), which can be thought of as a matrix of gate arrays. By combining the CSL with a library of soft moduleswhich include serial communications, logic functions, and display driversTriscend delivers an easy-to-use, customizable solution for configurable SoC applications. It brings simple 8-bit architecture together with sophisticated communications, three 16-bit timer/counters, a watchdog timer, a full-duplex UART, and three external and 10 internal interrupts. Simple Software Configuration Implementation of CSL is hidden from the designer. Using Triscends point-and-click Fast Chip Design Studio, logic modules can be dragged onto the chip as needed. After you select modules from the Fast Chip Library (see Figure 2), Fast Chip provides you with interface code files that you include in your project. These files connect the selected peripherals to the application code. No need to spend hours managing cell addresses or peripheral implementation details while fighting tight development deadlines. Why Pick an SoC? Any of the following reasons could be justification for considering an SoC for your next product: Size constraints that dictate reduced parts count Editors Note John Rinaldis white paper, 5 Issues to Consider Before Designing With a System-on-Chip, is available free on request. Simply e-mail soc@rtaautomation.com or call 414-453-5100 with your street address. Also ask about his e-book, Industrial Ethernet in 90 Days or Less: A Plan for Product Developers. John S. Rinaldi is President, Real Time Automation, 2825 N. Mayfair Rd., Ste. 11, Wauwatosa, WI 53222; 414-453-5100, fax 414-453-5125, jsr@rtaautomation.com. Sensors September 2002 |
Cygnal should have its USB part out by mid year. USB 1 not 2.
Let's hope Cygnal or one of the other 80C52 sees the merit of getting a BASIC and FORTH operating system working on their parts as an alternative to the Keil development system. Monday September 9, 2002 20:52
AUSTIN SKYLINE by DAVID
LAMMERS
Cygnal: Proof it can be done Not too long ago, I thought of an idea for a decent story: Go around to the hot Austin startups and find out how things are going. Trouble is, the A company on my list, Alchemy Semiconductor, was snatched up by Advanced Micro Devices at a fire sale price. Banderacom Inc. was next on the list, but its target marketInfiniband has sputtered like one of those firecrackers that you have to relight a couple of times at the neighborhood driveway gathering on the Fourth of July. After B comes C, and Austin does have a startup success in Cygnal Integrated Products, a company that shows that, even in todays crummy business environment, a startup can gain traction. Cygnal was founded in 1999 around the idea of taking a fast 8051 processor core and surrounding it with mixed-signal circuits that would rival what the top tier analog companies provide. Flash cores from foundry TSMC are the third leg of Cygnals product strategy. The 64-person company has created 36 variants on its central theme. Some of Cygnal's system-on-chip deices are flyspeck-sized: a 25-Mips processor, an A/D converter, SRAM, a temperature sensor, parallel and serial ports, all fit on a chip measuring 3 mm^2 While some companies, like Banderacom, are vulnerable because they need to gain traction in one application, or with a few major customers, Cygnal has more than 1,000 customers, ranging across the whole gamut of electronics systems. How does a small company with only six field application engineers attract, and keep track of, so many customers? The short answer: the Internet, which marketing vice president Don Alfano calls the engineers best friend. Customers can download Cygnals own development tools, order parts and so on, online; Nothing too unusual there. Cygnal also has an internal customer- tracking tool that allows the sales reps to funnel customer information back to Cygnal over the Web. This self-developed knowledge management system gives Cygnal the ability to sort customers by volume, provides order fulfillment data and drives new-product design strategies. Cygnal remains private, and it is hard to tell just how well it is doing. The company recently raised $20.5 million, and revenue is said to be growing each quarter. Still, customers are a lot more cautious now than they would be in a normal business environment. The overall business sentiment is not as bad as last year, but we are still not in a steady-state economy, Alfano said. Send feedback to dlammers@cmp.com. Electronic Engineering Times September 2, 2002 |
The 8051 is hardware extensible in that additional Special Function Registers can be defined by users.
What the electronics worlds needs is a super-fast 80C52 with USB and DMA support. It does not need a new 32 or 64 bits microcontroller for the reason that data transfer with USB 2.0 must be done in hardware NOT SOFTWARE.
Note that Goal advertises "single cycle" as opposed to about 4 cycles used by other 8051 vendors. Thursday August 28, 2002 10:19
8051 RISC Processor Features DSP Capabilties
Equipped with a single-cycle, 8051 RISC processor, the VERSA MIX VMX 1020 microcontroller is said to deliver an average of eight times more processing power than a standard 8051 MCU. The device integrates a hardware MAC unit coupled with a 32-bit barrel shifter capable of implementing both 16- and 32-bit DSP operations. The device also has on-chip analog peripherals such as a 4-channel a/d converter, four PWM- based d/a converters, a voltage reference and an uncommitted op amp.
Other features include a 56 KB block of programmable flash, an 12C interface, two UARTs, a differential RS-485/RS-422 transceiver, and 1.28 KB of RAM. The SPI can be configured as either a master or slave interface and up to four devices can be addressed in master mode via four programmable chip enables. In addition, the device features a programmable current source that is calibrated to deliver 33 pA or 133 pA and comes in 64-, 52- and 44-pin QFPs. For price, call GOAL SEMICONDUCTOR, Quebec, Canada. (831) 429-6288. EE Product News August 2002 |
Texas Instrument selected the 80C32 for its TI/Burr-Brown delta sigma MSC1210 converter.
Now TI is using fast 80C51 [80C52s] for the microcontroller for its USB products!
What is not needed is a 32 or 64 bit microcontroller. USB is too fast for software.
What is needed is fast USB and DMA hardware engines. And, of course, a BASIC and Forth operating systems with incremental compilation, assembly, and interactive development environment on the 80C52!
This is opposed to the Keil C [which is not an operating system] cross development system. Both Cypress and Cygnal use the antiquated Keil development approach.
TI set trends. TI president Tom Engibous recently explained TI acquisition of Burr-Brown for the astronomical price of $6.1 billion!
Engibous apparently told the shareholders to trust him that he was doing the right thing.
System on a chip [SOC] technology looks to be the winner when the electronics industry pulls out of it current slump. Using 80C32s, of course.
David Manner's article explains what will happen to those companies who missed going with the fast 80C52s as their core. Monday August 19, 2002 10:53
The green box at the low far right says 80C51 (60Mhz) |
Iran and Iraq, of course, is on Bill's mind. Friday August 16, 2002 19:45
Engineering Ethics
Jack G. Ganssle In 1985, 1 sold six in-circuit emulators to the government of Iraq. Today, Iraq is part of the Axis of Evil Then, they were our allies, partners in the U.S. Mideast strategy. I suppose our government still had some reservations about sending high-tech equipment there, since getting a valid export license took six months and several trips to the Department of Commerce in Washington, DC. But the approvals arrived and we shipped the gear. Iraq paid on time. We never heard a word from them again. Where were these things used? Were they part of a humanitarian medical program? Were they used in the development of weapons of mass destruction? At the time, Iraq was engaged in a mortal struggle with Iran. Iran was our enemy, Iraq our partner in the region. After the 1990 invasion of Kuwait, two burly men unexpectedly showed up at the office, sporting suit coat bulges in the wrong places and business cards pretending employment by the Customs Department. The phone number was one I recognized from many years spent in the intelligence business, one that led to Langley, not DC. These gorillas were mostly interested in all of the paperwork associated with the years-old sale. Every document related to the export licenses was duplicated and questioned; the products technical capabilities and features went unexamined. We were squeaky clean and quickly satisfied their inquiries. Fast forward to 2002. Everything has changed. Our relations with Iran are warming, those with Iraq headed from bad to worse. Today it might be possible to send equipment to Iran, but surely Iraq is a no-ship zone. treaties, members of various mutual defense pacts, whose democracies neither oppress their people nor antagonize the world. Yet one cannot ignore the shifting alliances of Realpolitik. My parents were part of a generation where even some of these countries were quite at odds with America and others. The enemy is a nebulous character whose identity varies with the times. Can we separate our engineering efforts from politics? In my naive youth, we claimed that everything is political, that even personal relationships impact the world at large. Those nearly-Marxian dialectics were more the product of sophomoric philosophies and frustrations with the times. But now I cannot help but wonder if, indeed, even the actions taken by us little people have some grave consequences of international scale. Its perhaps extravagant to suggest that Though our paperwork satisfied the government inspectors, I was left in a quandary. What was right? Was being legal the same as being ethical? Should we have sold these products to such an unhappy part of the world? Like most tool vendors, we had distributors in all of the major technology- consuming countries of the world, particularly in Western Europe and Japan. These are safe counties, bound by the sale of a handful of emulators for 8-bit processors may have contributed to significant evil. But did my sale of these productsthis decision made by inconsequential me and my tiny companyincrease the amount of evil in the world? Its too easy to suggest, as the ethical alternative, that we ship high-tech products only to the most benign of countries. These tools could be used to build products that pump clean water into villages in Pakistan, generate power for a community in Tibet, clean up pollution in Romania. How do we make these decisions? Searching ones own soul, especially one much more grounded in ones and zeroes than international politics, doesnt always provide the right answer. The law of unintended consequences means we cannot understand the implications of our actions over the long haul. It can twist our very best efforts into the worst of results. Yet, in my opinion, we cannot abdicate our responsibility to strive for the maximum amount of good. We engineers are the architects of the new order, and as such must consider how our work might spawn good or evil. My experience with the Iraqi sale taught me that its not enough to simply trust the government to decide. Ethics are personal. Talking shop These thoughts came to mind after facilitating a Shop Talk discussion on Engineering ethics in June at Chicagos Embedded System Coference. I raised the big issues of products and politics, using the crisis in the Kashmir as a focal point. What does that mean to those doing business with either India or Pakistan? Does arming the belligerents constitute a violation of our moral responsibilities? I was looking not for an answer, but for discussion and thoughtful insights. The group wasnt interested. Perhaps those locales are too remote; maybe Im too far off into metaphysics for nuts-and- bolts engineers. Maybe people feel powerless and unable to pursue change on regional scales. To my frustration, several people suggested that we should simply do what the boss wants. Im sure many people would react appropriately to an outrageous order. But big evils grow from small decisions. The Nuremberg trials brought monsters to the dock. But what of the little people, the clerks who stamped the papers ordering a family onto the train to Dachau? If those people had even a glimmer of the consequences of their actions, they were active participants in an unimaginable horror. Yet they were doing what the boss ordered. Small cogs in a huge machine who were perhaps unable to effect change. Does that excuse their actions? Does it excuse ours when we follow the companys marching orders even when were uncomfortable with the possible outcomes? I asked the group about other ethical issues. A letter in the RISKS digest suggested that ignoring buffer overflow problems is far more than bad engineering. Maybe its criminal, he suggested, especially when repeated attacks against this oh-so-preventable bug cost customers big bucks. Its trivial to write code thats immune to a buffer overflow attack, yet so many of Microsofts products, for instance, have been repeatedly compromised. Dont Bill Gates and his company have an ethical and legal responsibility to their user community to fix at least this simple vulnerability? Most of the attendees agreed, but its easy to blame others, and easier to target Microsoft. All ethics are personal The ethical life is not that of a moral dilettante; its one that encompasses sometimes heavy burdens. I was struck by comments made by the commencement speaker at a high school graduation this week. He said ethical behavior means understanding the difference between right and wrong, and then accepting both the responsibility and the accountability to do the right thing. Wowwe have to seek out accountability for our actions. Too many of us avoid that at all costs. The crowd was full of happy parents and relatives, each cheering on their own graduate. Many parents had tears in their eyes; quite a few were single moms who had raised these kids more or less alone. Where was dad? Divorce does not mean abandoning ones children. Dads should be held accountable for doing the right thing. These are difficult times. Kennedys ask not what your country can do for you admonition seems a quaint reminder of a kinder age. This feels like the gimmee and get-rich-quick era. The headlines speak of corruption and dissolution. Enron, Anderson, Tyco, Merrill-Lynch, and too many other corporate names splayed across the front pages this summer suggest that corporate America is the realm of sleaze, that CEOs will do anything, legal or otherwise, ethical or not, to inflate stock prices and build personal wealth. Its up to each of us, individually, to effect change. Ignore the headlines. Act ethically. Deal with the agony of divorce but take care of the kids. Ones own perceived needs pale compared with the responsibilities procreation incurs. Our responsibilities transcend the rules set down by the courts or the legislature. This week I met an electrician at a local boatyard, a guy who volunteered many free extracurricular hours to help rewire a friends mast. George is one of the nicest, friendliest, and funny fellows at that yard. Later I learned he had manipulated his companys health care system to wangle eight months of disability pay for what was the most minimal of injuries. Though not at all an uncommon practice in todays scheming society, this is deceitful and unacceptable to me. At work and at home we must adhere to the highest of standards. Even when its hard. The Brooklyn Bridge is one of Americas icons. The first large suspension bridge, it incorporated many new construction ideas, including the massive use of structural wire rope. Roebling designed and constructed the bridge, yet his company, the best wire rope vendor of the day, lost the wire contract to another firm. That firm provided, knowingly and with almost criminal intent, substandard material that could have jeopardized the safety of thousands of commuters. Only Roeblings sensationally redundant design saved the project. The history of civil engineering is filled with stories of crooked contractors and lousy materials. In pursuit of the quick buck, ethics are consistently tossed to the winds. We developers have a responsibility to our customers that closely parallels Roeblings corrupt wire rope vendor. Quality is hidden deep inside the product. No one sees the guts of our creations. A simple user interface might conceal hundreds of thousands of lines of code. Is it beautifully structured or a spaghetti mess? A couple of the Shop Talk attendees voiced what we all know but seldom admit: were lazy. Its easier to hack stuff out than do it right. Disciplined development is a core value of any workable approach to reliable firmware, but its tedious and, well, disciplined. Banging away on the debugger making motors whir and lights flash is a lot more fun than sitting in front of a desk thinking, especially when the cubicle is so noisy that deep thought is impossible. Im struck by the correlation between beautiful and reliable code. The Therac25, the earliest known embedded disaster, killed several people. Proximate causes included bad design and completely unstructured, unaudited, and totally convoluted code. A British helicopter accident resulted mostly from firmware so awful the review hoard gave up trying to read it after getting through just 17%. Check out the PC/OS RTOS (www.ucos-ii.com). Read the C listings. Then check out the source code to Windows CE (www. microsoft. corn/windows/embedded/ce/tools/source/default.asp.) One is beautiful, written almost like poetry, with pride and discipline. The other looks like the mutterings of an insane software terrorist. One is safety- certified to DO-178B standards. The other, well, lets just say its great in easily rebooted hand-held appliances. The beauty of the great code lies deep in its innards, invisible to any consumer. Its elegance cannot be observed functionally. Youll never hear a customer say, Hey, this thing hasnt ever crashed! Yet the beauty that stems from making difficult and ethical development choices yields great, reliable, portable code. its a standard we must all hold ourselves to. Think globally, act locally Act now. Do the right thing in your daily engineering efforts. Dont wait for others to take the lead. Do you check for buffer overruns? Vast bodies of experience show that input strings from untrusted sources crash code. If you skip these trivial checks, youre writing code that increases the unhappiness in the universe. If you use malloc(), do you check its return value? We all know that heap fragmentation can lead to malloc failures, so Id argue that writing code that assumes success is more than poor design; its unethical. Its dumping problems that we should deal with ourselves onto our users. Lots of us feel cranking code is a lot more fun than detailed design, code inspections, and adherence to standards. Yet I contend its worse than lazy to jump into coding. It leads to lousy products, frustrated users, and is a fundamentally unethical way to build a product. Is it ethical to accept an arbitrary, capricious, and impossible delivery date from the boss? Thats engaging in a dysfunctional cycle of lying, one thats doomed to get worse. Ethical behavior means accepting responsibility for your actions. As Harry S Truman said, The buck stops here. On your desk. Not your bosss. esp Jack G. Ganssle is a lecturer and consultant on embedded development issues. He conducts seminars on embedded systems and helps companies with their embedded challenges. Contact him at jack@ganssle.com. http://www.embedded.com august 2002 |
As we all should know
Those who cannot remember the past are condemned to repeat it.
George Santyana, 1905
Now is a good time to prepare for the next upturn in business.
Fix known mistakes.
Upgrade to new technologies like 80C52 SOCs and USB 2.0. Monday July 22, 2002 09:39
SEMICONDUCTORS
In the Throes of Death BY DAVID MANNERS Soon, the Silicon Cycle will be claiming its victims. Semiconductor companies die, not in the depths of the down-cycle, but early in the up-cycle when it becomes clear who has the right new products and the capacity to make them. Who now remembers Clevite, Amelco, Transitron, Philco, Sylvania, Tung-Sol et al? Sic transit gloria transistor. The transistor knocked out the vacuum tube manufacturers; chips overtook discrete transistors; MOS overwhelmed bipolar; CMOS swamped NMOS; ASICs took over from TTL; and programmable logic is taking over from hardwired ASICs. And so on, and so forth. In 1955, three years into the commercial manufacturing of transistors, only four of the top 10 vacuum tube manufacturersGeneral Electric, Sylvania, RCA and Westinghousemade the technical transition to transistor, making them sufficiently successful to be in the top 10 transistor makers. New companies, setting up as transistor makers without ever having been vacuum tube manufacturers, had done a better job. A decade later, in 1965, when the chip market was getting going, Sylvania, Hughes, Westinghouse and Clevite were elbowed out of the top 10 semiconductor manufacturers by pushy newcomers. Much of it comes down to people because only people can both correctly catch the technological drift and have the rare skills to get new technologies to work. For instance, largely through its invention of the planar process, Fairchild, led by a team of the brightest young scientists in the United States, became a top 10 semiconductor company eight years from startup. But when the Fairchild founders left to pursue MOS, Fairchild failed to make the transition from bipolar to MOS and declined to the point where it was sold to a French oil-field services company. Sparky newcomers such as Mostek and Intel were the best at making MOS and soon became the industry leaders of the 1970s. When Ferranti single-handedly invented the semicustom industry by coming up with the idea of the gate array, it dominated the world market for gate arrays for a decade. Although it is only recently that the gate array market has started to decline, Ferranti was out of the game more than a decade ago because it had missed two fundamental changes in the way in which gate arrays were made: the move to automated design and the move from bipolar manufacturing processes to MOS. Ferranti's original big idea was still valid, but, by failing to see that the rules of the game had changed, Ferranti allowed others such as LSI Logicto usurp its idea. LSI set up to make MOS gate arrays using powerful EDA tools running on mainframe computers, and it took over Ferranti's position as the No.1 merchant market manufacturer of gate arrays. However the gate array market met its Waterloo in the shape of a better big idea called user-programmable logic. Two companies became dominant in the new technologyAltera and Xilinx and the measure of the success with which they rode their big idea was that they saw off an astonishing number of major companies which tried to muscle in on it: Intel, TI, National, AMD, Philips, GEC-Plessey, AT&T Microelectronics (later Lucent Microelectronics and later still Agere Systems), Motorola and Toshiba. While old technologies hang around for many years, it is the new ones which decide leadership, and it is the downturns in the Silicon Cycle which accelerate the process of determining the new winners and the new losers. The downturns of 1971, 1975,1982, 1985, 1990 and 1996 all exacted their toll. The year 1975 saw Sprague, General Electric, Transitron, Raytheon and Philco pushed out of the top 10 by newcomers National Semiconductor, Intel, Rockwell, Signetics and AMI. By 1985, AMI, General Instruments and Rockwell had been replaced by Harris, Mostek and AMD, but the devastating rise of the Japanese based on superior manufacturing technology was about to revolutionize the industry's leadership. In 1988 Fairchild, RCA, General Instrument, AMI, Rockwell, AMD, National Semiconductorall from the United States were ejected from the top 10 and were replaced by six Japanese companies: NEC, Hitachi, Toshiba, Fujitsu, Mitsubishi and Matsushita. Japan's dominance was based on success in MOS memory and, when the arrival of the Koreans commoditized that product sector, the Japanese companies could not diversify. By the late 1990s, the top 10 consisted of TI, Motorola and Intel of the United States; NEC, Toshiba and Hitachi of Japan; Philips, Infineon and STMicroelectronics of Europe; and Samsung of Korea. And that's pretty much how it is today. But history says it won't stay that way for long. The 2001 downturn will accelerate changes among the leadership based on skills in new technologies. Will that skill be expertise in SOC? Or in new types of processors? Will it be supremacy in a totally new type of product such as a one-for-all purpose memory via the magneto-resistive MRAM route or the ovonics unified memory route? Or will it be something no one's yet thought of? Only time will tell. David Manners is an editor for Electronics Weekly, a sister publication of Electronic News. Electronic News June 24, 2002 |
Lots of the SOCs will be 80C52s.
But
will one be able to get sufficient parts for a project?
Making money requires that the design and development stage terminates to transition to a build and sell phase.
If chips are changing constantly, then those who use these chips can't make money. There is a very serious problem with this in the electronics industry.
Also obsolete chips turn into ewaste.
Bill's other project is looking more promising than the electronics industry. Sunday June 9, 2002 06:11
Semiconductor firms face lean future, says
analyst
By Ron Wilson SANTA CLARA, CALIF. As if a lousy business climate wasn't bad enough, semiconductor makers have more to worry about. "The simple fact is that we won't be needing so many chip vendors in the future," Dataquest analyst Jim Tully said last week. "Many firms that are now fabless chip houses should seriously consider becoming IP [intellectual-property] vendors instead." As for that next "killer app" for which the industry is holding its breath, Tully says forget about it. "We are always asking, 'What will be the next killer application?' "he said. "The truth is there will be no more killer apps. There will just be a growing network of hardware resources onto which increasingly compelling software applications can be loaded." At the Fabless Semiconductor Association Technical Workshop in Santa Clara last week, Tully based his less-than-reassuring outlook on a model of the end markets for electronic equipment, the needs of those markets and the rapidly changing technical environment in which all IC vendorsfabless or notfind themselves.' "We used to live in an industry that was driven by the personal computer," Tully observed. "We saw that change into an industry driven by the dot-corn explosion. Now we are in the post-PC, post- dot-coin era." Today, Tully said, systems- on-chip are consumed mainly by wireless handsets and consumer entertainment devices, with networking and storage applications a distant third. That the sockets are essentially in consumer markets, he said, will cause profound changes. "Consumer markets are fast and fickle," Tully warned. Technical issues also are changing. The increasing integration of SoC products means that fewer chips are going into a system. And the spiraling cost of a design both in nonrecurring expenses such as mask costs and in design complexity and riskare putting SoC design beyond the reach of more and more companies. This has led, Tully claimed, to a decrease in chip design starts. "Starts peaked in 1997, and we predict that the decline will continue indefinitely." Instead of SoC technology spreading across the industry, Tully painted an industry restructuringand contracting around a handful of vendors that could still afford to complete SoC designs. Many of today's fabless companies, he suggested, will continue to serve the industry, but as chipless IP providers rather than as semiconductor vendors. Those that continue to produce silicon will increasingly design standardized platform chips, capable of serving a range of applications via reconfigurable hardware and software, thus spreading the multimillion-dollar cost of a design across a range of applications. Services over hardware This reconfigurabiity also played a second role in Tully's view of the future. Increasingly, he said, the consumer markets that take SoC devices will be dominated by services, not hardware sales. New content, features and functions, would make up a larger portion of the revenue than the hardware, until the SoC-based system simply becomes a delivery vehicle for revenue-producing services. Oems, which now own most of the teeth in the food chain, will be squeezed between IC platform vendors that demand better margins and content and service providers who demand their pound as well. This scenario puts a premium on the flexibility and extensibility of the hardware platform. And it discourages system architects from locking differential advantages into hardware. Hence, the industry will swing away from its tradition of starting a new SoC design for each new application, instead adapting platform chips to cover new opportunities. The analyst suggested that the ideal SoC of the future would be a computing and networking platform, capable of adapting to a range of specific applications through reconfigurability. It could, on one side, configure itself for any of a wide range of communications environments and protocols, and on the other side would have sufficient headroom for a long line of enhanced user features. Though not a view of the future that beleaguered semiconductor executives were eager to hear, Tally's prognosis correlated with their anecdotal experience. One manager at a wireless SoC startup who wished not to be identified spoke of the difficulty of getting B-round funding. "Last year a lot of companies were able to get a first round. But now they all have designs that are just about done, and they need more money to tape out and produce chips. The money just isn't there. "So there's a growing pool of companies alinost ready to tape out, but trying to cut spending while they find enough customers to get another round," the manager related. "It's particularly bad in the network processor area, where there are whole designs that may never tape out." Nonetheless, Tally's analysis left open some important questions. First, is the trend toward fewer design starts and fewer tapeouts indeed a structural phenomenon, or is it an artifact of the present popped-balloon market? Once another idea shows apparent growth potential, will investors rush to fund hundreds of indistinguishable startups once again, reigniting growth in SoC design starts? A second question involves market life cycles. The scenario Tully describes is well-known for maturing markets. Growth slows, reducing investment. Differentiation gets harder and produces less revenue growth. Standardized computing plat- forms become capable of supporting most of the features that add market value, so the importance of vendor-specific hardware declines. And there's a shakeout. Is Tully simply documenting the maturity of the existing SoC markets, then? Or will another set of consumer appetites appear that once again justifies the rush to build specialized hardware? If the change is indeed structural, there is in fact a train coming down the tunnel. But if there is a next big thing out there that lies beyond the reach of existing platform architectureseven in advanced processesthat glimmer in the dark might be daylight after all. Electronic Engineering Times June 3, 2002 |
Communications errors occur at the start and stop of rs-485 and mil std 1553 frames. The lines seem to take a set when they've been idle.
USB 2.0 may have similar problems. Or even noise problems.
Bill has some practical experience doing EMC TEMPEST work. Even using Murata-Erie parts.
One should use the ideas of experts .... like Murata.
Do your USB cables conform to what Murata suggests? If not there may be practical trouble ahead for you ... or us! Thursday June 6, 2002 20:05
End USB Noise Without Stopping the Bus
by Jonathan Davis, Gerry Hubers, Deryl Kimbro, and Shuji Mikami, Murata Electronics North America
As universal serial bus (USB) 1.1 becomes stable ground for developers, the introduction of USB 2.0 makes the ground uncertain again. The new specification calls for operation up to 40 times faster than USB 1.1, resulting in speeds of 480 Mb/s. With the major goal of USB still being to replace the traditional ports of the PC, it now has the speed to guarantee that it will be around for a while. The USB primarily is a PC bus that can be readily applied to any device that would use an RS-232 or parallel port. Just about every PC on the market supports at least one USB port. USB is more flexible and more internally complicated than the interfaces that it replaces. USB 1.1 supported two bus speeds: 1.5 Mb/s and 12 Mb/s. Low-speed applications are interactive devices such as a keyboard, a mouse, and other low-end peripherals. Full speed is designed more for phone, audio, and compressed video. Now with USB 2.0 high-speed performance, bandwidth-hungry applications such as video and storage devices are easily connected to a PC. USB 2.0 can support higher speed applications such as high-density storage devices, high- resolution scanners, and printers. As with 1.1, the host still is in charge, allowing peripherals to be simpler and cost effective. Table 1 describes the relationship for the range of data traffic workloads that USB can handle. With the increase in speed comes the growing concern for EMI, where high- speed signal lines are notorious for generating noise. USB 1.1 and 2.0 use differential signaling because it is less prone to generating noise. Ideally, the magnetic flux produced by the differential signal currents has a cancellation effect and, for that reason, is not a large contributor of EMI noise. However, harmonic noise from IC drivers or peripheral circuits can radiate onto the signal lines and be transmitted through the USB cable, which can act like an antenna. Signal coupling by parasitic inductance or capacitance can cause common-mode currents. The common-mode currents are the dominant source of EMI problems for high-speed interfaces. As shown in Figure 1, all common- mode current flows in the same direction so the magnetic flux is not canceled as with differentialmode currents. USB 2.0 presents new challenges to the design community because of its higher frequency band. The typical USB 1.1 EMI filtering solution consisted of ferrite beads on the power and ground along with a ferrite bead on each signal line to reduce the level of noise. In USB 1.1, the signal frequency band and the noise frequency band are far enough apart that the ferrite bead can reduce the noise and not affect the signal. The higher speed of USB 2.0 makes this solution ineffective because the signal and noise frequency bands ire close together or possibly occupy the same frequency band. To use the same ferrite-bead filtering solution for USB 2.0 would suppress both the noise and signal frequencies. An effective way to provide the required noise attenuation includes a common-mode choke coil that reduces the noise while maintaining the integrity of the signal. As displayed in Figure 1, the magnetic flux created by the common-mode current increases and produces an impedance while the magnetic flux from the differential-mode current cancels, resulting in no impedance. The common-mode choke coil must have two key features to make it an effective solution: a high common- mode impedance and a high magnetic coupling factor. The effective common-mode impedance provides a higher level of common-mode noise attenuation. The test setup shown in Figure 2 was constructed to evaluate the effect of the common-mode choke coil. Testing was performed in an anechoic chamber with the USB Controllers separated by a distance of 2 meters. The noise radiation was measured using the standard 3-meter test method. Initial readings were measured and recorded.
A 9O Ohm common-mode choke coil (Murata part No. DLW21SN900SQ2) was tested. The data presented in Figure 3 shows that the 904 common-mode choke coil reduced the noise level by 6 dB at the peak levels. A 260-Ohm common-mode choke coil was tested next and reduced noise levels by 9 dB at the peak levels. The signal integrity (eye pattern) was minimally affected by the insertion of either filter.
This is a case where more may be too much. The common-mode impedance should not be too high; otherwise, it might distort an end- of-packet (EOP) waveform associated with USB 1.1 due to an inductive voltage effect. Put simply, the EOP signal in both USB 1.1 and 2.0 allows the receiver to know when the packet is complete. If the receiver does not receive the EOP, then the data will run together and become corrupted. Selection of an effective common- mode impedance value is key to preventing EOP waveform distortion in USB 1.1. USB 1.1 and USB 2.0 each use different encoding of the EOP signal. However, since USB 2.0 initially connects as USB 1.1, it will need to comply to the USB 1.1 EOP signal standards. As illustrated in Figure 3, the 904 choke coil has caused minimal distortion of the signal as compared to the 2604 part. Even though the 2604 part creates greater noise suppression, it distorts the EOP signal of USB 1.1. The 90-Q coil has a high coupling coefficient to provide minimal distortion of the differential signal while providing common-mode noise suppression. The common-mode impedance is such that the EOP of USB 1.1 is not distorted. Once the USB 2.0 connection is established, the distortion of the EOP signal is not a concern. USB 2.0 offers greater speed improvement to meet higher capacity needs of consumers. With the increase in speed, the design emphasis moves more in the direction of EMI filtering. The EMI solution for USB 2.0 isn't as simple as for USB 1.1; the ultimate solution is a common- mode choke coil that can reduce the noise without attenuating the signal. Selection of the correct common- mode choke coil is the key to reducing noise, maintaining high- speed signal quality, and preserving the EOP signal. After testing and reviewing the data, the 90-Q common-mode choke coil proves to be the most effective choice. Since USB 2.0 is projected to become the standard interface for most PCs, the need for effective and efficient filtering of EMI should be addressed and planned for early in the design phase. About the Authors Jonathan Davis is a product engineer for the EMI filter and chip inductor group at Murata Electronics. He recently joined Murata after completing his bachelor's degree in computer engineering technology at Southern Polytechnic State University. Gerry Hubers is the market segment manager at Murata Electronics North America. Since joining Murata in 1990, Mr. Hubers held positions in product management and product marketing. He has a technical background in electronics engineering from Radio College of Canada. Deryl Kimbro serves as a product manager for Murata Electronics North America. Mr. Kimbro earned a B.S. in electrical engineering technology from Southern College of Technology in 1995 and has been employed by Murata since 1997. Shuji Mikami joined the company in 1987 as an EMI engineer and today is group product manager for Murata Electronics North America. He became the EMI product manager in 1990 in Murata Europe and moved to North America in 1999. Mr. Mikami received his bachelor's degree in electrical engineering from Fukui University of Japan. Murata Electronics North America, 2200 Lake Park Dr., Smyrna, GA 30080-7883, 770-436-1300 EE June 2002 |
FORTRAN
Real time, or close approximations thereof,
computations are best done in a PC rather than in a peripheral microcontroller.
One reason is that a PC has a much more software-rich environment.
Also, in practice, PCs are infinitely fast, have an infinite amount of memory
and disk space. They cost practically nothing too. Maxtor 80 gig 7,200 rpm
drives are going for $99 in Albuquerque this week.
High speed USB 2.0 coupled with the lack of quality software for powerful peripheral microcontrollers is likely do cause some bankruptcies.
While C is available for peripheral microcontrollers and some operating system are available, FORTRAN is usually not.
Numerical analysts simply love FORTRAN. The reason is that extensive mathematical libraries have been developed over the years have FORTRAN code. And they know FORTRAN.
FORTRAN compilers stress turning-out code which almost matches the speed of assembler code.
In the middle to late 1970s IBM or Control Data FORTRAN compilers produced more efficient code than an assembler programmer could produce!
Rondall Jones was a mainframe computer consultant at Sandia labs for the CDC 6600 NOS computers.
Sandia physicists maintain large nuclear weapons code.
One physicist was complaining that his code took too long to run on a 6600. Jones suggested to the physicist that he switch on the optimizing compiler options.
The physicist did this. And the time to run the code went to ZERO seconds.
The compiler found a mathematical identity.
Complex computations, like Young's modulus of elasticity 1 2 with temperature compensation included, usually requires some scientist with an advanced degree and lots of mathematics knowledge.
These scientists require a FORTRAN compiler with libraries to make their computations.
So the approach of a super-fast 80C52 communicating
with a PC for USB 2.0 is attractive because the scientists can do their
computations on a PC rather than in a peripheral super-fast 16, 32, or 64
microcontroller with relatively few software tools.
Keep watching the
core meltdown feed.
Dr Richard J Hanson of LINPACK provides us with a list of FORTRAN compiler vendors.
Here another good Fortran link.
There are lots of reasons to favor the PC connected to custom hardware through a fast 80C52 with a USB 2.0 engine ... running Forth or BASIC, of course, opposed to tons of software on some peripheral microcontoller.
One point of this FORTRAN lecture is that it is the role of an embedded controller hardware/software designer to help solve global problems, not create a hardware/software mess.
This means that the embedded hardware/software
designer must take into account what other workers on a project will need to
make the total project a success. Wednesday June 5, 2002 10:01
Csd is following the port of the Palm OS from Dragonball to ARM core.
Porting Forth from the 8085 to the 8051 was no trivial matter.
Jerry Boutelle refused to write U/, -FIND, ENCLOSE and the assembler. So bill got to write these! Bill learned lots from Boutelle about metacompilers.
Palm Shares Hit Seven-Month Low
NEW YORK (Reuters) - Shares of Palm Inc. (NasdaqNM:PALM - News) on Friday fell to their lowest level in seven months after the dominant maker of handheld computers warned that quarterly sales would fall far short of target and pushed back its time table for achieving an operating profit. ....
Another real world "high tech, big
wreck" story. People usually underestimate the cost of computer hardware
and software development.
Csd is real
curious if the Palm OS was developed from a metacompiler. This is the best way
to do it.
Keep up-wind. Friday May 31, 2002 18:41
ARM cores may not do as well as expected because of USB 2.0.
Where do you want to do the computations? In a peripheral
microcontroller? Or in the software-rich PC environment?
Think the Tyson chicken model!
Charles Moore, inventor of Forth, created a very valuable operating system. But it's slow.
Speed of operating system execution is unimportant in most application of an interactive incremental assembler is resident with a Forth operating system.
Hardware accelerators to make Forth run faster are a passion with some. Like Charles Moore, Forth's inventor, for one.
ARM and Nazomi are into a lawsuit!!!
ARM Denounces Java Lawsuit as Frivolous
CAMBRIDGE, England, May 28 /PRNewswire-FirstCall/ -- ARM [(LSE: ARM) (Nasdaq: ARMHY - News)], the industry's leading provider of 16/32-bit embedded RISC microprocessor solutions stated today that it will vigorously defend itself against claims of patent infringement made by Nazomi Communications Inc. (Nazomi) in a lawsuit filed May 28, 2002, before the Federal District Court for the Northern District of California.
The suit claims that ARM's Java acceleration technology infringes U.S. patent number 6,332,215. ...
A lawsuit could, of course, ruin both Nazomi and ARM. Or even a government!
Please read about patent 6,332,215 and its relationship to Forth and Charles Moore. Wednesday May 29, 2002 14:02
Csd has been unavailable for unknown reasons for several days. So has the front page for the other project.
We've had to switch from ALOPress of HotMetaL PRO 6.0 for the reason that Comcast broke AOLPress.
HotMetaL PRO has as many bugs as AOLPress. And both a five start rated programs ... for good reasons. Both are outstanding html web authoring tools.
One of the more frustrating bug in HotMetaL PRO is that high-lighting text, then cut or copy doesn't always work very well. Also paste frequently doesn't work right. This has made it difficult to move articles from the home page to archive in file folders.
One of the reasons we've not been present is that we were afraid that the homepage got too large!
DON'T BUY USB HUBS. We're receiving reports that they crash windows XP. Tuesday May 28, 2002 08:15
Here's another 80C52 hot product. Csd is anxious to get Forth working on one or more of these SOCs.
Below is a ARM922T core with a Java accelerator. Since Java is a Forth derivative forth execution would be much faster than on an 80C52.
However, ARM is an expensive solution while the 80C52 is inexpensive.
Csd will continue to focus on 80C52, pci/cardbus, and USB 2.0 solutions. They look to be economic winners. Monday May 20, 2002 14:36
Chip Serves As Bridge Between PCI And Hi-Speed
USB 2.0
Said to be the worlds first PCI-to-Hi-Speed USB 2.0 device controller, the NET2280 chip provides an outside-the-box means of tapping the plug-and-play benefits of USB while retaining the high performance of PCI-based systems, such as video, data acquisition, and wireless network systemsthe traditional way of accomplishing this is by installing the PCI card inside the PC.
Providing a direct bridge between a PCI ASIC or CPU and Hi-Speed USB 2.0, which inns at speeds of up to 480Mbps, the new IC includes a 32-bit, 33-MHz PCI bus master or slave interface, a 4-channel DMA controller, an analog Hi-Speed USB 2.0 transceiver, a Serial Interface Engine (SIE), 12 endpoints, memory, an 8051 CPU, and a 30-MHz PLL. The 120-pin TQFP chip is sampling now, with production to begin in Q3. Price: $7 each / l00,000. NETCHIP TECHNOLOGY INC., Seattle, WA. (650) 526-1490. New SoCs Claimed To Knock Socks Off Competition sharpmsa.com In a product launch embracing a gaggle of new microcontroller (MCU) and system-on-a-chip (SoC) devices grouped under the tradename BlueStreak, the company announced a brand new, highly integrated SoC that combines an ARM922T core and a hardware Java accelerator in the form of a JIT Java compiler. The LH7A405s hardware Java compiler takes the place of either a dedicated Java processor or a software-based solution, with the latter traditionally running on a CPU and requiring large amounts of memory, reportedly resulting in a significant performance reduction and power drain. The LH7A405 SoCs JIT compiler is also said to allow optimization of Java program and cache instruction streams, as well as allowing other engines to be bolted to the hardware to process compressed files, such as MPEG and HTML/XML.. The new SoC packs a programmable LCD controller that supports STN, color STN, TFT, HR-TFT, and advanced TFT with up to 64 colors and 15 shades of gray and up to 1,024 x 768 resolution. Other key on-chip features include: a controller for fast processing of vectored interrupts; a USB host that allows the LH7A405 to communicate with devices such as digital cameras without going through a PC; a MultimediaCard (MMC) interface with Secure Digital Interface; four PWMs to generate tone and/or control LCD contrast; an ADC and touch- screen controller; and Smart Card and AC97 interfaces. The LH7A405 SoC is supported by an array of development tools and an evaluation board and is priced at under $32 each/l0,000. Production is scheduled for Q4. SHARP MICROELECTRONICS OF THE AMERICAS, Camas, WA. (360) 834-8002 EE Product News May 2002 |
Differential signaling, like RS-485 and MIL STD 1553, can cause bad data to be received, especially in the first one or two characters of a frame. And maybe at the end of a frame.
At csd we've only seen the first character being incorrectly received with 485. This usually occurs after the lines have been at rest.
Csd added pull-up and pull-down design tested in a perimeter security 485 installation in South Africa. This design did not entirely solve the problem but it helped.
The jury is still out on how lvds communications in general and USB 2.0 will work in practice. Thursday May 9, 2002 08:54
COLUMN > NO MYTHS ALLOWED
An ugly little secret about differential modes BY DR. ERIC BOGATIN Contrary to popular belief, there is no such thing as a differential mode. There is no such thing as a common mode. Ten years ago, 50% of the PCBs produced were controlled impedance boards. Today, virtually 100% of all boards are controlled impedance. Likewise, today about 50% of boards are designed with differential pairs, and in five years that figure may be closer to 100%. With so much emphasis on differential pairs, it is important that everyone uses the right terms in the right context so we all refer to the same things. An added benefit is that if we use the terms correctly, we wont be confused about the difference between odd mode impedance and differential impedance, as they refer to different quantities. A mode is a voltage pattern that is very special for a differential pair of transmission lines. If you launch this voltage pattern into a differential pair, it will propagate undistorted down the line. For a pair of microstrip lines, if you launch a pattern that is + 1 v on one line and 0 v on the other line (voltages with respect to the return plane), the voltage pattern will change as it moves down the lines. After all, there will be far end noise on the quiet line, which will grow with the coupling length. At the same time, the +1 signal will degrade as energy couples over to the quiet line. This voltage pattern is not a mode. However, if the voltage pattern is a + 1 v on one trace and +1 v on the other, there is no voltage difference between the two linesso there will be no crosstalk and this pattern will propagate undistorted. This voltage pattern is called a mode. It happens to he the even mode. The other voltage pattern that will propagate undistorted occurs when a +1 v is applied to one trace and a -1 v is applied to the other trace. This is called the odd-mode voltage pattern. The impedance looking into either line will depend on how they both are driven. When we drive the pair in the odd mode and look at the impedance of one line, we see that we have to put some extra current into it to not only drive the line but also couple to the adjacent trace with the large dV/dt between the two lines. This extra current means one trace has a lower characteristic impedance when the pair is driven in the odd mode. Drive the lines in the even mode and the current into one line drops, so the characteristic impedance is higher. Just looking at one line and measuring its characteristic impedance, we find that it depends on the voltage pattern imposed on both lines. This is why we have to specify the voltage mode on the lines to know the characteristic impedance of either line. When the signal on the differential pair is the odd-mode pattern, the characteristic impedance of one of the lines is called the odd-mode characteristic impedance. Likewise, when both lines are driven in the even mode, the impedance of one line is the even-mode characteristic impedance. Typically, the odd-mode characteristic impedance is lower than the even-mode characteristic impedance. When we apply a signal to the two lines that matches the odd mode, we usually say we are driving the lines differentially The differential signal is the voltage difference between the two signal lines. For a pure differential signal, this voltage is actually twice the voltage between either line and the return plane. The differential impedance is the impedance the difference signal sees. Since the difference signal is twice the voltage on either line, the differential impedance is twice the odd-mode characteristic impedance. If we think of there being a differential mode, then trying to figure out what the difference could be between the differential impedance and the odd-mode impedance is confusing. In fact, there is no such thing as the differential mode. There is only odd mode, odd-mode impedance, differential driving, differential signals and differential impedance. The odd-mode impedance is the impedance of one line when the pair is driven differentially. The differential impedance is the impedance the difference signal sees when propagating down the differential pair. Eric Bagatin is the VP and chief technical officer at Gigatest Labs [www.gigatest.com] and a frequent speaker at the PCB Design Conferences. Printed Circuit Design May 2002 |
The AMI product sounds similar to the Mentor Graphics USB/8051 product. You buy the design kit, turn out an FPGA, see if it works, then move on to an asic.
Csd has pretty-well figured-out what its local pci bus problem was. There has to be a coordinated 80C52, pci bus, OHCI and EHCI software solution to get both 80C52 Forth and BASIC52 working on one of these SOCs [system on a chip].
This is going to be really neat!
Here's a jpg of Sandia National Laboratories weapons component first 8051 printed circuit Forth boards.
Before this board, all systems were wire wrapped.
Those are 8kx8 ram and eproms. This was before 32kx8s were available. In about 1982-84.
In the upper-right are two 8255 ports. The ZIF socket is for programming eproms.
We first used Sandia's MNOS [metalized nitrided oxide semiconductor] eeprom to hold crypto key for the small missile launch control system. $1,500 per chip!
The first system was an 8085 Forth system. But this was changed to an 8051. Lot's fewer parts! Monday May 6, 2002 10:55
Leading edge
What's hot Design kit targets USB MI SEMICONDUCTOR has launched a design-and-development kit that simplifies the integration of USB interfaces into system-on-chip ASICs for advanced communication products. Suitable for both initial ASIC design and FPGA-to-ASIC conversions, the design kit provides all of the intellectual property, supporting tools, and documentation to deliver a complete USB device. The kit combines modular RTL- code descriptions of the USB- interface engine and other key USB-device-design blocks with a simulation testbench that features test and verification scripts and a USB test host. The kit comes with an 8- bit R8051 RISC-microcontroller core, but it is also compatible with other third-party microcontrollers. The kit supports USB 1.1 and 2.0 standards as a low- or full-speed device, and you can easily integrate the RTL blocks into variety of USB-device configurations. Configurable RTL module available with the kit include the USB-interface state machine. The state machine handles all low-level USB-protocol functions, monitors the status of addressed end point buffers, and manages data transfer between endpoint buffers and the USB port. The kit implements all endpoint buffers with a single dual-port RAM within the endpoint- group RTL module. The device architecture can support as many as 31 fully configurable endpoints. Other hardware features of the USB device include external microcontroller ports, a DLL clock generator, and a USB-transceiver pad that sources and sinks serial USB data. AMI Semiconductor, 1-208-233-4690, www.amis.com. EDN May 2, 2002 |
Another heavy-weight has entered the 8051 core arena. Analog Devices.
NEC is starting to come-through with µPD720100A documentation. The requests went to Denver. Santa Clara appeared not to have responded properly.
Whoops!
It looks like the User manual hasn't been written. Or, better yet, likely not translated from Japanese to English! Wednesday May 1, 2002 07:49
www.analog.com Electronic Design April 29, 2002 Who out of the 80C52 core companies is going to do well? We think all of them are going to do just fine. Just like auto dealships learned that when they grouped, they all sold more cars. But especially those who get BASIC52 and 8052 Forth working on their parts are going to do the best of all. Of course! Faster, cheaper software and hardware development. More reliable in the field. Good hardware and software developers need an interactive operating system with incremental compiler and assembler with source code availability. Both at the 80C52 and wdm driver sides. Csd is working on this. And other things too. While, of course, planning even more fun projects! |
Csd did a bit of telephone intelligence work yesterday.
Cygnal has a major blunder to overcome. The transformer to generate voltage for their board.
Cygnal should have emulated Cypress and taken power from the USB connector.
The good news is that Cygnal is going to have a USB version of its development board hopefully to be released this summer.
The Cygnal USB news was first heard about a month ago by phone from Phoenix and verified yesterday by a phone call to Dallas.
The Dallas caller said that Cygnal was using a low drop-out voltage regulator.
Reinvention, as opposed to adopting the good ideas of others, is very bad.
Rejection of others' bad ideas equally important.
Then, of course too, Cygnal has software problems. Friday April 26, 2002 08:27
There is no need to reinvent a circuit design for converting from USB 5.0 to 3.3 volts. Cypress/Anchor engineers have done this for us. The trace going vertical to the left of R11 is input from the USB bus. Let's hope Cygnal engineers didn't reinvent. Reinvention in the software arena is an equally bad idea. Here's the a/d description of Tom Katausky's BASIC52 taken from the Jameco Catalog Vol. 222 April-July 2002 80C52 BASIC Microcontroller Let's all hope some of the 80C52 SOC [system on a chip] manufacturers sees the merit of porting BASIC52 to their chips so that it can be FLASHED. Csd is focusing on 80C52 FORTH as another alternative to the Keil C/assembler development environment. Both BASIC and FORTH are real time operating systems. C is not an operating system. Also BASIC and FORTH have field-proven to be more reliable by orders of magnitude than C/assembler systems. The former USSR fuzes its weapon systems using FORTH. And the US space program runs on Forth too. Jameco also pushes Parallax BASIC stamp products which are mounted on hybrid carriers.
Very surprising is that Parallax is using the Scenix/Ubicom microcontroller which was designed as an internet micrcocontoller! Maybe Ubicom has shown exceptional brilliance by going the BASIC software route rather than the C/assembler route? |
Csd still hasn't gotten any manuals on the NEC µPD720100A yet.
Csd will phone for the third time and try an email too.
Looks like NEC is willfully withholding technical information from those who want to develop software for the µPD720100A.
There are lots of USB 2.0 pci cards on the shelf at CompUSA.
Let's see what happens with USB 2.0 adoption.
Csd is remains fairly concerned about how the USB 2.0 lvds drivers will work in the field.
Both RS-485 and MIL STD 1553 use differential drivers. Both have problems with leading characters in a frame. And 1553 with trailing characters.
Opto drivers would be lots better! Wednesday April 24, 2002 07:57
USB 2.0 Expected to Achieve Rapid Success BRIAN OROURKE THE UNIVERSAL SERIAL BUS (USB) IS ONE OF THE most successful interface standards in the history of PCs. USB was designed as an open interface standard to provide a common interface between PCs and peripherals, eliminating the need for legacy connections such as serial, parallel and PS/2 ports Over time, the USB standard has evolved. USB is a master/slave architecture requiring a host, usually a PC, and has evolved overtime to include some consumer electronics products. Host, peripheral and operating system all must support USB for the interface to operate. By all accounts, the USB standard has been a huge success, having been installed into approximately one billion devices at the end of 2001.
The original USB 1.0 standard, with a maximum data rate of 1.5Mbits/sec., was released in 1996. The USE 1.1 standard followed, with a maximum data rate of l2Mbits/sec. and backward compatibility with USB 1.0, opening the market for peripherals and consumer electronic products that required higher throughput. USE 2.0 is the latest update to the USB standard, with a maximum data rate of 480Mbits/sec., significantly raising the total amount of bandwidth available in a single host design. The increased pipe should help expand the penetration of USE in high-bandwidth applications such as external storage devices and PC cameras. The introduction of USB 2.O in PCs will happen in three phases. The first phase was the introduction of PCI add-in cards in late 2001. The second phase was the introduction of a discrete host controller in the motherboard of some high-end PCs, such as Gateway Computer Corp.s 700XL desktop PC introduced in January 2002. Intel Corp. also released two desktop motherboards, with NEC USB 2.0 host controllers. The third phase will be the integration of USB 2.0 capability into a core logic chipset. Intel will offer this capability in its 845G chipset in the second quarter of 2002, with other chipset manufacturers expected to follow in the second half of the year. Once USB 2.0 begins to penetrate the PC market, USB 2.0-equipped PC peripherals and consumer electronics should follow. In addition to external storage and PC cameras, scanners, printers, network hubs and digital modems are expected to quickly adopt USB 2.0. In consumer electronics, MP3 players, digital camcorders and digital cameras will quickly adopt USE 2.0, too. USE 2.0 should gain rapid acceptance in the marketplace. According to the In-Stat/MDR report USE: The Universal Connection, USE 2.0 will be a worthy successor to previous USE specifications, with a 220 percent compound annual growth rate predicted between 2001 and 2006. Brian ORourke is a senior analyst for multimedia service and Internet access devices. He can be reached at borourke@instat.com. In-Stat/MDR is owned by Reed Business Information, the parent company of Electronic News. Electronic News April 8, 2002 |
Csd phoned TI on Thursday April 18 at about 16:00 to request literature and one sample.
Early Friday morning, two samples of the MSC1210Y4 arrived!
We are seeing the start of complete 80C52 systems on a single chip.
Soon csd expects to see 64 kbyes of program FLASH or RAM and 64 K bytes of RAM data memory.
Hopefully the designers will have an option for ANDing -PSEN and -RD so that either than Harvard or Princeton architecture can be implemented.
Look what Cygnal is up to
Cygnal Releases SOC with
CAN Bus ConnectivityThe C8051F040 is a complete mixed-signal system-on-a-chip with CAN bus connectivity The device has a fully compliant CAN 2.OB bus with 32 message objects seamlessly integrated to the companys 25 MIPS 8051 core. The programmable gain amplifier offers offset and gain control into the 12-bit 100 ksps ADC over a 60 V common-mode input range. The device is suitable for CAN industrial automation and sensor applications. Other features include a 25 MIPS CPU; CAN 2.OB with dedicated 32 message object RAM; two UARTs, SMBus, SPI bus, five 16-bit timers, six-channel PCA; 64 KB in-system programmable Flash; external data memory interface; 12-bit 100 ksps ADC; two 12-bit DACs; 60 V input programmable gain amplifier with gain settings from O.05x to 16x; comparators, reference and temperature sensor; 8-bit 500 ksps ADC; and JTAG non-intrusive, in-system debugger with trace buffer (no emulator required). The device operates between -40 degree C to 85 degrees C and is packaged in a 100-pin TQFP and 64-pin TQFP. Cygnal Integrated Products, Inc., 4001 Westbank Dr., Bldg. B, Ste. 100, Austin, TX 78740; (512) 327-7088; Fax (512) 327-7087; www.cygnal.com.
ECN April 2002
Now, of course, we have to get a good operating system ported to some of these super-80C52s.
Wouldn't it be neat to port Tom Katausky's [spelling?] BASIC52 operating system to one of these new 80C52s?
To this end csd is posting the source of BASIC52 [basic.asm] and floating point extension [Fp52.ASM] so that it could be ported and installed in FLASH on one of these SOCs.
Former Intel employee Mark Thompson [read book preface first paragraph] gave bill the copy of the posted source code. Bill has also seen the code posted at other locations.
One would also want to extend the BASIC with an incremental 80C52 assembler! This has been done, csd thinks.
Let's look at the Forth and BASIC competition!
Triscend E5 Support
The Triscend ES family of Configurable System-on-Chip (CSoC) devices is based on a performance accelerated 8-bit 8051 microcontroller. The ES Family consists of several devices which vary in the amount of on-chip programmable logic, RAM, and I/O. The ES is ideal for embedded systems applications that demand both fast time-to-market and high levels of customization.
Software development tools (assembler, ANSI C Compiler, and target debugger) from Keil Software fully support the ES and provide a platform for programming and debugging your real-time, in-system target hardware.
Configurable System-on-Chip (CSoC) Designers may configure ES CSoC chips using the Triscend FastChip Development System. FastChip software customizes the processors peripherals using a drag and drop methodology which allows you to create your own 80S I derivatives on-demand. Simply select the desired peripheral from the library and drag it into place. FastChip makes the rest of the process automatic so you can begin code development without worrying about the implementation details of the peripheral set.
JTAG Debugging with pVision2 The Triscend ES family supports target-level debugging via the built- in ITAG interface. Using the Triscend JTAG driver with the Keil pVision2 Debugger, you may test application programs running on your actual target hardware.
The JTAG interface allows you to download program code, set breakpoints, watch memory locations, and single-step through your C or assembly programs. Since the ITAG interface works with the Keil debugger, you can instantly get started writing and testing embedded programs for the Triscend ES family. The JTAG debugger gives you full access to the Triscend ES peripherals. ....
Getting Started
The best way to start your own CSoC project is to get the Triscend ES Development Kit. This kit includes the E5 evaluation board that is designed to help you create working programs using the ES family of devices. All features of the board are fully supported by FastChip and pVision2.The ES Development Kit provides a flexible and powerful platform for quickly creating, developing, and debugging embedded system designs based on the 8-bit E5 family of CSoC devices. It features the ES evaluation board, the ES base board, a power supply, and a parallel cable for download and debug.
E5 Development Kit
Features
Triscend TES2O CSoC Device
128 KByte Flash Memory <<<< dig this!
Serial Configuration EEPROM Socket
40 MHz Oscillator
JTAG Debug and Download
Two 7-segment Displays
RS-232 Serial ChannelThis complete kit is available from Triscend. For more details, visit www.triscend.com.
So we all have to get serious about getting BASIC52 and FORTH working on the new 80C52 SOCS!
Or do you want to send lots of your money to Germany ... for old technology? Monday April 22, 2002 09:34
Data Converters Lowest noise precision data system on a chip.
The MSC1210 from Texas Instruments achieves a new level of high performance and functionality in mixed-signal processing. This 24-bit, low-power (4 mW) delta-sigma ADC integrates an enhanced 8051 processor core, Flash memory and a variety of on-chip peripherals including an additional 32-bit accumulator and an SPl-compatible serial port. The high-performance 8051 core executes up to 3X faster and at lower power than a standard 8051 core. With its high level of analog and digital integration, the MSC1 210 is ideally suited for a wide range of precision applications requiring small size, high integration and user-selectable features.
Burr-Brown Products Applications Features www.ti.com/sc/msc1210 The Wold Leader in DSP and Analog EETimes April 15, 2002 NEWS ON TECHNOLOGY R&D, PRODUCTS, AND BUSINESS Precision DAS Integrates 24-Bit ADC With 8051 Core And Flash Memory Targeting applications ranging from weigh scales to industrial process control, Texas Instruments Burr- Brown Division has developed a highly integrated data acquisition system (DAS). It merges a 24-bit delta-sigma (delta-sigma) analog-to-digital converter (ADC) with an 8051 microcontroller and flash memory on the same piece of silicon. Implemented in 0.5-µm CMOS, the MSC 1210 is the first fruit of this integration. It boasts on-chip a 6-MIPS 8051 core and up to 32 kbytes of flash memory and 1 kbyte of SRAM. Four different memory configurations let users partition the flash between program and data memory for each configuration. The flash memory can be independently programmed at 3 and 5 V using both serial and parallel programming methods. And, the 24-bit ADC offers up to a 1 ksample/s rate. Other features on-chip include an input multiplexer with eight differential/single-ended channels, a buffer, a programmable-gain amplifier (PGA) with maximum gain of up to 128, a temperature sensor, a voltage reference, an adjustable digital filter, a 32-bit accumulator, three 16-bit timer/counters, system timers, a watchdog timer, and power-management control. Plus, it offers special function registers (SFRS) for control and status purposes. Typical power consumption for the unit is 4 mW at 3 V. The MSC121O uses separate analog and digital supplies. Available in a 64- lead TQFP, it costs $8.95 each in 1000- piece quantities. CIRCLE 504 Texas Instruments Inc., www.ti.com; (520) 746-1111. Ashok Bindra Electronic Design April 1, 2002 |
Csd is experienced in furnishing real time dll drivers for Windows 3.11, 95, 98 for industrial controls. This works. But is it not without problems.
Reliable Windows 2000 with wdm drivers are replacing dll, or even vxd, drivers.
Csd prefers to put a 80C52 between Windows 2000 and custom hardware. But csd has done direct coupling to custom hardware.
USB looks to replace serial and paralell communication. Provided that the lvds drivers are resistant to noise corruption.
Csd has lots of experience with RS-485 drivers. 486 and MIL STD1553 have nasty habits of corrupting the first one of two characters in a frame.
1553, csd was told, tosses away the first two characters and the last character as part of the protocol! And 1553 goes into F16, F18, and A10 fighters!
Csd heard a story where a new 1553 box was plugged in to an F16 ... and the plane quit working at the end of a runway! Tuesday April 16, 2002 12:54
PC control hits factory floor
By Charles J. Murray CHICAGO Driven by rising cost concerns, manufacturing engineers are taking a second look at PC technology and finding ways to cut dollars off the factors floors bottom line. Engineers are shaking off long-held belief that the personal computer isnt up to the task of controlling manufacturing operations and are instead looking to leverage the PCs commodity nature. The result: Sales of PC control software have climbed to about $75 million annually, up from about $40 million two years If the trend continues at its current pace, experts say, it could signal a start to the long-expected change in the shape of the $1.4 billion U.S. market for programmable logic controllers (PLCs). The change is finally starting to take place, said Dick Slansky, senior analyst for the ARC Advisory Group (Dedham, Mass.). Were seeing manufacturers making a major architectural shift toward PC- based controls. Indeed, General Motors Corp., the worlds biggest manufacturing enterprise, said last week that it has embarked on a program to adopt a hybrid-type control technology that mixes the characteristics of PCs and programmable logic controllers (PLCs). The automotive giant said that it plans to use the technology across its entire North American manufacturing operation. GM has already adopted it in three manufacturing plants in Moraine, Ohio, and Oklahoma City, where it builds the GMC Envoy and Chevrolet Trailblazer; and in Lansing Grand River, Mich., where it makes the Cadillac CTS sedan. Were using it for all new programs going forward, said Phil Disch, engineering director for controls, robotics and welding at General Motors. If we put an all-new body shop or assembly operation in North America, were going with this technology. GMs move mirrors similar decisions by other major manufacturers. At National Manufacturing Week in Chicago two weeks ago, Microsoft Corp. teamed with 30 industrial partners in its show booth, including such giants as Siemens, Rockwell Automation and Dell Computer, all of which are involved with Windows-based manufacturing solutions. Industry analysts say that such partnerships reflect a growing acceptance of PCs not just for the storage of information, but also as a tool that can be used to control automation equipment. No turning back "Theres no doubt now that industry is finally heading more toward PC-based systems, noted Tom Bullock, president of Industrial Controls Consulting Fond du Lac, Wis.). And were going to see a lot more of this as Ethernet use grows on the factory floor. Experts say that the shift toward such controls is taking several forms, ranging from traditional white-box PCs with windows-based control software to solutions that mix the characteristics of programmable logic controllers and PCs. As a result, industry analysts say that the industrial-controls market is changing in away that makes the phenomenon difficult to quantify. Most of the big industrial-controls vendors, such as Rockwell Automation Mayfield Heights, Ohio) and Siemens Energy & Automation software packages that enable PC-like industrial controllers to emulate the behavior of PLCs. As the lines between the two technologies blur, experts say that customers are often unaware that their PLC may actually be a PC in disguise. Whatever forms the solutions take, analysts say they are a reflection of a sometimes-grudging form of acceptance of the PC by industrial-controls engineers. Goodbye to status quo Its less of a religious issue for them now, noted industry consultant Nat Frampton of Real Time Development Corp. (Pearl River, La.). Its becoming more of a functional matter. OEMs are saying, I need this capability. Supply it to me. Such attitudes represent a sharp departure from the old status quo in the manufacturing world, where engineers long held that PCs simply werent cut out for the rigors of factory floor control. PCs, they said, reached obsolescence too fast; they werent robust enough; they lacked the ability to stand up to dust and vibration; and, worst of all, they lacked so-called determinism. Determinismthe ability of a machine to focus on a critical taskwas key to the automation community. Engineers said they didnt want a PC to be polling its hard drive while a factory floor operator was frantically pressing an emergency stop on the assembly line. For that reason, many engineers dug in their heels and maintained their loyalty to the PLC, which has dominated the factory floor for nearly 30 years. But while determinism was a issue three years ago, experts say that its not anymore. Operating systems such as Window CE now offer scan times of less than 5 milliseconds, well within the hard-real-time definition set by the Open Modular Architecture Users Group. Such Windows systems also have cycle variations, or latency, of less than 100 microseconds, also within the OMAC standards. As a result, some industry consultants now believe that PCs are more deterministic than PLCs. By the current definition of real-time, you could easily make the case that PCs are deterministic and PLCs arent, Frampton of Real Time Development said. Almost everyone agrees that determinism is no longer in issue for PCs, added Bullock of Industrial Controls Consulting. PCs are so fast that it reality doesnt matter anymore. Still, PLCs clearly dominate the industrial-controls market, amassing about $18 in sales for every dollar spent on PC-control software. Even though people have touted soft control, PLCs have stood their ground, noted Ron Bliss, Logix Netlinx software marketing manager for Rockwell Automation. Theres already an enormous installed base, and the attitude is, If it aint broke, dont fix it. Suppliers such as Rockwell Automation are developing an insurance policy against PC encroachment, however, by offering so-called soft-PLC products. Such software programs behave like PLCs, even allowing developers to code them using ladder logic, but are built on PC- type architectures and use Windows operating systems. Rockwells RSLogix 5000, which allows engineers to create a PLC-type controller but employs PC programming techniques, serves as a cornerstone for GMs recent industrial-controls decision. GM uses the software in general assembly operations and even in more complex systems, such as stamping presses. The companys engineers say that RSLogix provides them with a common platform for all automation, and supplies then with an open architecture, which wasnt the norm during the heyday of the PLC. As a result, the automotive giant expects to save on manufacturing expenditures. Now, all our panels and all our equipment are identical from plant to plant, said Disch of GM. All of our programming is the same, too. Because we write all the code in templates, we can autogenerate that code and use it over again in new plants. A programming edge Disch said that RSLogixs PC- type programming techniques provide an advantage that was not available from old-style PLCs. If we write logic for a clamp-open procedure in a simple assembly machine, then we can use the same logic again in a clamp-open for a more complex machine, such as a stamping press, he said. Like many new controls products, RSLogix falls into the gray area between the PC and PLC, Therefore allowing GM to continue in a PLC environment, despite the fact that RSLogix can be Windows-based. GM is running the software on a Control- Logix 5550 industrial controller, which Rockwell stresses is neither a PC nor a PLC. Ultimately, makers of pure PC-control solutions hope to capture customers who are hovering in that area between PLCs and PCs. Companies such as Entivity Inc. (Ann Arbor, Mich. ), which makes soft- PLC controllers that are Windows-based and can run on white-box PCs, believe their products will serve as the next logical step for such customers. Pure PC products, they believe, can be more easily integrated with the front office, as well as with customers and vendors computers. Analysts believe that the economy may ultimately be the driving force that pushes the automation community toward the PC. When the economy was moving along at full tilt two years ago, the order of the day was for engineers to go to their regular suppliers and use their traditional solutions, noted Slansky of ARC Advisory Group. But since the economy has gone down, theyve had time to analyze their technical requirements, and theyve found that PC-based control can help them. Like GM, many manufacturers are finding that PCs can offer economies of scale that can translate to lower engineering costs and lower initial system costs, Slansky said. For that reason, PCs are displacing PLCs at the high end of the market. The PLCs stronghold, however, remains in the low end of the market, where PCs cant compete with the $300 to $500 costs of so-called shoe box PLCs. By some estimates, such products now account for more than half of the existing PLC market. During the next five years, however, engineers expect PC and PLC hardware to continue to merge, in the same way that PC and PLC software is merging today. At some point, we are looking to build a hardened version of a PC that would run inside our hardware and provide a PLC life cycle, said Bliss of Rockwell. Eventually, we see the market migrating that way. Electronic Engineering Times April 8, 2002 |
Csd is tracking Comcast for several reasons.
1 that where it has one of it Internet connections.
2 Comcast is going to redo the work that Excite did ... and went broke doing.
Also we're following what could be a very disasterous financial year for high tech - except for the fast 80C52s with a reliable inexpensive operating systems hosting its own incremental compile and assembler, of course! Thursday March 21, 2002 14:44
Cassel's Corner by Paul Cassel Cassels Picks and Pans And the Slack-Jawed Imbecile award goes to. ... From time to time I write a picks-and-pans type of column, which praises or condemns whatever computer-related vendor, product, etc., I think warrants such attention. This is one such column. Read on at your own peril. .... .... I used to recommend Comcast's cable Internet service. I am now lukewarm at best. The company moved from good, cable speed at a good price with a good ISP @Home/Excite to having still good connection speeds, but a complete botch-up job at the ISP end. Instead of a simple manual setup for mail and news, the new outfit tries to rope you into installing its miserable add-ons to your Web browsers. Usenet newsgroups, once elegantly handled with @Home, now are some sort of Web-based mess. The entire Web site that is the gateway to various services is defective in design as well as operation.
My guess is that Comcast decided that its customer base, or its desired customer base, consisted of dimwitted refugees from AOL who desire to be controlled, as they were when they were with the Internets Great Satan: Here I award the Slack-Jawed Imbecile trophy to Comcast for not knowing that its base is the experienced user who is willing to pay more for high bandwidth. The new Web-based misery which "munged two computers I tried it on as thoroughly as the previous champion, AOLis an insult to Comcast's own base. It remains an utter mystery to me why Dell has the reputation as the professional choice over other vendors, such as our own locals, or even national rival Gateway. This is the slimiest computer-related company I know of in business today. This outfit will lie, cheat, steal, and sell worn, recycled parts as newand yet it marches on; spreading its poison across the land. There seems to be some sort of weird dampening field around the company that suppresses accurate word of mouth about this outfit. I hereby award the Sauron Trophy to the great Texas Computer Company, Dell. The Saruman/Asmodeous/Endor Award this year once again goes to the perpetual winner, Michael Dell. Some of you have wondered why I never trash local vendors like I do the nationals. The reason is simple. By the time I get enough dirt on a local to go public, its out of business anyway. Dell can cheat or steal from as many New Mexicans as it wants to, and get away with it because our self-hatred will cause us to blame ourselves, not the Texas Wonder. However, let a neighbor do even half of what that outsider does and were on him like a bag of hot sweat. Its either that, or the complaints I get are more aesthetic than worrisome. For example, I got a complaint that a fellow walked into a local computer store and didnt get waited on because the rather young, vigorous and good-looking sales staff was engaged in variously staged mating rituals. Now, that guy was flustered, but not harmed in any way, so I didnt think it was worth a mention in the column. I still dont think its worth a mention, so dont expect to read about it here. Finally, kudos to both AMD and Intel for keeping the chip wars going. This has resulted in better products at lower prices for all of us. AMD XP vs. Pentium 4? Ive used and like both. I still will prefer AMD to Intel,:. but I sure Wouldnt refuse the Pentium if it were offered at a good price. Keep your mainboards cool, your spreadsheets hot, and Ill see you all in the ether. Have an opinion that differs from Cassel's, or the views of any of ComputerScenes writers? If so, let the rest of us knowwe welcome your comments. E-mail your views to: editor@cscene.com from the publisher If there is one aspect of life that is constant, it is change. Now, after more than two years of my involvement with ComputerScene Magazine in New Mexico, we are going through a significant transition. Often, changes occur not, because they are desired, but due to external forces. Circumstances dictate a new view and varied approaches. Over the past two years, the economy of New Mexico has dealt devastating blows to the computer businesses in the state. The number of computer dealers and Internet services has dropped significantly. As a consequence, ComputerScene revenues dropped to a level that had made me seriously question the present and future viability of the magazine. I didnt see a way to continue operations, and had in fact made the decision in late January to cease publishing as of this issue. This was communicated to our advertisers in February. I did not feel that it was possible to cut costs enough to stem the losses, nor was it reasonable to expect the advertisers who had managed to weather the tough times to assume more burden. I had already started the closing process. Early this month, the staff in Albuquerque approached me with an alternative plan. If there were enough commitment from our current advertisers, ComputerScene would return to monthly publication with a greatly reduced local staff (one person, Rebecca Lynch). I said that I would consider the possibility, if Rebecca decided that it would be worth her time to assume some of the risk personally. We discussed the terms, she talked to the advertisers, and we are moving forward with a monthly publication starting on May 1. Patrick Can, who has done a magnificent job of producing ComputerScene single. handedly, will continue with his graphic design business. Over the years weve worked together, he has never needed or desired any pestering from me, and has reliably met all deadlines while producing outstanding quality. His departure will necessitate some cosmetic changes in the name of cost-cutting, but you will see this in May. Capitalism is a democracy. Votes are cast with purchasing decisions. Those businesses that do not receive the votes do not deserve to continue. Everyone must make the decisions that are best for them individually, whether it is as a consumer or a businessperson. If you buy from our supporters, you are saying that you appreciate the service we provide. If you buy somewhere else, we will eventually get the message that were not doing the job, and either change or disappear. For now, were changing. ComputerScene has nearly 100,000 readers in New Mexico. Only a few of you read this column. Please save your fellow readers time and aggravation by pointing out to your friends and cohorts that we are now published monthly. The next issue will not be on the stands until May 1. We will be significantly increasing the number we print, so that ComputerScene will be available all month. Jack Dunning ComputerScene April 2002 |
Csd is zeroing-in on the source of the Cypress EZ-USB-FX Windows 2000 crash.
It appears to be related to a code called fastinst.
Not really understanding both hardware and software
before committing to either can result in field disasters.
Thursday
April 4, 2002 13:57
Csd downloaded the 3671 [and discovered it is 3671 not 3761!] software yesterday.
Csd's original download was about December 24, 2001. The new revision date is 02/21/02.
Csd discovered that the .dsw and .dsp files have been removed. For business reasons?
In summary, csd observed that the Monitor originally loaded at start-up, then quit doing this. The monitor must now be loader manually.
Then the Cypress software crashed Windows 2000.
Csd tried to install the system downloaded yesterday. It, too, crashes Windows 2000.
But csd is getting an idea where the problem lies.
The Cypress solution is audaceous in that it is all
software. But will it work? And be economical? We'll see.
Tuesday
March 26, 2002 07:44
Csd can fix its Windows 2000 disk by removing the Cypress drivers and .inf file. So the problem is definitely in the Cypress driver software.
In trying locate the exact cause of the Cypress problem, Csd decided to recompile the code from source. And got a somewhat nasty surprise.
The code Cypress sent csd on cd rom does not contain the dsw and dsp files necessary to directly use Microsoft Visual C/C++.
Csd previously compiled the Cypress code. The downloaded code contains the dsw and dsp files!
In order to rebuild the Windows 98 disk trashed when trying to recover from the Cypress bug, cd bought a Maxtor 40 gig. Accompanying the drive is
Csd's speculation that a lot of products using USB 2.0 may be miniport drivers not conforming to the full USB 2.0 protocol.
CompUSA is now selling an new Adaptec USB 2.0 six port card with "contains driver software" printed on the box. $69. Monday March 25, 2002 08:09
Here's the directory from the cd rom Cypress sent to csd. Not no dsw or dsp. files! You can't directly build the project. You might conclude that you need to buy Compuware/Numega Driver projects to build the project from the software on the disk. And, yes, Compuware/Numega has a subdirectory on the cd advertising its products. |
Here's the directory from the download from Cypress 1. Note that the dsw and dsp files are present. This means you can build the project using Visual C/C++ 6.0 after C/C++ converts from 5.0 to 6.0. |
Csd is still in the process of trying to recover from the Cypress EZ-USB-FX problem.
The problem apparently was caused partially by windows 2000 handling of .inf files.
But csd learned that the Cypress development system works fine under 98! And the problem didn't occur in 98.
Csd is currently testing a Maxtor 20 gig with Maxtor PowerDiag 2.5 prior to reinstalling Windows 2000.
Yesterday csd formatted the 20 gig with MaxBlast, then copied contents of a 98 slave to the 20 gig. Then installed a windows 2000 upgrade. Windows 2000 reported that there were problems on the slave and proceeded to do BAD THINGS to the 98 slave.
98 still boots but lots of errors are reported at boot up. So don't try this.
While involved in the mundane tasks of getting about three disks repaired and looking at a lot of books, csd points out that having a microcontroller talking to windows lots more complicated than talking to dumb hardware.
Baker has a neat statement about how much a driver writer should know about hardware
The Hardware Environment
For some people (you know who you are), hot solder is the only true programming language. If you're not in that category, this chapter will give you a gentle introduction to those aspects of hardware that have an impact on writing drivers. You'll also find here a quick tour of the major bus architectures supported by Windows NT, and a few words to the wise about dealing with hardware in general.
2.1 HARDWARE BASICS
There are a number of things you need to know about a peripheral device before you can design a driver for it. At the very least, the following items are important:
How to use the devices control and status registers
What causes the device to generate an interrupt
How the device transfers data
Whether the device uses any dedicated memory
Whether the device can be autoconfiguredThe Windows NT Device Driver Book
A guide for Programmers
Art Baker
Now that we have some real intelligence with hot 80C52s at the peripheral end, things become lots more interesting. This, of course, is why a good interactive operating system on the 80C52 will make life lots easier and, thus, fun. Friday March 22, 2002 08:24
There sure are a lot of companies now focusing on embedded software. And with the IP core meltdown in progress likely fewer processors to port operating systems too.
But the 80C52 family is going to do just fine.
Provided it has fast usb communications with powered from the cable. No wall transformers allowed. Or interface modules to the development board.
And a decent interactive operating system with incremental compiler and assembler compiling, loading and linking source code on the 80C52. Thursday March 21, 2002 07:20
Mentor Takes On Wind River
Acquisition spree aims at market leader in embedded system OSes BY ED SPERLING & GALE MORRISON Mentor Graphics is on a buying spree to gain ground in the embedded systems world. While some companies live for acquisitions, Mentors history has been exactly the opposite. The company went through three years of on-again/off- again negotiations before announcing last week that it would acquire Accelerated Technology Inc. But with customers clamoring for one-stop shopping, not to mention the shadow of Wind River Systems Inc. getting increasingly longer, Mentor believes it no longer can afford to sit back and watch. We have an ambition to be the largest player in embedded software, says Mentor President Greg Hinckley. Right now were at $30 million and Wind River is at $400 million. Obviously we cant do it all internally. Meanwhile, Wind River doesnt believe Mentor can do it at all. The company argues that buying up small real-time operating system (RTOS) vendors is no indication it understands the bigger changes in the embedded software world. Its nice for them to try to supercharge their embedded business, says Dave Fraser, VP and general manager of Wind Rivers networks group. But its a little bit lateespecially since the embedded business has changed completely. Its not about operating systems and tools anymore. Its not about just supplying a little VRTX (Mentors RTOS) or a Nucleus, in the case of Accelerated. Fraser says the real challenge now is solving business problems, such as how to improve the efficiency of their development programs, how to get development teams from various acquisitions working together, and how to become more competitive. Customers are definitely looking for OSes and tools, but they need a full range of things, he says. They need networking stacks. They need professional services and a global partner. Our customers are giant companies with huge development departments looking to gain big efficiencies. Quite frankly, we are more like a business re-engineering company than the embedded OS company thats our heritage. Fraser also contends that Mentors track record in the embedded world is mixed. When Mentor bought Microtec, they got a superb debugger in XRay. They took that from being a very nice and strong product to being irrelevant, Fraser says. Ten years ago when they bought Ready Systems and Jim Ready was an icon in the industry at that time.. .During that transition (the integration into Mentor), VRTX went from being the No. 1 kernel from the No. 1 company to essentially being a dead product, to just disappearing. Jim Ready, who left Mentor in 1999 to found embedded Linux player MontaVista, warns that Mentor is taking on a lot with a challenge to Wind River. You can imagine that Mentor sees the same market research that others have, Readysays. [Embedded software] ought to be a very good place to be. But Wind is a formidable, well-financed, obviously successful company. I would not want to go after them on their turf. But Ready says there is a place for Mentor, if it can play from its strengths. Mentors strength is on the EDA side, the tie-in to Seamless, Ready says. Seamless is Mentors hardware/software co-verification tool, which boasts 90 percent market share, according to Gartner Dataquests most recent numbers. Dataquest forecasts this to be about a $25.4 million market this year. Where Seamless is strongest is the smaller, less connected devices. Some customers can really benefit from Mentors strengths, Ready says. Theyve found the closest coupling with the kind of things that Seamless can support: the smaller, more compact kernels... .Mentors strengths are Seamless and the XRay debugger. Thats why they bought Microtec, and they put a lot of work in that. It makes sense to pour gas on that fire. Fraser and Ready contend that Mentor, acquisitions or not, wont be playing in the mainstream embedded systems world where connectivity is paramount. Mentor says it will. Moreover, Hinckley says that with Neil Henderson running the embedded systems division Henderson was president of Accelerated Technology until the acquisition and will now become general manager of the Mentor embedded divisionthe team is set to make those acquisitions work. But it clearly will have to work far more quickly than in the past. Mentor has a lot of lost ground to make up, and it has been extremely slow at closing deals. It spent four years acquiring Ikos Systems, which it also announced last week. In the case of Accelerated Technology, Henderson says the deal was almost closed six months after negotiations began, but the two companies couldnt agree on their combined direction. It took another 42 months to finally strike an agreement. A lot of time was spent on potential integration issues, Henderson says. Price was less of an issue, and executives involved say it will be even less of an issue in the future. With a market cap of between $1.5 billion and $1.75 billion, and most software companies valued in the range of less than $3 million, Mentor executives believe that adding new companies will be a relatively simple matter. In fact, what slowed down the acquisition of Accelerated was not money; it was a question of who would run the embedded group. With that issue resolved, Henderson says the combined company can go out and spend its money to gain market share and put together more complete solutions. Its shopping time, he says. We cant do all of that organically. Electronics News March 111. 2002 Embedded: Past, Present and Future Ch-Ch-Ch in Embedded BY MICHAEL KASKOWITZ We have been building embedded systems for a quarter of a century. That time has seen immense change in the electronics world. As the term embedded systems has now found its way into common use, it is worth considering where we are heading. As the technology of electronics has evolved, the development of embedded systems has changed radically. The earliest embedded systems were very different from what we see today in two important respects the electronic design was much simpler and the software was considered an almost trivial afterthought. Typically, the same engineer who designed the board would hastily put together some code to bring it to life. In electronics terms, a modem embedded system is very different. The microprocessors are more complex and operate at higher frequencies. These devices may actually be highly integrated microcontrollers with the silicon containing much more than just the CPU. There may be multiple processors on the board, which are likely to include semi- and/or full-custom ICs, ASICs and programmable logic. Typically, a sizeable team is involved in the development of the hardware alone. Embedded system software has grown all out of proportion. The effort expended on software and, therefore, often the team size frequently exceeds that applied to the hardware. There is an increasing tendency to consider the reuse of softwareboth complete systems and individual components. This is manifesting the growing popularity of object- oriented languages, C++ being the primary example, and methodologies. It is also apparent that reusing software from other sources is becoming more widespread. There is a clear perception that the embedded software market is growing both in terms of development tools and services, and the provision of software IF. Even the well-established leaders in desktop systems, Microsoft and Sun Microsystems, view embedded as having great potential for the future. Standards are a good thing. In non-embedded contexts, the examples are easy to see. Windows has made application software more widely available because it provides a clear, defined platform upon which to build. Nonproprietary (open) standards offer more benefits. The Internet is an excellent exampleprotocols such as FTP and standards such as HTML simplify and propagate its use. Open standards have had very limited success historically in the embedded world. Often this is because the difference between embedded and desktop systems is not clearly appreciated and embedded is considered to be a minority interest. A full implementation of ANSI C++ for embedded is very difficult to achieve (even ANSI C had its problems). POSIX has been touted as a good standard applications program interface (API) for RTOSs but, despite having an embedded subset in the standard, it has failed to make an impact. A recent open standard that seems to have captured the imagination of the computing world is Linux. Even large corporations are using it or considering its use. It was almost obvious that the idea of trying to take this free software and employ it for embedded would occur to someone. Linux, like all standards, has pros and cons. In its favor it is an open standard with no royalty issues, there are a wide range of device drivers, etc., available and a good choice of support options. The downside is that it is quite large for an embedded operating system and it is not real- time. The success of Linux in this space is dependent upon the real-time problem being solved, without breaking Linux drivers or compromising the integrity of the system. The embedded software industry has gone through considerable change in recent years. Once a tiny part within the design process, embedded software has evolved into a key piece of the electronics industry. Moving forward the embedded software industry path will only grow more instrumental in system design. The only question is what standard it takes to get there. Michael Kaskowitz is general manager and VP of the Mentor Graphics Embedded Systems Division. Electronics News March 111. 2002 |
You might wish to look at the core meltdown feed below. Wednesday March 20, 2002 11:12
Embedded start-ups at crossroads
Still reeling, many IP core companies face an uncertain future BY DARRELL DUNN The embedded-processing market is undergoing a massive shakeout, forcing companies to exit the sector, shift strategies, or put up a for sale sign. Although a recovery in the chip industry is under way, a lack of venture capital and the yearlong industry slump have left many Iicensable-IP and fabless IC-core start-ups weak and vulnerable. Theres always the tendency to overfund and put too much VC money into too small a market, said Mike Gulett, chief executive of ARC International plc, at the Embedded Systems Conference here last week. Its happened over and over again, and that contributes to a natural shakeout. This recession has also hit everybody hard, and a lot of the smaller guys arent going to make it through. A start-up company right on the edge cant afford 18 months of recession. They have to get acquired or change their business model. Gulett became chief executive of the San Jose supplier of RISC and DSP processor cores in December, following a housecleaning of the companys management staff. Gulett said he and several other senior-level executives have been brought in to ARC to address an unfulfilled business opportunity, but with $175 million in cash still on hand, the company can weather any storm that comes up. Other start-ups in the processor- core and IP market may not be as fortunate. Several companies are struggling to keep their efforts afloat, analysts and industry insiders said last week. The shakeout has already started for a variety of reasons. In the whole IP world, there are not a whole lot of exit strategies, said Eric Mack, director of strategic marketing at 3DSP Corp., Irvine, Calif. You go public, you get bought, or you fade away. Going public is the difficult path, and you really cant do that on licensing alone. You cant get to a $100 mil- lion a year on licensing fees, or at least it takes a long rime. Changes under way Earlier this year, Lexra Inc., a supplier of MIPS-based clone processor cores, and picoThubo Inc., a supplier of ARM-based processor clones, were forced to abandon their efforts when they lost successive long standing court baffles with MIPS Technologies Inc. and ARM Ltd. Upon settling its suit with MIPS, Lexra said it would pursue plans to become a fabless semiconductor company At the time, the company said it had to revamp its strategy anyway after losing three licensing agreements with major telecom companies. Plans at picoTurbo, which had to cease all sales and marketing activities and transfer its product designs and IP assets to ARM, are unclear. Chameleon Systems Inc., a San Jose supplier of reconfigurable DSP chips, last month changed its management, cut 50% of its staff, and admitted it had to rework its architecture. Having retreated to the development lab after its Reconfigurable Communications Processor (RCP) failed to find a market in wireless infrastructure, Chameleon plans to reemerge next year sporting a new product line and a somewhat different focus. According to Peter Peist, recently appointed chief executive, the original RCP is now a history lesson, but its engine a powerful streaming-data processor will be at the heart of a new thrust into areas such as image processing and voice-over-IP While the company hasnt abandoned all hope of participating in the wireless basestation market, Feist said the reconfigurability that once made its product attractive for 3G applications will be considered a secondary benefit. Alchemy Semiconductor Inc. ended us run as a start-up provider of MIPS-based processor cores when Advanced Micro Devices Inc. acquired it for $50 million in cash. Phil Pompa, a founder and executive at Alchemy who now serves m vice president of marketing at AMDs Personal Connectivity Solutions unit, said acquisition was the best outcome for Alchemys investors and employees. It was the best way to get payback, and it was certainly something we needed to do, Pompa said. If you talk to investment bankers about start-ups that have what they call a positive exit, either through an acquisition or an IPO, 75% to 80% is through an acquisition. The reality in the semiconductor business, even in the best of times, is the chances are greater that you will be acquired than make it to an IPO. VCs taking closer look Venture capitalists are scrutinizing which start-ups they will continue to support m the wake of the industrys unprecedented downturn, he said. Investors are weighing how much cash remains on hand, at what rate it is spending cash, and how long it will take the company to move into a positive revenue flow. "A lot can depend on where the company is in the funding cycle, Pompa said. If theyre looking for a third round, they need to be hitting that magic number of about $2 million a quarter in revenue. Those companies that are still pre-revenue and are looking for more money are finding it hard to attract more capital. One company being watched closely is BOPS Inc., a Mountain View, Calif., supplier of DSP IP that has been talking to suitors about a possible acquisition, according to analysts. Carl Schlachte, chairman and chief executive of BOPS, last week denied the company is pursuing acquisition, but admitted that were interested in talking. Like any other start-up, if the right offer comes along, you have to listen. Its no big secret that the last year was difficult. The drying up of capital forces you to look at alternatives. BOPS, like many companies, suffered through a drop in revenue in 2001, and a number of expected licensing agreements were pushed out, Schlachte said. We believe were in position for substantial growth in 2002, but cash is king and you dont say no to the right offer. BOPS is similar to a number of start- ups that were getting very close to making it over the hump, and then the 9/11 incident and the whole state of the economy resulted in their VC money running out and design wins being put on hold, said Markus Levy, an analyst at Micro- Design Resources Inc., Sebastopol, Calif. BOPS had done a lot of things right and had some momentum, and really, theyre not dead yet. But, like a number of cornpanies, theyre having to regroup and try to figure a way out of this. Some start-ups have been tweaking their operational models, such as ARC, which has gone from being primarily a licenser of stand-alone processor cores to a system-level solutions provider through a series of acquisitions. The same is true of DSP specialist 3DSP, which is now a fabless IC provider instead of a pure IP company. I think thats the right way to go for anyone, said Will Strauss, an analyst at Forward Concepts Co., Tempe, Ariz. "The pure IP model just doesnt have enough return to excite the VCs when theyve got to wait three years for a royalty stream. The shakeout in the IP and processor core market is still in the early stage, observers said, and exactly what the landscape will be in a few years is debatable. ARCs Gulett said as few as three or four of the start-ups will survive long term. AMDs Pompa believes fewer than 50% will be able to keep their business models intact. "Attitudes have changed. The market and funding environments have changed. And consequently, the business plans have to change as well, Strauss said. Were entering a second stage, and perhaps its the darkest-be- fore-the-dawn stage. We see signs of recovery, but some companies are not going to be able to hang on that long. Many VCs are becoming very hesitant to throw good money after bad. EBN March 18, 2002 |
The NEC µPD720100A USB 2.0 will be superceded with USB 2.0 chips on the motherboard chipsets. Wednesday March 20, 2002 10:37
Radar 3.18.02
INTEL TO UNVEIL UPDATED GRAPHICS CHIPSET FOR PCs Intel Corp., Santa Clara, Calif., in May will launch a revamped version of its integrated graphics chipset (IGC), according to sources with knowledge of the companys product roadmap. At a time when graphics IC heavyweights ATI Technologies Inc. and Nvidia Corp. are training their own IGCs on the high end of the PC market, Intel is planning to direct its device at the low-end PC segment, where chipset makers such as Acer Laboratories Inc. and Via Technologies Inc. have been gaining market share. The 845G integrated chipset, code-named the Brookdale-G, will feature a unified memory architecture and is expected to include a USB 2.0 controller. Bruce Gain EBN March 18, 2002 |
Archimedes C appears to be out of business.
It's good to have the source code for the development tools to avoid unpleasantness of having a vendor go out of business.
American Raisonnance has an interesting 80C52 website. Saturday March 16, 2002 12:55
With the demise of this venerable name in the embedded market you may be looking for a new source for 8051 embedded tools to support your applications. To help you out with your decision process American Raisonance is offering a 10% discount of current V5 or V6 users of Archimedes tools. |
SMC, of course, like Cygnal and Cypress an maybe Mentor needs a development board and associated inexpensive real-time operating system with source code availability.
Note that with USB 2.0 the data transfer needs to be done in hardware because of the speed. Also the bitness [ 8, 16, 32, 64 etc.] is unimportant because the processor is merely responding to interrupts and setting buffer addresses and other menial tasks.
Forth only runs at about 10% as fast as straight assembler-produced machine code. But this is not a problem. The reason is that the operating system and background tasks are running at the lowest priority. When an interrupt arrives, control is immediately transferred to an assembler-produced interrupt service routine.
BTW, Microsoft Visual C/C++ only runs about 1/4 the speed of an in-line assembler-produced ring 0 similar program, we've found out from experience.
However, an interactive real-time operating system with integral incremental compiler and assemble is essential to get the software done quickly, inexpensively, and to trouble shoot problems interactively over Internet using Netmeeting or Messenger.
Messages and dead ends
Kim KOMANDO
For the Journal[T]heres a horse coming up on the outside, though Microsofts Windows Messenger (messenger.msn.com). This Windows Messenger is easy to setup. And it handles voice and video, in addition to written messages. Web cams, which are inexpensive, can be used to transmit the video image. Users can talk for hours over the Internet, without incurring long- distance charges. Windows Messenger requires Windows XP. Older versions of Windows can use MSN Messenger (messenger.msn.com/default.asp?client=1). It does not include video.
If you have not made the jump to XP, you might have NetMeeting on your computer. NetMeeting also supports video. It requires you to find an Internet Locator Service (ILS) server. That server is used to hook users up with one another.
With the advent of Windows Messenger, I expect NetMeeting to be phased out. ...
Business Outlook, Albuquerque Journal Thursday March 14, 2002
Csd pages show netmeeting examples were on Excite which is gone. These pages may be restored when Comcast makes web space available.
The important part of both netmeeting and messenger is that an engineer can log on to an 80C52 family microcontroller connected to internet to troubleshoot hardware/software problems. The post a hopefully-software fix! Hardware problems are lots more difficult and expensive to correct! Thursday March 14, 2002 10:40
USB 2.0 peripheral device controller added by
SMSC
HAUPPAGE, N.Y. Nearly four years after it began shipping USB 1.1 peripheral controller chips, Standard Micro systems Corp. (SMSC) has introduced two USB 2.0 controllers, the USB97C2O1 True Speed" ATA/ATAPI/CF bridge controller for external disk drives. and the USB-97C210 Memory Card Controller for USB memory card readers. Many of the USB 1.1 controllers were shipped under customers brand names, but the new devices will carry an SMSC label. Marshall Lacoff, vice president and general manager of SMSCs Embedded Products Group, said the U5B97C201 USB 2.0 ATA/ATAPI/CF controller chip, which is sampling now, includes UMDA technology from the ATA/66 interface and supports UMDA modes 1 through 4. Thus, is can transfer data packets to and from ATA/66 and ATA 100 disk drives at the maximum sustainable USB 2.0 burst data rate of 53 Mbytes/ second. The chip is based on a patent-pending programmable architecture that, together with process technology, minimizes its die size, yet it can support ATA-6 drives up to 2,048 Gbytes and implement hard drive password security features. The U5B97C201 combines a PHY with an 8051 microprocessor, and is designed so as to keep the MPU 8051 out of the data path between USB and IDE buses. As a result, data bytes flow internally at 60 Mbytes/second. This design also uses a 1.25-kbyte on-chip data buffer that is smaller than that on other USB 2.0 ATA/Atapi solutions. It comes in a 12 x 12 mm, 100-pin TQFP and costs $4.95 in 100,000 quantities. Designed for multi-function card readers, SMSCs U5B97C210 USB 2.0 memory card controller consists of a USB 2.0 physical layer transceiver (PHY) and serial interface engine (SIE), data buffers, an 8051 microprocessor with expanded scratchpad and program SRAM, and dedicated controllers for each memory card type. It interfaces to Compact-Flash (CF), SmartMedia(SM) Memory Stick (MS), Secure Digital (SD) and MultiMedia-Card (MMC) memory cards, as well as to ATAMicrodrives. It is packaged in a 128-pin TQFP and priced at $7.50 in 100,000 quantities. Call (631) 435-6000 www.smsc.com Electronic Engineering Times March 11, 2001 |
Here's a jpg of the windows 2000 screen showing the Numega DriverStudio monitor and some of the icons csd uses to do its work.
Csd's icon for cdpci1.exe, which is in the process of getting its local bus accesses fixed, comes from the cover of Embedded Controller Forth for the 8051 family. 1
Csd is continuing to experiment with both the NEC µPD720100A and Cypress 3671.
There apparently is a bug in the local pci/cardbus experimental software csd is trying to correct.
This software is no longer posted since Excite removed the files and Comcast apparently will not have web space available until later.
Compuware/Numega's ring 0 monitor prints the Cypress progress message.
The selection below is printed when one clicks on Hold on the Cypress control panel, then clicks Run.
Note the exuberance "yahoooo". This is very complicated software and it's suprising when things work! Wednesday February 13, 2002 10:17
224.528445 Default URB status Ezusb.SYS: Enter
Ezusb_Create() 224.528547 Default Ezusb.SYS: IRP_MJ_DEVICE_CONTROL 224.528599 Default Ezusb.SYS: Enter Ezusb_VendorRequest - yahoooo 224.528654 Default Ezusb.SYS: enter Ezusb_CallUSBD 224.528704 Default Ezusb.SYS: Calling USB Driver Stack 224.529103 Default Ezusb.SYS: return from IoCallDriver USBD 103 224.529153 Default Ezusb.SYS: Wait for single object 224.532858 Default Ezusb.SYS: Wait for single object, returned 0 224.532915 Default Ezusb.SYS: URB status = 0 status = 0 irp status 0 224.532970 Default Ezusb.SYS: exit Ezusb_CallUSBD (0) 224.533031 Default Ezusb.SYS: Vendor Request returned 0 bytes 224.533153 Default Ezusb.SYS: Enter Ezusb_Close() 226.279222 Default Ezusb.SYS: Enter Ezusb_Create() 226.279322 Default Ezusb.SYS: IRP_MJ_DEVICE_CONTROL 226.279373 Default Ezusb.SYS: Enter Ezusb_VendorRequest - yahoooo 226.279428 Default Ezusb.SYS: enter Ezusb_CallUSBD 226.279478 Default Ezusb.SYS: Calling USB Driver Stack 226.279873 Default Ezusb.SYS: return from IoCallDriver USBD 103 226.279924 Default Ezusb.SYS: Wait for single object 226.283713 Default Ezusb.SYS: Wait for single object, returned 0 226.283771 Default Ezusb.SYS: URB status = 0 status = 0 irp status 0 226.283826 Default Ezusb.SYS: exit Ezusb_CallUSBD (0) 226.283886 Default Ezusb.SYS: Vendor Request returned 0 bytes 226.284002 Default Ezusb.SYS: Enter Ezusb_Close() monitor -- end -- |
Here's an ATMEL TCP/IP 80C52 with flash like the Cygnal. And another evaluation board!
Keil is selling software for the ATMEL part.
Csd does not believe in NIH [not invented here]. Unnecessary creativity is, in reality, unintelligent.
Design goals should be set. Then figure-out how to meet the design goals the easiest way possible.
Here's a jpg of the original Sandia seismic data authenticator motherboard.
Design goals included not using any of the 8051 hardware features since they might be needed in the app.
The Keil software monitor uses one of the serial ports on the CY 3671. This is probably a good idea now. Monday March 11, 2002 08:44
There are or were a several commercial board built from the schematics in Embedded Controller Forth.
Below is another. Tuesday February 12, 2002 06:42
Here's the original Sandia seismic data authenticator motherboard. The width of the board could not change since it had to fit in a tube. But the length could be extended, if necessary. Experience tells us that it is best to keep the digital parts as close together as possible. And the analog parts shielded! The Cypress CY3671 board is seen above.Thank goodness someone else [Cypress] upgraded this hardware! Let's try to find out what Mentor graphics did. Princeton or Harvard architecture?
The Princeton or von Neumann architecture on this board is implemented with a 'HC08 AND gate. 1 2 3 4 See U4 in Sandia tech report. 4 Note the use of the PC connector for add-on electronics. One should be very careful to shield analog electronics from the 8051 family. This is easy to do with the wide spacing between the expansion cards. 8051s are electrically noisy. You get black lines on your, and perhaps you neighbors', TV and you cb radio doesn't work very well when an unshielded 8051 is running. New hardware development is no longer necessary. Cypress built about an ultimate 80C52 Forth motherboard with its 3671 and 3681 EZ-USB-FZ/2 development systems! Casual tests indicate that the Cypress CY7C64313 is fairly quiet electrically. |
Csd contacted Mentor at 800 547 3000 to get details on the development board.
One might have to buy the IP core to get the development board!
And look at the speed of the Mentor part!
Fastest 8051 Soft Core available The M8051E-Warp is the only core that can surpass 50mips performance-reliably. Unlike other 8051 cores, the M8051E-Warp has a frequency-scalable state machine and is silicon-proven at 120MHz on 0.18µ process technology.
Cypress may have some real problems.
Contrary to the Cypress EZ-USB-FX designation, the designation COMPLEX-USB-FX would likely be more descriptive.
But not to worry once 80C52 Forth is working on the Cypress development board! Friday March 8, 2002 09:38
Dual-Role IP Core Meets Full-Speed USB 2.0
The industry's first dual-role U 2.0 controller macrocell is now available from Mentor Graphics. It complies with both the current USB 2.0 standard for full-speed function and the new On-The-Go supplement for point-to-point communication. The Inventra MUSBFDRC synthesizable macrocell supports host a peripheral functions in point-to-point data transfers. With it, designers can include a single core for both peripheral and host applications, configuring the core as required. The On-The-Go supplement gives portable devices a cost-effective means of conducting point-to-point communications using USB protocols. It lets portables connect directly to a peripheral device for point-to-point file exchange without a host PC. The soft IP core and an evaluation board are both available. The board shows how the core can be configures as a standard USB peripheral within a mass-storage device when connected a PC. It also demonstrates file transfer from a PC to a peripheral over On-The-Go connections, as well as data transfers in the form of print files when configured as a host. The core complies with the USB standard for full-speed (12-Mbit/s) functions. It supports session-control and host-negotiation protocols. And, it's configurable for up to 15 addition transmit endpoints and 15 addition receive endpoints. Also included is PVCI-compatible CPU interface. P AMBA-AHB bridge is optional. USB device-layer firmware is provided to facilitate peripheral application code development. The firmware gives designers "hooks" for application-specific software, such as a host stack, and handles device configuration and control management. It's delivered in C in embedded development tools an compiles from 10 to 20 kbytes. The Inventra MUSBFDRXC soft IP core and evaluation board will be available in the second quarter. A reference design for On-The-Go configuration will be available early in the first quarter for developers. Contact Mentor for pricing and delivery information. Mentor Graphics Corp., (503) 682 7000; www.mentor.com/inventra. David Maliniak Electronic Design March 4, 2002 |
Jim Turley comments on core meltdowns and embedded processors are posted at this website.
Turley wrote The Death of Hardware Engineering 1.
Turley failed to mention including hidden features as a disadvantage to this approach. Like having your car's computer announce to you,
"You've driven 200,000 miles.
Time for you to buy .
I'm giving you 50 more miles to drive me to a junk yard.
If you drive more than 50 miles, then I will destruct."
This is call "spiking" a chip. Here's a real world example. 1
Turley also failed to mention all of the software bugs which cause endless chip revisions.
These cause a hardware configuration management nightmare.
Better to do it in sofware, like the Cypress ez-sub-fx, and let users download the most recent revision over internet.
Csd likes the idea of source code.
Loading binaries on the Cypress ez-sub-fx is not a good approach. And, of course, csd is working on a solution to permit the 80C52 to assemble and compile source code! Thursday March 7, 2002 10:37
Csd got a response out of Cygnal for source code to update the flash in the adapter which goes between the PC and its development board.
Not good.
Cypress has been generally excellent about providing most, if not all, source code for operating its development boards. But here may be a problem here. Loader 80C52 source code. Monday February 25, 2002 20:34
Subject: RE: Cygnal did not mention the source code
availability of ECReset yet. Date: Wednesday, February 06, 2002 9:42 AM Hello Bill, I cannot provide you with the source code for ECReset. If you are interested in programming devices through he JTAG interface please refer to Application Note 5 available on the Applications webpage at www.cygnal.com. Best Regards, |
Here's a new link on Microsoft
problems 1. Monday February 25, 2002 17:09
Disaster struck on Saturday evening.
Csd converted from Excite@home to Comcast.
AOLPress crashed with Comcast's system!
AOLPress works fined under New Mexico Online and Excite@home.
AOLPress is used by csd to help build its web pages. Along with raw html coding.
Webcommando rate AOLPress as one of the recommended html editors. AOLPress, in csd's experience is What You See is Close to What You Get. It's got bugs but is a very nice software package.
Csd downloaded both arachnophelia and Macromedia Homesite. After experimenting the Homesite, csd uninstalled it.
Paul Lutus is author of AppleWriter and GraForth 1.
Lutus has scathing comments on Windows XP!!!
Arachnophelia fairly slow to build web pages with. So Csd is using a combination of AOLPress and arachnophelia until Comcast get's its problems solved. Or csd decided to do something else.
Comcast's decision to try to redo software what Excite@home successfully did may be a mistake. Csd emphasizes again that software costs are easy to underestimate for the reason that experience show that only about 20% of the cost has been expended when one thinks the software is working.
Csd is following Comcast, Excite, and now Quest. Quest is involved in Global Crossing. And New Mexico senator Jeff Bingaman's wife, Ann, made an impressive $2.5 million for 6 months of work.
Honestly, of course.
The cost of getting USB 2.0 working right is great. Not only on the PC driver side but the peripheral side too.
But, of course, csd is working to reduce 80C52 software costs on the Cypress EZ-USB-fX development boards.
While MCII emphasizes mass market USB products, there will be industrial USB applications which will require a super-fast USB 2.0 miniport driver which will not be required to be interoperable. Friday February 22, 2002 07:40
COMMENT Test Strategies Drive USB Product Success THE SUCCESSFUL ROLLOUT of the high-speed version of USB is dependent on three key points the availability of enabling silicon, development of firmware and driver software, and easy access to compliance testing. Following the publication of the USB 2.0 specification in the spring of 2001, silicon controllers and software support both emerged by midsummer, but testing badly lagged behind until late last fall. The less-glamorous cousin to silicon and software, testing is not an optional formality in the product-introduction process. In the beginning, compliance testing for USB devices was a relatively cursory process of simply plugging-in to see if the device enumerated. Since then, testing has evolved into a robust compliance program with detailed test matrices and multiple tests, including all-important interoperability tests. Todays USB test procedures were developed by the Universal Serial Bus Implementers Forum (USB-IF), with help from Intel Corp. and other companies. The procedures check a devices electrical characteristics, verify the operational characteristics of the device, and test general interoperability between the device being tested and other known-good devices, hubs and hosts. Based on real-world experience, the tests identify real problems correlated to device design quality. The USB -IF, a nonprofit corporation founded by the group companies that developed the USB specification, requires that all USB products must be completely interoperable, functioning together without interference. Each class of USB productlow-speed, full-speed and high-speed must pass specific testing to prove they can function together in the same USB tree. Full-speed devices must be able to work properly in a high-speed environment, and so on. And for the newer high- speed devices associated with USB 2.0, compliance testing ultimately deems whether a product is certified to carry the USB -IF Hi-Speed USB logo. Therefore, well-executed testing strategies are nothing less than the cornerstone of a products ultimate success in the marketplace. However, until the recent certification of independent test facilities, the only venue outside of Intel for USB testing were USB compliance workshops, or plugfests. Plugfests are indeed excellent venues for networking with other US vendors and with the USB-IF compatibility team, and they provide exceptional visibility to potential customers. But the inconvenience of testing at Plugfests can sometimes be contrary to the importance of testing and to the need to test products early in the development cycle and often multiple times before going to market. Formal compliance testing is usually the best way to uncover silicon, firmware and driver problems, all of which can be long lead items to fix. Because of the cost of design changes, experienced engineers plan to formally test their products several times during the development cycle: once early in prototype development, once prior to committing to mass production and once to verify the final product. If problems are found at any stage, additional tests may be needed to verify product corrections. To help vendors get USB products to market more quickly, the USB-IF has authorized the opening of high-speed USB testing labs that charge on the order of $5,000 to run a product through the rigorous battery of tests to determine adherence to the specification. That $5,000 to test a product before it ships is an incredibly sensible investment far cheaper than a recall and easier than diagnosing a problem remotely, or worst of all, having to fly an engineer to a customer site. But it is the time-to-market issues that dominate the need for independent USB testing labs. A good, independent USB testing lab can turn around device-, product- and system-level testing in less than two weeks. More importantly, the best independent labs often bring USB design expertise to the table, providing client companies technical counsel down to the silicon level if necessary. With the advent of independent labs, USB product makers can finally embrace test strategies that are timely, confidential and that take place at test facilities that add real value to the process. This will finally make testing a strong third leg, equal to silicon and software in its ability to quickly and effectively drive new USB products to market. Terry Moore is the president and founder of MCCI of Ithaca, N. Y; the company provides USB-IF logo compliance testing as part of its services. Electronic News February 18, 2002 www.electronicnews.com |
You may wish to read Windows XP: Why Bother 1.
And Sifting Through XPs Built-In Tools 2.
Both the Milcoch and Cassel articles are posted at this website.
Bill went to CompUSA yesterday. Lot's of Adaptec USB 2.0 pci bus adapters now in stock!
It took the SIIG adapter to convince Csd that the problem is in Windows 2000. Thursday February 21, 2002 08:15
Here's good evolutionary history of the 8051 family
related to USB.
The Anchor/Cypress parts are very complicated from a software standpoint. This means big dollars in software development.
Csd, however, sees a way to decrease software development time on both the PC and 80C52 sides.
And have peripheral 80C52 software which is on an order of magnitude more reliable in the field than software produced by other technologies.
One of the more important principles in practical design is that digitization should occur at the sensor.
If one wants to get the same values independent of the computer/microcontroller reading the digital sensor data, then, as a practical consideration, digitizing must be done at the sensor.
And it is very important to have quiet regulated power at the a/d converter.
Proper board layout is critical too. A/d converter manufacturers have app engineers who help you with board design.
Then, too, it is important to shield the a/d converter from electrical noise.
Digital data is best transmitted to the computer over fiber optic links. These are immune to electrical noise.
Electrical noise can be so bad as to cause an analog mux to skip channels. Believe us!
Speed of communication with fiber optics is not limited by cable size. Clerk Maxwell 1 2 results applies to wires not fiber optics.
All of this means that USB, especially USB 2.0, is very important in a PC analog measurement environment.
Csd has been involved with a/d converters on ISA bus boards.
Not a good idea if one want to get the same values out of the sensor independent of the computer system. The same applies of a/d converters on pci bus boards.
The 80C52 family is ideal for synchronous serial mode 0 reading values from a/d serial output converters. Wednesday February 20, 2002 07:55
Intel/Cypress 8x931 and 8x930
The 8x93 1 and 8x930 are full-featured chips that are compatible with the 8051 microcontroller family. The 8x93 1 is an 8051 with an embedded USB controller. The 8x930 is compatible with Intels enhanced, fast 8051, the 80251. Intel was the first to release peripheral chips with USB controllers, but is now focusing on the PC side of USB and has granted Cypress the license to their USB chips. Both chips have full USB capabilities in a microcontroller thats powerful enough to control complex tasks. Some versions have an embedded hub that can connect to additional USB devices. These chips have two advantages: many system designers and programmers are already familiar with the 8051 family, and language compilers and other development tools are available. The architecture and programming are compatible with other 805 is and 8025 is, except for the USB-specific registers and instructions. If youre not familiar with the chips, you will have a learning curve, as the architecture is complex. The 8x930 has over 300 machine-code instructions. CPU Architecture The 8x93 1 is based on the 8xC5 1 FX, which in turn is based on Intels 8051, the original member of an early but still popular microcontroller family. The chip can access up to 64 kilobytes each of external data and program memory and has 32 general-purpose I/O bits, three timers, an asynchronous serial port, and a keyboard interface. The 8x930 is based on Intels 80251, which has the same general architecture as the 8x93 1. But the 8x930 is much faster and has more RAM, the ability to access more external memory, and more USB endpoints. The 8x930 is faster than the 8051 for two reasons. It has a redesigned core that takes from two to ten clock cycles per instruction, compared to the 8051s twelve. Also, the 80251 has added hardware and 62 new machine-code instructions that speed program execution. For example, a second general-purpose register enables faster data transfers, and page mode enables fast data access within a 256-byte block. Also, code fetches are 16-bit instead of 8, and some internal data transfers are 16- or 32-bit rather than 8-bit. To remain as compatible as possible with the 8051, the 80251 supports two modes: source and binary. A configuration bit determines which mode the chip uses. Binary mode is compatible with the 8051, while source mode is faster on devices that make liberal use of the new 80251 features. In binary mode, each of the new instructions has a prefix of AS, which is the 8051s single unused machine code. In source mode, the new instructions dont use the prefix, but 159 of the original instructions do require it. The USB Controller Both the 8x930A and 8x93lA contain a peripheral-side USB port. The 8x930H and 8x931H add hub circuits that enable attaching downstream devices. The chips have special-function registers (SFRs) that store received data and data to transmit as well as status and control information. The transmit and receive buffers can each hold two sets of data. The chips can also be configured to automatically reread or retransmit data after an error has been detected. Of the 8x93 ls three endpoints, Endpoint 0 supports control transfers only, Endpoint 1 supports all four transfer types, and Endpoint 2 supports all but isochronous transfers. On the 8x930, Endpoint 0 supports control transfers, endpoints 1 through 4 support all types, and endpoint 5 supports all but isochronous transfers. To transmit data, the firmware writes the data to a transmit buffer and stores the number of bytes written in the TXCNTL register. This tells the USB controller how many bytes are available to transmit when the host polls the device. In a similar way, the USB controller stores received data in a receive buffer and generates an interrupt to let the firmware know that the data is available. Intel provides source and object code for a host driver that handles enumeration and bulk transfers with the chips, as well as sample firmware. Cypress Semiconductor EZ-USB Cypress EZ-USB is another 8051-compatible family of chips with USB capability. These chips uses a different approach to storing firmware. Rather than storing the firmware on-chip, an EZ-USB can store its firmware on the host, which loads it into the chip on each power-up or attachment. The EZ-USB was originally a product of Anchor Chips, which began as an independent company but in 1999 became a business unit of Cypress Semiconductor. Having the firmware stored on the host has pluses and minuses. The obvious advantageand its a big oneis easy updates to firmware. To update the firmware, just store it on the host and the driver will send the firmware to the device on the next power-up or attachment. Theres no need to replace the chip or use a special programmer. The downsides are increased driver complexity, the need to have the firm- ware available on the host, and longer enumeration time. Cypress helps with the driver by providing complete source and executable code for a driver that handles the downloading of the firmware, enumeration of the device with the new firmware, and other data transfers of all types. You can use the supplied driver as-is, or use it as the base for a custom driver. If you dont want to download from the host, the chip also supports storing its firmware in an external serial EEPROM. On power-up, the code loads into the EZ-USBs RAM. CPU Architecture The EZ-USBs architecture is similar to Dallas Semiconductors DS80C320, which is another 8051 with a redesigned core. The chip uses four clock cycles per instruction cycle, compared to the 8051s twelve. Each instruction takes between one and five instruction cycles. The CPU is clocked at 24 Megahertz. On average, the EZ-USB is 2.5 times as fast as an 8051 with the same clock speed. The instruction set is compatible with the 8051s. All of the chips 8-kilo- bytes of combined code and data memory is RAM; there is no non-volatile memory. However, the chip does support non-volatile storage in its 12C serial interface that can read and write to serial EEPROM. There are 32 general-purpose I/O pins. USB Controller The chip supports the maximum number of endpoints: one control end- point, plus 30 additional endpoints and all four transfer types. The enumeration process for this chip is unique. As with any USB device, when the host detects the EZ-USB, it will attempt to enumerate it. But how can it enumerate the device when it has no stored firmware? The answer is that the chips core includes circuits that know how to respond to the hosts enumeration requests. The circuits also respond to the vendor-specific request AnchorLoad, which receives and stores the downloaded firmware. These circuits are independent from the 8051 core. They communicate with the host while holding the normal 8051 circuits in the reset state. On power-up, before enumeration, the core circuits attempt to read a byte from a serial EEPROM on the 12C interface. The results tell the core what to do next: use the default mode, identify the device from EEPROM bytes, or load firmware from EEPROM. Default Mode for Basic Testing. If the core detects no EEPROM, or if the first byte read is not BOh or B2h, the core reads no data from the EEPROM. This is the most basic mode of operation. When the host enumerates the device, the chips core circuits respond to requests. During this time, the 8051 is held in the reset state. This reset state is controlled by a register bit in the chip. The host can write directly to this bit to place the chip in and out of reset. This reset affects the 8051 circuits; its unrelated to the reset state of the USB interface. The descriptors retrieved by the host identify the device as a Default Anchor Device. This instructs the host to use Cypress General Purpose Driver to communicate with the chip. The driver uses the AnchorLoad request to download firmware (provided by Cypress) to the device. The firmware contains a new set of descriptors and code that enables transferring data using all four transfer types. On completing the download, the 8051 exits the reset state and is configured so that the firmware, rather than the core circuits, responds to requests. On exiting reset, the firmware causes the chip to electrically simulate a disconnect and reattachment to the bus. When the host detects the simulated re-attachment, it enumerates the device again, retrieving the new descriptors. Cypress has trademarked the term ReNumeration to describe this process. This mode of operation is intended for use in debugging. You can use it to get the chip up and running and to transfer data without having to write any drivers or firmware. Identify the Device from EEPROM Bytes. The core can also read identifying bytes from the EEPROM on power-up, and provide this information to the host during enumeration. If the first value read from the EEPROM is B0h, the core reads EEPROM bytes containing the chips product, version, and device IDs. When the host enumerates the device the first time, it uses these bytes to find a matching INF file that identifies the driver that specifies which firmware to download on re-enumerating. Load Firmware from EEPROM. A third mode of operation provides a way for the chip to store its own firmware. If the first byte read from the EEPROM is B2H, the core loads the LEPROMs entire contents into RAM. This must include the Vendor and Product ID bytes as well as all descriptors required for enumeration and whatever other code and data the device requires for USB communications. When the chip exits the reset state, it has everything it needs for USB communications. When the host enumerates the device, it reads the stored descriptors and loads the appropriate driver. There is no re-enumeration. For a quick start to EZ-USB designs, J. Gordon Electronic Design (http://www.jged.com) has an inexpensive business-card-sized SimmStick with an EZ-USB and related circuits. John Hydes excellent book USB Design by Example includes assembly code and circuits for many projects with this chip. Johns website at http://www.usb-by-example.com has more information and code examples.
USB Complete Everything You Need to Develop Custom Peripherals No custom dirivers needed - use the Win32 API and Visual Basic author of Parallel Port Complete and Serial Port Complete |
Here's a Motorola Coldfire development system in the
foreground and the Cypress EZ-USB-FX system the the background.
Coldfire is used in about all H-P printer. And lots of other printers too. There's probably a 8051 in your PC keyboard.
The question, of course, is how best to develop software on these 'peripheral' processors.
Csd advocates use of an operating system. Especially an operating system which hosts its own incremental compiler and assembler.
These two systems represent very different philosophys.
The Coldfire emphasizes lots of compute horsepower at the peripheral end.
The cypress emphasizes USB and DMA high speed hardware engines with a hardware traffic cop microcontroller. Where do you want to do your computations. In the PC or in the peripheral? Monday February 18, 2002 08:38
Here's a Motorola Coldfire MCF6902FT25 Arenwsh / P&E development system. Csd hasn't tried this system it bought in 1997 yet! Look at the old 54 pin DRAM module. Obsolete! High tech frailty! What both these systems have in common is that they obviate need for an expensive development system. And there looks to be several Forths for the Coldfire family 1 2. However, the Coldfire is capable of supporting operating systems like linux. |
The statement
The problem looks to be with the NEC chip decoding.
appears to be incorrect.
The function number can be any value betwee 0 to 15. A pci/cardbus device is supposed to do a full decode.
Windows decides the addresses of each device.
The 8 bit was decoded okay in the NEC µPD720100A.
So perhaps the problem is in Windows 2000?
And will be fixed first in XP! For business reasons, of course.
Csd thinks it knows what wrong with the pci/carbus local register Reads and Write. Now to attempt a fix. Wednesday February 13, 2002 06:27
Csd was curious about the apparent decode problem in the Adaptec USB 2.0 board.
What happened is that the apparent same hardware was being detected at muliple function numbers. This confused Windows 2000 plug and play which doggedly reinstalled a driver for the apparent second set of hardware.
So csd bought a SIIG USB 2.0 board for about $40. All seem to use the NEC µPD720100A pci to usb 2.0 chip.
The SIIG card come with a 3.5 diskette containing drivers.
Here's what happened when csd tried to install the drivers
Windows 2000 'blue screened.' Even in the Safe mode!
After manually removing the .sys files and .inf file, csd downloaded a new driver at the SIIG web site. The new drivers partially installed.
Csd enumerated the pci bus to locate the SIIG card bus, device, and function. Here's what happened.
This looks just like the Adaptec card! The problem looks to be with the NEC chip decoding.
Here's memory map for the functions
6 E801C000 256 bytes
2 E801C000 256 bytes
0
E801A000 4096 bytes
1 E801B000 4096 bytes
4 E801A000 4096 bytes
5
E801B000 4096 bytes
It appears that the NEC µPD720100A isn't decoding the 4 bit of the function number! Hmm.
Csd hasn't done a USB project before.
We're curious what engineers and programmers really do compared to what the protocol specifiers say they should do.
There is a possibility that some of software writers, such as for the FujiFilm digital camera, are bypassing the higher-level microsoft drivers and going directly to hardware.
It's becomming increasing apparent that USB 2.0 has a bit to go.
Csd is studying the Compuware/Numega USB codes. These examples are closely related the the Cypress codes. Here's the wdm examples diretory
We can see how UBS dominates focus.
Csd, of course, wants to connect UBS to a Forth operating system running on the Cypress 3671 and 3681 development systems. Tuesday February 12, 2002 09:00
USB 2.0 will work first on XP. With Microsoft's usb 2.0 drivers, of course.
CompuUSA has on its shelves three
brands of pci to usb 2.0 cards
10 Belkin
28 SIIG
9 CompUSA
in Albuquerque. 47 units total on the shelf!
Note that Adaptec card was not among these.
In the real world what pci/cardbus designers do versus what they should do are two different things. You have figure-out what the designers, in fact, put into their products by experimentation.
Csd speculates that the Adaptec USB 2.0 pci card may have failed Microsoft's quality tests for the reason that it apparently didn't completely decode pci functions. This caused Windows to think that the card contained more functions than, in fact, it did.
All use the NEC µPD720100A pci to usb 2.0 chip. This gives you 5 usb 2.0 ports with any one of these pci cards.
All cards closely resemble the NEC demo board. Demo or development system boards with schematics, of course, hold down hardware design costs.
One borrows the circuit design from these boards.
But that's what the chip manufacturers want. They are helping, to those who don't want to reinvent, build good hardware.
NIH - not invented here - is bad.
The NEC pci vendor number is 1033. The Adaptec vendor number is 9004. 1
Csd sees this on the Adaptec card. There appears to be 6 pci function numbers. But this isn't the case.
One space has 256 bytes.
Two have 4096 bytes.
Here's the enumeration using the code at 1
For the Adaptec card with the NEC USB 2.0 chip, the card appears to have 5 function 0, 1, 2, 4, 5, and 6.
Two of the functions apparently have 4096 bytes of memory. The second function starts at E801B000.
The apparent same function 2 and 6 appear in device manager as
and occupy 256 bytes of memory.
Here's what Adaptec warned us about.
No device drivers are included with the Adaptec card.
Windows 2000 has problems installing device drivers. Multiple drivers were installed for the two 4096 bytes functions. Csd deleted these.
Here's a view of the configuration registers.
Offset 11 contains the Subsystem Device ID and Vendor Device ID. 9004 is Adaptec.
The subsystem device id is 0E. Both for NEC and Adaptec.
What does this mean? We did some searching and found.
Vendor ID: 0x1033
Short Name: NEC
Web Site: http://www.nec.com/Device ID: 0x0035
Chip Number: uPD9210FGC-7EA
Description: USB Host ControllerDevice ID: 0x00E0
Description: PCI to USB Enhanced Host ControllerVendor ID: 0x9004
Short Name: Adaptec
Web Site: http://www.adaptec.com/
Csd tried to write to the local pci bus registers. And crashed Windows 2000.
The problem may be in the device driver for the pci/cardbus. It was not totally check-out before turning the source code over to others.
Csd does this type of software for money. And csd may see a need for a miniport driver for the NEC µPD720100A connected to the Cypress 3671/3861 development systems. Friday February 8, 2002 08:14
The handheld market and ARM cores financial futures will be interesting to watch 1 2. Tuesday February 12, 2002 08:54
Core meltdowns 1 Tuesday February 12, 2002 06:54
Intel Targets Wireless
New Chips May Be Built in N.M. Journal Staff and Wire SANTA CLARA, Calif. Intel Corp. introduced a family of microprocessors Monday that promises to improve the performance and increase the battery life of - handheld devices such as cell phones and palm-sized computers. The Intel PXA250 and PXA210 processors will allow improved music, movies and games on next-generation gadgets, which are expected to be popular as wireless networks are upgraded in coming months. More than likely, the new chips will be built at Intels plant near Rio Rancho, said Intel spokesman Terry McDermott. Were talking about a new technology, and you can be sure Intels largest plant (the New Mexico plant) will be involved, McDermott said. Intel officials said they picture the handheld market growing in the coming years. We want to make this explode and be the next $100 billion market, said Mark Casey of Intels handheld-computer division. To make that happen, you need good clients and a fast wireless connection. The PXA210 will run at speeds up to 200 megahertz and will be used in cell phones, entry-level handhelds and wireless devices. The larger PXA250 will run at speeds up to 400 MHz and will be part of high-end handheld devices. Products built with the new processors are expected to be available by the middle of the year. The technology, which is called XScale, is based on the StrongARM architecture that Intel acquired as part of a settlement with Digital Equipment Corp. in 1997. Cambridge-based ARM Ltd.. licenses its chip designs to semiconductor companies, which then manufacture the processors. Albuquerque Journal Tuesday February 12, 2002 |
InteractiveWeek
ROB FIXMER: THE INTERACTIVE ENTERPRISE THE IMPORTANCE OF GREEN HARDWARE REMEMBER THE NEW ECONOMY? TURNED OUT TO BE NOTHING more than a barnacle on the butt of the old economypainful but probably inevitable. Human progress seems handcuffed to human error. At any given moment, its hard to distinguish vision from delusion. Unfortunately, the distinction is becoming more urgent because as the time between generations of technologies shrinks, so does the time available for course corrections. This situation may cause errors that do irreparable damage to our economy or social institutions before we detect the mistakes. Since the dawn of the computer age, competition in the processor market has been driven by a single priority price/performance ratios. This was not the result of industry shortsightedness. Manufacturers have long realized that power consumption could not be allowed to grow indefinitely at the exponential curve described by Moores Law. Indeed, advances in chip design between 1985 and 2001 did keep increases in power consumption to a multiple of 75 while the number of transistors grew at a multiple of 256. But that's not good enough. As documented in our Jan. 21 issue (see Page 41), even that rate, were it to continue, would send the economy and environment into a meltdown by me end of the decade. Yet, as Intels announcement of the power-sucking Pentium 4 Jan. 7 clearly showed, competition is still defined by megahertz benchmarks. Why? Because when IT pros, the biggest buyers of computer gear, try to deliver the most bang for the buck, its tempting to measure their bucks in the most advantageous lightsimple price/performance ratios isolated from associated costs such as power consumption for air conditioning equipment or ever-expanding demand for facilities space. These items often end up on some other departments bottom line or can easily be rationalized away as unavoidable costs of doing business. But Californias recent power crisis reminds us that even the digital revolution must live within its means, and utility analysts warn that delivery and distribution problems are only going to get worse. Until informed IT pros fully anticipate this reality, expand their concerns to the entire enterprise and make their revised priorities known to vendors, the microelectronics industries will have little incentive to move forward wisely. e Can high-tech companies stay in the black by going green? Let me know at rob_fixmer@ziffdavis.com. eweek February 4, 2002 80C52 POWER-SAVING MODES OF OPERATION For applications where power consumption is critical the CHMOS version provides power reduced modes of operating is a standard feature.. The power down mode in HMOS devices is no longer a standard feature and is being phased out. CHMOS Power Reduction Modes CHMOS versions have two power-reducing modes, Idle and Power Down. The input through which back-up power is supplied during these operations is VCC. Figure 27 shows the internal circuitry which implements these features. In the Idle mode (IDL = 1), the oscillator continues to run and the Interrupt, Serial Port, Timer blocks continue to be clocked, but but the clock signal is gated off to the CPU. In Power Down (PD = 1), the oscillator is frozen. The Idle and Power Down modes are activated by setting bits in Special Function Register PCON. The address of this register is 87H. Figure 26 details its contents. In the HMOS devices the PCON register only contains SMOD. The other four bits are implemented only in the CHMOS devices. User software should never write 1s to unimplemented bits, since they may be used in future MCS-51 products. IDLE MODE An instruction that sets PCON.0 causes that to be the last instruction executed before going into the Idle mode. In the Idle mode, the internal clock signal is gated off to the CPU, but not to the Interrupt, Timer, and Serial Port functions. The CPU status is preserved in its entirety, the Stack pointer, Program, Counter, Program Status Word, Accumulator, and, all other registers maintain their data during Idle. The port pins hold the logical states they had at the time Idle was activated. ALE and -PSEN hold at logic high levels. There are two ways to terminate the Idle. Activation of any enabled -interrupt will cause PCON.0 to be cleared by hardware, terminating the Idle mode. The interrupts will be serviced, and following RETI the next instruction to be executed will be the one following the instruction that put the device into Idle.
The flag bit GP0 and GP1 can be used to give an indication if an interrupt occurred during normal operation during an idle. For example, an instruction that activates Idle can also set one or both flag bits. When Idle is terminated by an interrupt, the interrupt service routine can examine the flag bits. The other way of terminating the Idle mode is with a hardware reset. Since the clock oscillator is still running, the hardware reset needs to be held active for only two machine cycles (24 oscillator periods) to complete the reset. The signal at the RST pin clears the IDL bit directly and asynchronously. At this time the CPU resumes program execution from where it left off; that is at the instruction following the one that invoked the Idle Mode, As shown in Figure 25, two or three machine cycles of program execution may take place before the internal reset algorithm takes control. On-chip hardware inhibits access to the internal RAM during this time, but access to the port pins is not inhibited. To eliminate the possibility of unexpected outputs at the port pins, the instruction following the one that invokes Idle should not be one that writes to a port pin or to external Data RAM. POWER DOWN MODE An instruction that sets PCON. 1 causes that to be the last instruction executed before going into the Power Down mode. In the Power Down mode, the onchip oscillator is stopped. With the clock frozen, all functions are stopped, but the on-chip RAM and Special Function Registers are held. The port pins output the values held by their respective SFRs. ALE and -PSEN output lows. The only exit from Power Down for the 80C51 is a hardware reset. Reset redefines all the SFRs, but does not change the on-chip RAM. In the Power Down mode of operation, VCC can be reduced to as low as 2V. Care must he taken, however, to ensure that VCC is not reduced before the Power Down mode is invoked, and that VCC is restored to its normal operating level, before the Power Down mode is terminated. The reset that terminates Power Down also frees the oscillator. The reset should not be activated before VCC is restored to its normal operating level, and must be held active long enough to allow the oscillator to restart and, stabilize (normally less than 10 msec). Intel 8-Bit Embedded Controllers 1991 |
At the end of the Oney excerpt is the coff dump of USBD.SYS.
Csd compiled some of the Compuware/Numgea code. Compilation was without errors. Saturday February 9, 2002
Anchor/Cypress 3671 and 3681 development board are audacious in the sense they are RAM-based.
Oney writes
One of the key features of the Anchor Chips chip set is that you can download firmware over the USB connection easily. For a programmer like me with a phobia for hardware in general and EEPROM programming in particular, that feature is a godsend.
Csd is looking at options for the PC side of USB communications.
Should we use Oney, Cypress or Compuware/Numega USB 1.x, not 2.0 yet, software or do something else? We don't know yet.
Oney's and Microsoft C++ approach has problems too.
It's slow and the C++ code is abstract and hard to read for embedded controller programmers..
Oney writes in Chapter 11, The Universal Serial Bus,
Rather than communicate directly with hardware by using hardware abstraction layer (HAL) function call, a USB driver relies heavily on the bus driver (USBD.SYS). To send a request to its device, the driver creates a USB request block (URB), which it submits to the bus driver.
This may not work with USB 2.0 until Microsoft releases the drivers for XP. So we might use some HAL calls. Friday February 8, 2002 08:45
NOTE ON SAMPLE PROGRAMS Anchor Chips, Incorporated (http://www.anchorchips.com), kindly provided me one of their EZ-USB development kits. The Anchor Chips USB chip set revolves around a modified 8051 microprocessor and additional core logic to perform some of the low-level protocol functions mandated by the USB specification. The development board also contains additional external memory, a UART and serial connector, a set of push buttons, and an LED readout to facilitate development and debugging of 8051 firmware using Anchor Chips software frame- work. One of the key features of the Anchor Chips chip set is that you can download firmware over the USB connection easily. For a programmer like me with a phobia for hardware in general and EEPROM programming in particular, that feature is a godsend. The USB sample drivers on the companion disc illustrate the simplest possible USB devices and stand alone as examples of how to perform various tasks. If you happen to have an Anchor Chips development kit, however, you can also try out these samples with real firmware. Each sample contains a WDM driver in a SYS subdirectory, a Microsoft Win32 test program in a TEST subdirectory, and a firmware program in an EZUSB directory. You can follow the directions in the HTM files included with each sample to build these components or to simply install the prebuilt versions that are on the disc. A word of caution is in order here. Anchor Chips provides a reduced- function version of 8051 development tools authored by Keil Elektronik GmbH. Youll need an unlimited version of those tools (which you must license separately from Keil) to develop real firmware and even to build some of my samples. You might also need some perseverance to get past the rather dated interface offered by these 16-bit programming tools. But, by the time you read this, Keil will have introduced new, considerably improved 32-bit tools for the 8051 called uVision2. Programming the Windows Driver Model The offical guied to the Microsoft Windows Driver Model Microsoft Press Keil with uVision 2 is downloading software over the serial port! Microsoft (R) COFF Binary File Dumper Version 6.00.8447 Copyright (C) Microsoft Corp 1992-1998. All rights reserved. Dump of file C:\windows\system32\drivers\usbd.sys File Type: DLL Section contains the following exports for USBD.SYS 0 characteristics 37FFA896 time date stamp Sat Oct 09 13:41:58 1999 0.00 version 1 ordinal base 33 number of functions 33 number of names ordinal hint RVA name 4 0 0000350E USBD_AllocateDeviceName 5 1 00001EB0 USBD_CalculateUsbBandwidth 6 2 00001C02 USBD_CompleteRequest 7 3 00003DDA USBD_CreateConfigurationRequest 1 4 00003C32 USBD_CreateConfigurationRequestEx 9 5 00003806 USBD_CreateDevice 10 6 00000AE0 USBD_Debug_GetHeap 11 7 00000ADC USBD_Debug_LogEntry 12 8 00000AF6 USBD_Debug_RetHeap 13 9 00000BAE USBD_Dispatch 14 A 00003E82 USBD_FreeDeviceMutex 15 B 00003588 USBD_FreeDeviceName 16 C 0000339A USBD_GetDeviceInformation 17 D 00001C54 USBD_GetInterfaceLength 18 E 00001F62 USBD_GetPdoRegistryParameter 19 F 00001C64 USBD_GetSuspendPowerState 20 10 00000B96 USBD_GetUSBDIVersion 21 11 00003A30 USBD_InitializeDevice 22 12 00001DD0 USBD_MakePdoName 23 13 00003DB8 USBD_ParseConfigurationDescriptor 2 14 00003D2C USBD_ParseConfigurationDescriptorEx 3 15 00003CF8 USBD_ParseDescriptors 26 16 00002018 USBD_QueryBusTime 27 17 00001CA0 USBD_RegisterHcDeviceCapabilities 28 18 00001C8C USBD_RegisterHcFilter 29 19 000035F6 USBD_RegisterHostController 30 1A 00003974 USBD_RemoveDevice 31 1B 00001EF8 USBD_RestoreDevice 32 1C 00001C76 USBD_SetSuspendPowerState 33 1D 00003E64 USBD_WaitDeviceMutex 8 1E 00003C32 _USBD_CreateConfigurationRequestEx@8 24 1F 00003D2C _USBD_ParseConfigurationDescriptorEx@28 25 20 00003CF8 _USBD_ParseDescriptors@16 Section contains the following imports: NTOSKRNL.EXE 1204C Import Address Table 14408 Import Name Table 0 time date stamp 0 Index of first forwarder reference E1 RtlQueryRegistryValues F6 ZwClose 57 IoOpenDeviceRegistryKey DE RtlInitUnicodeString 104 ZwQueryValueKey 7 ExAllocatePoolWithTag 8C KeReleaseSemaphore 9C KeWaitForSingleObject 76 KeDelayExecutionThread 2F IoAllocateMdl 37 IoCancelIrp A0 MmBuildMdlForNonPagedPool 95 KeSetEvent 74 KeCancelTimer 6F IofCallDriver 97 KeSetTimer 7A KeInitializeDpc 7F KeInitializeTimer 7E KeInitializeSpinLock 7B KeInitializeEvent 34 IoBuildDeviceIoControlRequest 4C IoGetDeviceProperty C ExFreePool DA RtlFreeUnicodeString 3D IoCreateSymbolicLink 70 IofCompleteRequest BD PoStartNextPowerIrp BB PoSetPowerState 27 InterlockedExchange 5C IoReleaseCancelSpinLock 29 IoAcquireCancelSpinLock 17 ExQueueWorkItem BA PoRequestPowerIrp B6 ObfReferenceObject B7 PoCallDriver 3A IoCreateDevice 3F IoDeleteDevice 40 IoDeleteSymbolicLink 45 IoFreeMdl CA RtlAppendUnicodeStringToString DF RtlIntegerToUnicodeString 61 IoSetDeviceInterfaceState 59 IoRegisterDeviceInterface CD RtlCompareMemory 106 ZwSetValueKey 7D KeInitializeSemaphore HAL.DLL 12040 Import Address Table 143FC Import Name Table 0 time date stamp 0 Index of first forwarder reference 6 KfAcquireSpinLock 9 KfReleaseSpinLock Summary 220 .data 4C0 .edata 160 .rdata 2C0 .reloc 3C0 .rsrc 1D40 .text 520 INIT 1B40 PAGE |
USB 2.0 must run convincingly at 480 megabits or about 60 megabytes per second to an from a peripheral device. Cypress claims to have a microcontroller which will do this on the peripheral side. The 3861 development board.
On the PC side the NEC µPD720100A appears to be the leader with the USB 2.0 to pci bus chip 1 2.
NEC literature update below. Monday February 4, 2002 11:05
Here's a jpg of the My Computer|Control Panel|Hardware| under
Windows 2000 of USB devices.
CSD has an Adaptec USB 2.0 pci card.
Note the Cypress driver. Next note the two Intel pci to usb ports. The motherboard on this system is a SOYO. Let's look at the Intel resources
Note that this is in the i/o space. Next let's look at the NEC resources
Memory mapped! Additional speed required by USB 2.0 over 1.x apparently requires the the 4k bytes of memory buffer and memory mapping opposed to a port map. Locating data sheets on the NEC USB 2.0 part on Internet proved not too productive. But here's some information posted in Germany. 1 Here's a NEC literature phone number 1 800 366 9782. This works, John, is getting sending a request to headquarters to have data sheets, user manuals, and app notes sent. Sure would be fun to write a Windows 2000 wdm miniport driver to connect the cypress 3671 ez-usb-fx to the NEC µPD720100S! 0113KV202 on the Adaptec pci to USB 2.0 board. Once we get a decent operating system running on the Cypress 3671 and 3681 development boards, of course! First things first. One needs decent operating system on the peripheral side to debug the wdm miniport driver. On second thought, one might plug one of the 5 USB connectors on the Adaptec board into one of the other to try communication. CSD is into miniport pci/cardbus drivers. 1 And needs to get into USB 2.0! USB 2.0 is the future! |
Senior techies 1 Thursday June 6, 2002 18:15 |
Memory 1 Tuesday February 12, 2002 06:54 |
Core meltdowns 1 Tuesday February 12, 2002 06:54 |
Palm operating system port 1 Wednesday February 20, 2002 12:31 |
Industrial espionage 1 Sunday March 24, 2002 08:43 |
E-waste 1 Tuesday May 28, 2002 07:56 |
Comcast and Excite@Home [Excite removed all files. Comcast has not yet provided web space] 1 Thursday April 4, 2002 13:38 |
Electronic and Internet Business 1 Monday February 12, 2001 09:34 |
Real-time operating systems 1 Tuesday January 1, 2002 15:12 |
Economics of Software 1 Monday January 28, 2002 13:08 |
USB 1 Tuesday January 1, 2002 16:09 |
IEEE 1394 Firewire 1 Sunday February 3, 2002 08:22 |
X86 processors and wdm drivers 1 Wednesday September 20, 2000 12:08 |
Architectures, bus board, and periperal microcontroller 1 2 Monday May 5, 2000 09:29 |
Bluetooth 1 Monday February 12, 2001 09:38 |
Wireless 1 Friday April 28, 2000 15:37 |
80C51/52 and wdms 1 Tuesday February 12, 2002 07:13 |
Micocontrollers 1 Wednesday November 15, 2000 10:29 |
Software reality 1 Friday January 4, 2002 09:05 |
Windows 2000, XP, ME 1 Thursday April 4, 2002 13:45 |
Electronic components and wdm drivers 1 Wednesday September 20, 2000 12:42 |
Hardware reality 1 Wednesday September 20, 2000 12:42 |
Here's results of the fig Forth VLIST on windows 2000 with the fantastic software productivity tools C/C++, Visual Basic 60, and DriverWorks 2.0 icons in the background. Wednesday December 19, 2001 15:16
|
Csd's first attempt to run a forth operating system in a DOS Windows 2000 window failed. So a second attempt was made.
It worked!
A televideo terminal emulator written in forth in the early 1980s is running on PC Forth operating system which has, more or less, taken over windows.
When IDS [initialize development system] is typed, an 8051 thinks that the keyboard, screen and one disk file is its.
The cold start message "8051 series... " message is coming from the 8051.
The second picture shows the 8051 system [to the right] with an experimental hardware board plus board which contains the 8051 forth operating system.
We have some ideas on how to modernize this operating system. The terminal emulator should be written as a Windows application in Visual Basic 6.0 communicating with the 8051 [52 now] through a dll to a wdm driver!
Having an interactive operating system of some sorts on a microcontroller really lowers the cost of development. Both hardware and software.
One cannot usually guess all of the failure modes of hardware or software, therefore an interactive operating system on a peripheral microcontroller communicating to Windows is very valuable. Wednesday December 19, 2001 14:40
|
USB 2.0 looks to be coming way to connect a high-speed peripheral to a PC. Here's an ad from a Circuit City included with the Sunday December 2, 2000 Albuquerque Journal.
Note the necessity for a Adapetec card - with no drivers! Monday December 3, 2001 09:19
Csd is examining usb 1.0 and 2.0 implementations.
We bought an Adaptec usb card. It has four ports on it. Three coming out the back of a pci card and the fourth internal.
We attended Microsoft's October 24, 2001 rollout of Windows XP in Albuquerque. Here's a jpg of the information bag handed-out to attendees
Note the list of approved vendors.
We then bought a FujiFilm FinePix 2300.
A shock came when we discovered that there are no usb 2.0 driver cd with the Adaptec product. Plug and play doesn't recognize the card.
The Adaptec card is a usb to pci bus interface using an unusal Japanese chip set. More about this later.
The Fuji usb interface software works perfectly with motherboard usb interface. Here's more about the FinePix 2300.
Csd is about ready to begin test of
for use to connect a PC to custom hardware.
From the information at the XP rollout, Microsoft seems to push approved vendors. This likely means that usb interfaces using Microsoft's usb class driver will work well. Add-on board interfaces might have problems. Tuesday November 113, 2001 12:25
NOTE |
A wdm driver should be written in as high of level language as possible. And as low of level language as needed. Like in-line assembler.
CSD uses Visual Basic 6.0 to debug its drivers. The reasons for using Visual Basic 6.0 have been eloquently spelled-out in Paul Sheriff's outstanding book Visual Basic 6, Que publishers. 1 2 2.5 3
Sheriff posted source code for examples in his book at a web site.
The book has, we believe, the best coverage of Active Data Object and ODBC.
CSD now gives you the opportunity to experiment with one of its wdm driver packages http://members.home.net/bpayne37/softwaredownloadtest.html.
This pci/cardbus driver works best on Windows 2000.
Windows 2000 is so much better than 98 for industrial applications, that CSD is only focuses on drivers for 2000.
A network administrator reported that his network was up 99.97% of the time using Windows 2000 server. Asked what the reliability was under NT 4.0 server, the administrator reported, "It sucked. It was so bad that I didn't bother keeping statistics.
VB 6.0 is a totally different software technology compared to the old Basics. The old Microsoft Visual Basics were based on Forth threaded code technology.
You, again, can read about how the old BASICs work
1.
CSD posts the a portion of Chapter 1 to Visual Basic 6.0
written by Sheriff.
Microsoft's "p-code" is not the old UCSD Pascal p-code but, a German reported, is better called t-code for threaded code! 2 Saturday June 30, 2001 08:01
Visual Basic Advantages
Visual Basic has many advantages over other development languages. Here's a list of some of these advantages: Has a shorter learning curve and development time than C/C++, Delphi, and even PowerBuilder. Visual Basic Disadvantages Although it has many advantages, Visual Basic still has a few disadvantages. Most of these are related to some limitations of Visual Basic in getting at real low-level functionality in hardware and the operating system. The following is a list of so of these disadvantages: Runtime distribution requirements are larger than those for C/C++. Code Generation Options There are basically two kinds of programming languages: interpreted and compiled Visual Basic lets you take advantage of both of these within one environment. is one of the best features of Visual Basic. An interpreted language waits until time to translate the source code into machine language. A compiled language translates all the source code into machine code, all at one time. The advantage using an interpreter is quick development time, whereas the advantage to using compiled language is quick runtime execution. With Visual Basic, you have the best of both, because Visual Basic has the ability run interpreted in the development environment but can compile the source mt machine code for the final runtime. Visual Basic uses a modified version of an interpreted language called p-code. Let's take a look at the differences between p-code executable and a natively compiled executable. p-code Interpreted languages must translate each line of source code into machine language (ones and zeros) each time that line of code is executed. Therefore, if a is executed 1 00 times, it's translated 100 times. To write an interpreted language you simply write the source, and then run it. Examples of interpreted languages the old traditional GW-Basic and most batch languages. Compiled languages translate all the source code just once into machine language To write a compiled application, you typically write the source code, compile a link the program, and then run the program. Examples of compiled languages COBOL, Fortran, and C. p-code is a hybrid of interpreted and compiled languages. A compiler compiles the source code, but it does not produce machine language. Instead, it produces a series of symbols for the source code. These symbols make a line of code that has five words into just one or two characters. When you run the program, a p-code interpreter interprets each line of the p-code, turning it into machine language. When p-code is executing, it's faster than interpreted code because there are fewer characters to read; however, it's still slower than compiled code. Visual Basic has the option of generating p-code or generating native code. Native Code Visual Basic also has the ability to generate native code EXE files. This can give you some significant enhancements in performance when doing CPU-intensive calculations. If you have a part of your program that's CPU bound (generating Mandelbrot sets, for example), you can compile that piece into native code. You might also want to write this piece as a C language DLL that you could call from Visual Basic. Which Option Should You Choose? When you're running a Windows program, much of the CPU time is used by Windows function calls or calls to external DLLs, not by the source code in your program. A lot of time is also spent waiting on the user to perform the next task, click the next button, or maybe type the next character. Therefore, Visual Basic p-code programs are generally fast enough for most business applications. Future of Visual Basic Since the introduction of Visual Basic for Applications (VBA), Microsoft has positioned Visual Basic as the premier general-purpose Windows development tool. It's the glue for assembling software components into software solutions. All the Microsoft Office products now include Visual Basic as their macro/development language. Many third-party companies are licensing VBA so that their customers can program their products. Microsoft has also created a scripting edition of Visual Bask for Internet development both in the browser and on a Web server called VBScript. Visual Basic's main competitors are Powersoft's PowerBuilder Borland's Delphi, and Microsoft's own Visual C++, Access, and Visual FoxPro. Although these are all fine languages, none enjoys the success or widespread use that the Visual Basic language does. |
Software code and documentation costs
Monday December 25, 2000
12:08
|
Real world
perspectives on
high
tech
for project
success