Aug 31st

Cunning Plan to Kill Windows 98SE?

One thing I've been wondering about since Microsoft introduced Windows XP product activation was whether Windows 98SE support would slowly fade, forcing everyone to use a newer OS.  For gamers who don't need Fort Knox security on their gaming rig, Windows 98SE plays more games in a more trouble-free fashion than any OS ever made by Microsoft.  Indeed, as more and more users get fed up with Windows Me (because it's awful) and Windows XP (because of the product activation scheme), they are going to places like ebay looking for old copies of Windows 98SE.  

Indeed, we have been trying to update our entire LAN to Windows 2000 over the last few months, but each time we take a step forward, we end up taking two steps backwards.  Just last week, after quite some effort to get all our favorite games working on a new Windows 2000 system here, we could not get several games to work at all, and ended up back with Windows 98SE. Now all is well again, and with substantially less effort. Many gamers out there have probably had similar experiences.

But now comes the zinger.  NVidia just released their beta 40.41 detonator drivers, supposedly offering a 25% performance boost (oh sure, just like all the other driver releases?).  But the first thing you'll notice is that they only support Windows 2000 and XP.  This, despite the fact that writing drivers for Windows 98SE is a snap in comparison.  So, my question is, will NVidia release a Windows 98SE version soon?* or is Bill asking companies to stop supporting the old, now re-popular operating system? Only time will tell, but it sure is a sneaky way to kill a good, simple, and very game-worthy OS. *(Will there be a Windows 98SE version of Direct X 9??).

                                    Dr. John

Aug 29th

Where is Unreal Tournament 2003?

Rumors abounded back in June that Unreal Tournament 2003 was just about to be released. Then those rumors shifted to July. Then the rumor mill went silent, and now it is almost September with no sign of the demo anywhere! This always happens with games, especially good games, so it is not an unusual situation. And I must admit, I'd rather have them get it right the first time, rather than having to release a whole bunch of patches.

So when it is it coming out? Even the people who made the game do not know for certain. But my guess is the demo will be out in early September. The game should be out by October sometime. For more info, listen to the latest Daily Rumor.

                                    Dr. John

Aug 27th

BAPCo story Reaches Phat Media

Like a juicy public relations nightmare, the Intel/BAPCo benchmark story has reached big media outlets, including the NY Times.  The questions I have are; will this change anything? or, is it a quick zinger story? The underlying issues here eat away at the computer industry like a cancer. Industry nepotism is rampant, and each industry clique vies against the other in sanguinary ways. "Standards committees" are often not nearly standard,  and wild marketing claims are dismissed as so much puffery. There is no safe harbor for customers buying a PC that says, computer chip A is X% faster than computer chip B, because consumers haven't insisted on it. If this story doesn't get their hackles up, nothing will.

                                    Dr. John

NY Times

EBN

Van's Hardware Folds

Right after attacking Intel and BAPCo with a vengence, Van's Hardware review site has closed shop, citing money and threats as the main reasons.  Van also has quite a rant on Tom's Hardware, actually on Tom, which is a real eye opener, and a window into the weird world of hardware review sites.  If you like Tom's site (which I must admit, I don't), you may prefer not to read the rant. I'll take the boy wonder, Anand, over Tom anytime (but lose some of the ads Anand... please!:).

                                     Dr. John

Aug 26th

Rampant Cynicism Good News for Intel

As the Intel/BAPCo fiasco plays out on the web, the entrenched cynicism among IT reporters and analysts makes change seem unlikely. Rather than displaying outrage, and spunk, as you might expect from the likes of the ever watchful Mike Magee of The Inquirer, we get cynical acquiescence. I understand that Mike has had numerous contentious interactions with both chip makers, but that doesn't mean it's time to give up. But rather than being persuasively pugnacious, Mike now seems resigned to the status quo, as though he can't make a difference anymore. I believe this is wrong, and that if players like Mike put Intel's feet to the fire, we might get better corporate behavior out of them. Not because they "feel bad" and "want to do good", but simply because they can't get away with crap without shooting themselves in those same fire-roasted feet.

But if Mike sticks to his guns on statements like "The performance argument is all part of the myth, fostered by the industry itself, that computers are like cars", we are in a significant deal of trouble.  Because computers are like cars, they do perform differently, and don't let Mike or anyone else convince you otherwise.  Maybe you'd prefer a 1GHz Celeron with 64MB of SDRAM and a built in VGA video card to play Unreal Tournament 2003?  Don't think so.  Gamers are not "the average PC customer", and to them, computers are exactly like cars, there are economy models, and performance models, and the difference is night and day when playing high-end 3D games.

Finally, I'd like to say that Mike and others on the news end of the spectrum have been bad-mouthing PR (performance ratings) for CPUs for some time.  This is mainly because Mike and others email and type on their computers, and that's about it.  That 1GHz Celeron would be more power than any email or word processing program could ever need.  But if Mike were digitally editing his home videos, or voice dictating into his computer, or, heaven forbid, playing Unreal Tournament 2003, he might be much happier with a 2600+ Athlon XP with 512MB of PC2700 DDR DRAM and a Radeon 9700 video card. 

The bottom line is this. As processors become more divergent in the ways they process data, and more companies provide low-cost, high performance options, an honest, non-partisan, dependable rating system will be essential for properly comparing the processing power of a CPU to the competition. Mike says: "Some will say we need an independent benchmark so consumers can really understand how a PC performs. AMD says that. But does anyone really think that's ever going to happen? Perhaps we should let Microsoft do the benchmarketing". Well, maybe if everyone loses a touch of that creeping cynicism, and gets back on their soapbox, the world may just end up with a standards body that has all CPU makers as members. But accepting the status quo is not the way to go. "Say it ain't so Mike!"

                                     Dr. John

First 'Bionic' Eye produced

The worlds first bionic or electronic eye has been developed in Australia, and is ready for human trials. The small circuit disk is implanted on the retina, and makes contact with the so-called 'ganglion cells' there, that send their nerve fibers to the brain. The chip picks up light, and transmits the data to the ganglion cells via small electrodes. Right now the chip can only deliver a 10 x 10 pixel image, but this will be improved over time. It's actually quite a fascinating invention, and it will be very interesting to see how well it works.

Bionic Eye

Aug 24th

SYSMark 2002 Biased Towards Intel

The big talk on hardware and IT news sites in the last few days has been on BAPCo's SYSMark 2002 benchmark, and it's bias towards Intel's Pentium 4 processor. 

Van's Hardware has a very interesting piece comparing SYSmark 2001 with SYSmark 2002 benchmark suites. The long and short of it are that Intel and BAPCo worked together to optimize SYSmark to favor the Pentium 4 processor. Indeed, Intel apparently loaned engineers to BAPCo in order to help them optimize the benchmark for Intel processors.

But it goes much further than mere optimization for Intel's processors. The Athlon XP came out right after SYSmark 2001 was released, and much to Intel's shock, the Athlon beat the Pentium 4 decidedly, even when running at significantly lower clock speeds. Then BAPCo and Intel went back to work, and quickly removed all tests where the Athlon beat the Pentium 4. Then they decided to have the benchmark suite repeat memory intensive tasks where the Pentium 4 had an edge  in order to skew the results further in favor of Intel's CPU. The end result was SYSmark 2002, the most unrealistic and biased benchmark suite ever produced on the planet.

Perhaps the most interesting part about this damning evidence against Intel and BAPCo is that Intel just got hit with a lawsuit complaining that the chip giant misled customers about the performance delivered by the Pentium 4 processor as compared with the Pentium 3 and Athlon. The radical transformation of SYSmark 2001 into SYSmark 2002 under Intel's guidance provides ample ammunition for the court case, and is certain to figure as a central piece of evidence in support of the litigants claims.

This embarrassing story for Intel is also an excellent reason why performance ratings are going to be essential when comparing the performance of future processors. Obviously a 1.3 GHz Celeron, 1.3 GHz Duron, and 1.3 GHz Athlon XP will all performed differently despite their identical clock speeds. How are customers going to be able to compare the actual performance of processors in order to make educated buying decisions? With the advent of the latency rich Pentium 4 processor, this CPU conundrum becomes ever more pressing. 

Despite complaints from Intel lap dogs, performance ratings in some form or another will be with us forever. There is no other way to compare processors with radically different internal design features. Customers need to know how fast a processor will run the tasks that they most commonly do with their computer, so that they can make an educated price vs. performance decision. So, for example, there could be a gaming PR rating, an office suite PR rating, and a digital encoding PR rating. But it is clear now that somebody other than BAPCo is going to have to produce a reliable and trustworthy PR rating system. Maybe it's time for a standards body to be formed to accomplish this task.

                                     Dr. John

Van's Hardware

Aug 21st

Athlon XP passes 2GHz

AMD is about to release tweaked versions of the Athlon XP (Thoroughbred) CPU, including 2400+ and 2600+ models. The 2600+ runs at 2.13GHz, finally breaking the 2GHz barrier with the Athlon. And despite it's dramatically slower memory and bus speeds when compared with the P4 "B", it still beats the new 2.53GHz P4 in a number of benchmarks. Read about it at the Tech Report.

Aug 20th

First Electronic Force Field Invented

"Shields Up!" Future tank commanders may actually issue this command. Scientists have invented an electronic force field which disperses shaped charge gas jets, the most common method of breaching tank armor. Inexpensive anti-tank weapons like the 'rocket propelled grenade 7' work by creating a super-hot jet of metal gas, which can burn through a foot of steel armor. Because this jet of gas is composed of metal atoms in the gas state, it can conduct electricity.  The new "force field" simply puts two plates of conducting material on the outside of the tank, with an insulator layer between them. One plate is then charged with high voltage, like a capacitor, while the other is grounded.  As the metal gas jet burns through the insulator between the plates, a circuit is completed and the high voltage vaporizes some of the metal gas, and conducts some of it to ground. The end result is that the outer plates take all the damage, and the tank remains unscathed. 

While this isn't really a force field, it acts much like one with respect to shaped charge weapons. High velocity (sabot) type weapons used by many tanks would not be stopped by the new shield system. But early tests show that shaped charge weapons are ineffective against shield-equipped vehicles.

                                     Dr. John

Telegraph

Windows 2K Service pack 3, Revisited

As I mentioned a few days ago, Windows 2K SP3 has lots of snoopware built in to allow Bill to rummage around your hard drive, and make changes Willy (Billy?) Nilly. But there are solutions to these nagging problems. Over at The Register they talk about how to disable the features that get installed with the new service pack. It's fairly straight forward, and sounds like it should eliminate most of the offending code.

Then there is the "Matt Solution" provided kindly by one of our message board users. These two small files should remove most of the offending code from SP3.

Instsrv.exe (service installer/uninstaller)

cleanup.bat (eliminate auto-update batch file)

Aug 19th

How Fast is the ATI Radeon 9700?

If the tentative benchmarks shown over at the Inquirer are accurate, the new ATI 9700 may not be all it was hyped up to be.  Of course these numbers could be wrong, but only time will tell for sure.  According to the few benchmarks shown, the ATI card only beats NVidia Ti4600 chips by relatively small margins in some benchmarks. The Ti 4600 still wins in a few, including Quake 3! This is quite contrary to the "50% faster than the Ti4600" talk we have been hearing.  Indeed, if the numbers posted are close to the real situation, then the Radeon 9700 isn't a GF4 Killer, it's merely a solid catch-up. 

Anandtech has a much more thorough analysis, and indeed the 9700 does look at least 30% faster in some benchmarks. If you like using antialiasing and anisotropic filtering, the Radeon 9700 does a far better job than the 4600. But without those enabled, the 9700 just keeps pace with the GF4 4600 in several benchmarks. Nonetheless, the 9700 is the first Direct X 9 part to market, and has many nice advanced features, so ATI fans will definitely be pleased, especially those who want to use antialiasing. NVidia fans are now salivating over the upcoming GeForce 5 (NV30), which is scheduled for a late Fall release. The Radeon 9700 should be generally available by the beginning of September.

                                     Dr. John

VisionTek Bankrupt

The NVidia card maker Visiontek is going out of business.  It's always been tough making money as one of NVidia's chip purveyors, because there is so much competition. Visiontek had entered the video card scene late in the game, and now exits it just as quickly. This leaves several significant players, including Asus, Gainward, PNY, Abit and MSI. Visiontek was important for system builders because they often had good quality cards to market before the others. They will be missed.

                                     Dr. John

Hard[OCP]

Aug 17th

Intel Sued Over Poor P4 Performance! :)

Much to Paul Hales' consternation (see Aug 13th story below), Intel is actually being sued for making claims about the speed of the Pentium 4 processor. The litigation claims that Intel's Pentium 4 processor is no faster than the Pentium 3 or Athlon at the same clock speed. In fact, that is an understatement. The Pentium III processor was notably faster when it got 512 KB of L2 cache. The Athlon XP has been much faster clock for clock since its debut. But that never deterred Intel from claiming that it had the world's fastest processor. Indeed, Intel does finally have a slightly faster processor in the form of the Pentium 4 2.53 gigahertz processor, which just barely beats the Athlon XP 2100+ processor by a few percentage points in benchmarks. Since the 2100 actually runs at 1.73 gigahertz, it is apparent that Intel needs at least a 700 MHz lead just to barely edge ahead of the Athlon.

Clearly, this is the first time that a new processor has been produced which was substantially slower than the older design. This move was purely intentional on Intel's part, because the slower processors could be run at higher clock speeds. But if a 2.5 gigahertz processor barely beats a 1.7 gigahertz processor that costs hundreds of dollars less, then Intel is selling an inferior product for more money.

It will certainly be interesting to see what happens with this class action lawsuit. Will it grow in size to include thousands or tens of thousands of irritated Pentium 4 owners? Or will the overcharged Intel fans just acquiesce? While the litigants are at it, maybe they ought to sue over Intel's claim that the Pentium 4 processor would dramatically improve Internet access, one of the most bogus claims ever made in modern processor marketing. Just like Microsoft, Intel feels that it is too big and important to be hassled for making false claims. Let's just hope they're wrong.

                                     Dr. John

PC World

Paul... Eat your hat.

Aug 16th

How to Tame Windows 2K SP3.

So you've been itching to install Windows 2K SP3, but you don't like the new EULA that gives Bill access to your PC and rights to your firstborn? Well, there may be an answer, thanks to John Lettice's snooping around.  He found a very nice utility at a very cool web site called securitysoftware.cc, run by "Pertinax". The utility is called "Enabler" and you can get it here. It lets you click on "Next", even if the "I do not accept the EULA" is checked while installing SP3. While it is not clear to me that this prevents loading of the onerous code that permits Bill rummaging around your hard drive, folks seem to think that such is the case.  Don't ask me, I just assume Bill delivers his payload regardless.  But it's certainly worth a try.

While your at it, pick up "Slap" which works in conjunction with Black Ice and Zone Alarm to send a disturbing port scan back at anyone who does the same to you.  He has even provided the perfect wav file to send along with your digital slap! (WAV). Other cool utilities he's written include a real Taskmanager for Windows 98!

Bill's gonna love this! Can't you just see the veins bulging on the sides of his head?  Enjoy!

                                     Dr. John

Aug 15th

NV30 "Taped-Out"

Anand has an article on the NV30 chip from NVidia, which will be the basis for the upcoming GeForce5 card. The new chip has been taped-out, which is industry slang for "we got one to work!" Now that they have shown the design works, and can be manufactured, it's a matter of refining, scaling up, and improving yields. Currently yields are in the 10% to 20% range, but that is the record on the first few trial runs. As they refine the process, yields will improve.

The NV30 is built at 0.13 microns, as opposed to ATI's R300 chip, which is made at 0.15 microns. Rumors suggest the NV30 is twice as fast as the NV25 (GeForce4), which would make it about 50% faster than the R300 from ATI. Expect actual cards to be at stores around late November or early December. That timing guarantees that the first ones out will be rare, and expensive.

                                     Dr. John

Aug 13th

Paul Hales and Dr John Debate AMD's PR System

Paul Hales of The Inquirer has written several short pieces recently where he has quipped at AMD and their performance rating system as "pure marketing baloney". I wrote an e-mail to Paul expressing my opinion, as many, many others also appear to have done. Paul was kind enough to respond, but I don't think my arguments got through to him. Nonetheless, our discourse probably mirrors the debate going on throughout the industry, and as such is interesting in its own right. Here goes...

Dear Paul,

You mentioned the review of Intel's Celeron 1.7 GHz processor which was posted over at the Tech Report. Our company acts as the hardware sponsor for The Tech Report. I'll be willing to bet that you did not read that article, because if you had you would have much more respect for AMD's performance rating system.

If you have the time, I suggest you read Scott's excellent review which makes use of several different benchmarks in different software categories. The first thing you'll notice is that the Duron at 1.3 GHz actually beats the Pentium 4 running at 1.7 GHz in two benchmarks!

Examples:

1) DivX encoding (xmpeg)
2) Sciencemark benchmark

It also beat the 1.7 GHz Celeron in most tests. So should AMD rate their chips by MHz, even though Intel's clock ratings don't reflect actual performance?

Scott sums up the introduction thusly: "We've also rounded up a diverse suite of performance benchmarks and a couple of competitors to show just where the Celeron 1.7GHz sits in the value processor field and why you can never trust clock speeds alone".

Why not write an unbiased story about the PR system for a real change? :)

John Moffett
KickAss Gear

Blimey,
Biased? Me?..

The clock speed a processor runs at is just that. The chipmakers have been hyping the big numbers to sell their chips. Now, when AMD's chips are more efficient (let's say) at lower clock speeds they are victims of the trap they themselves happily fell into.

It's AMD that needs to address the problem. It's my job to take the piss.

Anyhow, thanks for the note. I'll try and have a closer look at what Scott says, later..

PH

Hello again Paul,

Thanks for the reply, I did not expect one, and certainly not so soon.

I think this is the crux of the misunderstanding here, you wrote:

"AMD's chips are more efficient (let's say) at lower clock speeds they are victims of the trap they themselves happily fell into."

This statement is not true. It is Intel who reworked the Pentium 4 pipeline to be longer for the sole purpose of artificially inflating the MHz ratings that their chips can run at. AMD stuck with the basic pipeline architecture from previous chips, including those made at Intel. So AMD did not make a trap that they themselves fell into. Intel's longer instruction pipeline leads to large performance hits when branch misprediction occurs on the longer pipeline. But the end result is a chip whose MHz ratings can be increased to what has to be called artificially high levels that do not compare directly to shorter pipeline architectures.

Why is this a trap that AMD fell into? It's a marketing ploy on Intel's part, and highly disingenuous. Consumers are buying more expensive, slower processors, and reporters such as yourself aren't helping clue them in (which is your job, at least more than "taking the piss" from readers is). The Pentium 4 is the slowest CPU MHz for MHz available on the market today. But it has the highest MHz rating.  Who set the trap?

Hope this isn't perceived as just more piss....

Cheers,

John

Hi John,

What you say about Intel may be true but it doesn't make what i said about AMD untrue, I reckon.

Anyhow, thanks for your comments.

I think its safe to say we'll see more on this in the proper place..

PH

The Inq

Aug 11th

Microsoft plays monopoly, passes go, collects $200.

If the e-mail circulated at Dell Computers is real, it indicates that Microsoft is pushing its monopoly position for all it's worth. A fascinating move considering the penalty phase is still underway in their antitrust lawsuit case with the U.S. government. Of course, ever since George II became ruling monarch of the United States, the court case seems to have turned to mush. Microsoft knows this better than anyone, and has been letting their monopolistic tendencies boil to the surface on a regular basis.

According to the document, Dell has been forbidden to sell computers with the "no OS" option starting Sept. 1st. And since Dell is only allowed to sell Windows operating systems, this means that all Dell Computer's must have a Microsoft product installed before they can be shipped. Wintel lives, and it lives at Dell.

Moves like this on Microsoft's part will not endear them to system integrators or customers who like options, but for the general masses, this will go far below their radar screens.

Anyone up for quick game of monopoly?

                                     Dr. John

Slashdot

The Register

Aug 8th

NVidia Detonator 30.82

Today, perhaps in response to ATI's release of its new catalyst 2.2 drivers, NVidia is releasing its new detonator 30.82 drivers. They are available for download on NVidia's web site now. We have not had a time yet to test out the drivers to see if any performance increases have been achieved, but the last several driver releases have not offered significant performance boosts.

Rumors leaking from NVidia Central suggest that the GeForce 5 card will be over two times faster than the GF4 4600 card. That would give it a significant lead over ATI again, if NVidia can pull it off before the holiday shopping season. I'm not at all sure they can do it.

                                     Dr. John

Aug 7th

"Windows Flaw" Sounds Like a FUD-Bomb for Palladium.

Palladium is Microsoft's initiative to make the Windows operating system more secure. Palladium requires substantial changes to both operating systems and the hardware that runs them. While it is still uncertain exactly what palladium will entail, it is known that it will require new processors from Intel and AMD, and new applications running on a new operating system. Beyond that, the main point concerning palladium is that every process requires privileges to run on any machine. With that in mind, take a look at the website that discusses the hacker program known as 'Shatter'. The author points out a basic flaw in the Windows 32-bit API which allows messages from an unprivileged window to execute code in a window that has a much higher level security authorization. According to the author, this is an integral part of the Windows API and cannot be fixed without completely rewriting Windows.

All very interesting, but, it almost sounds like a plug for palladium. Aren't they basically talking about the same thing? With palladium, only 'trusted' data will be executed, and much of the internal workings of the operating system will involve keeping track of what is trusted on any given machine, and what is not. It almost makes me wonder if palladium was conceived because of the major flaw, or instead, the flaw was announced in order to drum up support for palladium. We will probably never know, but it sure it is an interesting coincidence that this flaw becomes public now.

Palladium is not intended simply to secure computers from hacking and other problems. It is also the core of the new digital rights management system that The Recording Industry of America and Motion Picture Association of America want implemented on all new computers to prevent copying music and movie files. All of this is still in the relatively early planning stages, so you've got at least a couple more years of unhindered computing. But the good old days of computing may just be that soon.

                                 Dr. John

Shatter

Internet News

PS, FUD = Fear, Uncertainty, Doubt

Dual-GPU Radeons???

A little snippet on the Internet suggests that ATI is planning, or at least contemplating, dual chip Radeon 9700 cards, much in the way 3dfx made the dual chip Voodoo Five card. Such a configuration would permit the use of quad monitors! You can bet you bottom dollar after ATI mentioned such possibilities that NVidia is looking into something similar for a maxed-out GeForce 5 card. And remember, NVidia bought 3dfx, and has many of their engineers on-board. Should be mighty interesting to see what comes of all this scrapping between ATI and NVidia.

Aug 5th

AMD to Hammer Intel: It's Official.

When the likes of Mike Magee, or small fry like myself, make comments about IT industry changes, big businesses and Wall Street dismiss the comments as so much cage rattling. But when Barron's comes out with a story about AMD's Hammer series of 64-bit computer chips, everyone sits up and takes notice.  Forget about the fact that many industry watchers and kibitzers said the same thing months ago, they aren't important.

But now that the 'scoop' has finally been made by those in authority, it's time to buy AMD stock before the "news" drives the stock price up.  While you're at it, NVidia is looking pretty affordable right now too. Of course, you may prefer to sit back and watch the ongoing meltdown on Wall Street, and keep your money under your mattress.

                                 Dr. John

PS, Those guys at Barron's want your money if you want to read their highly informative and influential article: "The Next Big Thing". Why get your news early and free when you can get it late and expensive?

The Inquirer

Aug 3rd

EULA-gy for Honest Micro$oft Service Packs

Windows 2000 users have been waiting for service pack 3 with bug fixes and security patches. Well...  Service Pack 3 is now available. But, there's a catch. As with all new Billware, you are signing over your firstborn to His Billness when you click on that innocuous "I accept" button and click on OK.  Well, not actually your firstborn, but your computers integrity from being vandalized by Bill in the future. Apparently, all new Windows updates starting with Media Player 7, heralded the new change in the EULA (end user license agreement) that now gives Bill permission to log onto your computer, analyze the contents (and presumably disable any unlicensed Billware), and proceed to "upgrade" your system without prompting or further permissions required. These new upgrades will of course continually give Bill more and more direct control over your computer, which can only be a very good thing indeed.... for Bill.

Is this the end of honest service packs from MS???

Yes. A moment of silence, please.

                                 Dr. John

Aug 2nd

ATI Winning the Graphics War

While old man NVidia slept, ATI kept plodding along with their new chip, the R300, which will be the basis for new Radeon 9700 video cards set to debut this month. ATI has also just released version 2.2 of their new, improved drivers, known as Catalyst, which provides big performance boosts with OpenGL games like Quake3. 

What does NVidia have to counter this threat? Zilch. The newest NVidia chip design, code named NV30, has not 'taped out' yet, which is industry jargon for, "it ain't ready yet!" This makes me wonder what NVidia has been doing these last 6 months since the GF4 chip 'taped out'. Let's see, making NForce2 chipsets, tossing out thousands of Xbox chips because Microsoft changed the protection codes without telling NVidia, and watching it's stock price drop in a death spiral. Ah the fickle world of finance.

So the way is clear for ATI to sweep the competition aside this holiday shopping season... unless that is, if old man NVidia awakens from his slumber and gets back to work.

                                 Dr. John

Aug 1st

Did MS Upgrade 6.0 Work?

It's hard to tell from reports, but surveys indicate that most MS license holders did not purchase Upgrade 6.0 before the July 31st deadline.  Now ask yourself, what was the deal with the deadline?  Why was there a limit to the time period where you could purchase this "upgrade package deal" from Microsoft?  They needed to keep the cash flowing during the PC slump.  PC sales were off by 13% from the first quarter, and that means 13% fewer copies of Windows pre-installed.  So Bill wants everyone to pay a bill for future software that may be produced, to save money if it ever materializes.

Two consequences are now apparent.  It didn't work very well in terms of getting the user base to pony up. Also, if even a few large corporations with many licenses did pay Bill for future wares up front, that will dent any future profits on whatever software does make it to market.  So Bill got a little cash now (like he needs it), but will have to forego some profits in the future as a result. The best laid plans of Bill and men....

So how many of you paid Bill up front for uncertain software updates in the future?

                                 Dr. John