|
|
|||||||||||||||
 |
|||||||||||||||
|
|
|||||||||||||||
|
|||||||||||||||
|
|
|||||||||||||||
|
|||||||||||||||
|
|||||||||||||||
|
|||||||||||||||
|
|||||||||||||||
|
|||||||||||||||
|
|
|||||||||||||||
|
|||||||||||||||
|
|
|||||||||||||||
| 2015 © KickAss Gear | |||||||||||||||
|
|
|||||||||||||||
|
|
| KickAss
Gear News Archive: April 2003
April 30th Bill Teel Spins Rambus Again You'd think Bill Teel would be getting tired of hawking Rambus and spinning their case after years of bad news and setbacks. But Bill is stubborn if nothing else. Bill owns lots of Rambus stock, which he bought because he thought they had invented and patented all forms of computer memory. This is of course absurd. The real story is one of sleazy businessmen and lawyers working with engineers to try to cover all forms of computer memory in a series of patent applications submitted and revised over years, while Rambus was part of a memory standardization committee. It's an old story, and I don't need to go over it again, but Bill Teel's new attempt to rehabilitate Rambus in the public's eye seems particularly absurd. His claim is that the memory standards committee JEDEC is jeopardized by the Federal Trade Commissions case against Rambus. Bill's claim is that JEDEC's common-sense rule that all members of a memory standards committee shouldn't be trying to patent any part of the standard that is being developed. Obviously, a standards committee couldn't work any other way, or they would just be unwitting facilitators of those who would seek to subvert the standards process for their profit, at the expense of the other members that were operating in good faith. But Bill's twisted take on this is that if companies are required to disclose any pertinent patent applications before participating in JEDEC, they will all leave because secretly they all want to operate just like Rambus. Bill seems to think that every company is trying to secretly patent portions of the JEDEC standard, in the hopes of squeezing huge royalties out of the other committee members after a patent is granted. Clearly, Bill's take on the world is that everyone, and every corporation, wish they could behave just like Rambus has over the years. Admittedly, corporations are in competition, but it is supposed to be fair competition, with rules that companies have to follow. Without rules, it's just a mud fight, and nobody wins. The FTC's decisions, even if they punish Rambus, will not drive companies from JEDEC because JEDEC serves an extremely important function. When common PC parts are not standardized, the consumer suffers, and in the end, the companies that produce the non-standard products may suffer as well. PC standards are necessary because many different companies have to make parts to rigid specifications. A fragmented, incompatible PC market wouldn't be good for anyone. But Bill can dream, can't he? April 28th Arctic Silver Voids AMD Warranty In what has got to be one of the dumbest PR stumbles I've ever witnessed, AMD admitted when asked that the use of good heat sinks and thermal compound on your Athlon will void the warranty. Yes, that's correct, if you upgrade your cheesy aluminum stock heat sink and wimpy thermal pad with a good copper heat sink, and use high-quality thermal compound, like Artic Silver, AMD will no longer honor the warranty. Their excuses sounded more disingenuous than a White House press briefing. Just to give you an example of how self-destructive and irrational the AMD admission is, take my own recent experience as an example. My Athlon 2700+ system was running hot with AMD's heat sink and thermal pad. In fact, it was running around 52 degrees Centigrade. I removed the stock heat sink, scrapped off the worthless thermal pad, and installed an all-copper Thermalright 800 heat sink with Arctic Silver 3. The CPU temperature dropped to 39 degrees Centigrade. My guess is that this is a corporate statement without any backbone. I don't see AMD refusing to replace bad CPUs because customers used better cooling solutions than AMD is willing to spring for with their retail, boxed chips. In fact, a little solvent will get all the Arctic Silver off of the CPU surface, so AMD would never know anyway. So basically it was just a dumb, unenforceable, and unfortunate statement that AMD will wish they hadn't uttered. April 27th Video Void You can't buy 'em for love nor money. Video cards that is. It hasn't escaped the notice of many that high-end video cards are very scarce nowadays. You can buy all the Radeon 7000 or GeForce MX cards you want, but just try and find a Radeon 9800 or even one of those wacky GeForce FX cards anywhere for sale. The shortage of muscle cards is pervasive and complete. So what's the deal? Is this the kind of scam like in 1999 when memory makers stopped making memory, sort of like some kind of memory chip OPEC? If you recall those crazy times, a stick of 128MB of SDRAM could run you almost $300. No, I don't think it has anything to do with market manipulation, despite the fact that good video cards, when you can find them, have gone up in price. That's just supply and demand. No... I'm still thinking it is the failure of the GeForce FX card. It's not that NVidia completely stopped their production, but they scaled back chip production drastically as they geared up for turning out the next chip in their line, the NV35. This left a very large vacuum in video card availability that hasn't been filled yet. NVidia put out 8x AGP versions of their GeForce4 line, but those are also in relatively short supply. ATI, on the other hand, is in a position where they need more capacity than they have, while also being stuck in the middle of a product line transition. So until both companies complete the transition to their newer chips, and can get those ramped in volume, you can expect shortages of high-end video cards. Rumors have it that both Radeon 9800 and GeForce "5" cards will be available in quantity in May, but I wouldn't hold my breath waiting. April 25th Intel Terrified by Opteron That's the conclusion I have to draw from the fact that Intel called all major computer makers with concerns that they were attending the debut of AMD's Opteron CPU. In fact, all vendors agreed that the Intel call was a veiled warning that their lucrative "CPU kickbacks" might be at risk. I'm sure most consumers don't know it, but Intel gives what would have to be called kickbacks (rebates?) to all computer makers that use Intel CPUs, and the kickbacks can be very substantial. I was always amazed at the checks we would get every quarter from Intel, back in the days when we were buying hundreds of Celerons a month. We don't sell many Intel CPUs anymore, so those checks are pretty scarce nowadays. Intel would not have even bothered pulling this kind of stunt if they were not concerned greatly about AMD's Opteron. Some of the attack stories I've seen about the uselessness of 64-bit computing over the last few days hint of Intel spin to me. But the number of positive reviews that have come out about the Opteron suggest that the industry is very eager to adopt the new, low-cost 64-bit technology. All of this is a direct threat to Intel's plan to populate coporatedom with Big-Price-Tag Big-Tin. In case you haven't been in the market for a business server recently, Intel Intanium processors can run you over $4,000 apiece. So a sub-$1000 64-bit CPU that is also 32-bit compatible can't sound like a good thing to Intel. Hence, the urgent phone calls, and the apparent panic. April 24th Opteron Rising, Itanic Sinking The early reports on the Opteron 32-bit/64-bit processor from AMD have been very positive. Currently, most benchmark comparisons between AMD and Intel involve benchmarking Opteron vs. Xeon processors. Here, Opteron clearly wins, despite its significantly lower price range. A set of benchmarks released by Microsoft shows a 4-way Opteron system beating a 4-way Xeon system from Intel by good margins. However, Xeon is just an expensive 32-bit CPU, so it really isn't a good comparison with the Opteron. The Intel Itanium is a true 64-bit CPU, with only wimpy 32-bit emulation. So all the hoopla about 64-bit computer for the masses has Intel worried. Just days after the Opteron release, Intel is talking about beefing up the 32-bit emulator in the Itanium. But putting lipstick on a pig has never been very enticing. AMD's Opteron has native abilities to run both 32-bit and 64-bit code, something Intel still feels is not demanded by the marketplace. That may change as more benchmark comparisons come out. It will be very interesting to see how the Opteron and Itanium stack up against each other running in 64-bit mode. My guess is the price-performance ratio on the Opteron will be far better than that with the Itanic. Intel will be reducing the price of Itanium CPUs, which also seems to be an indication that Opteron is worrisome for purveyors of Big Tin. April 23rd Opteron Bugged AMD has released it's official list of bugs, or errata, in the Opteron CPU. It's not a huge list, and most of the bugs seem minor, but it does mean AMD has more work to do to iron the kinks out of their new CPU line. I'm very glad to see them fessing up to "errata", and hope they can get them all fixed to everyone's satisfaction. April 22nd Opteron Benched, Baked, Beaten and Begun The long awaited day is here. The AMD Opteron chip has been officially released. It's not really a big story for home PC users, because the Opteron is geared to low-cost 2-way and 4-way business servers. But we do get a peak at how a 1.8GHz Opteron performs in single and dual-CPU systems. The numbers look quite good for a sub-2GHz chip. Indeed, if AMD can scale the Opteron to 2.5 GHz or higher, they will have a very nice product. And that is probably the biggest question on many people's minds. Can AMD debut the home version, known as the Athlon64, at speeds around 2.5GHz? If so, they may have an excellent contender this Fall. If you want to check out some of the benchmarks, have a look at the links below. April 21st How Popular Will the Opteron Be? That is the question that many people have been asking for the last year or more. On paper, it seems like a mixed bag. The fact that it is the first 32-bit/64-bit processor is very good, because it is both backward and forward compatible. But the actual clock speeds on the first releases look low, even when compared with the "Barton" version of the Athlon. That's bad. But with their newer technologies, AMD should be able to ramp speeds fairly well, so the lower debut speeds may not be so bad. Then there is the currently existing compatible 64-bit software support, which is fairly good so far, especially on the Linux front. In contrast, there is the fact that the 64-bit version of Windows for the Opteron will not be ready for some time, which is bad for the debut. Next is the fact that AMD has a good new chipset out for the Opteron, but the number of existing motherboards, and the number of third party chipsets, are very slim indeed at the present time. Tomorrow is the big day. The Opteron is in the supply chain, and motherboards are on the way. Right now you'll need to use a 64-bit compatible version of Linux to see how the Opteron works in 64-bit mode. And even if you can't test one yourself, you can expect a whole bunch of reviews to hit the web tonight at midnight, so by then we will know how the Opteron stacks up to other 32-bit and 64-bit processors. But this is really just the business debut of AMD's 64-bit initiative. The Athlon-64 for home computers will not be released until Fall, when many new motherboards, and the 64-bit version of Windows should be ready to go. You can also expect clock speeds to be substantially higher by then as well. April 18th Will Your Next Vacation be Suborbital?
April 16th Microsoft Shoots Self in Foot.... again For the last several years, Microsoft has been slowly adding more and more copy protection, and mandatory registration routines to it's software, including installing such routines with service packs and updates to existing applications or operating systems. For the most part, these wonderfully irritating routines are foisted on the general public, while corporate users are given more of a break from these "time-is-money"-wasting irritants. But the worm has turned, and now through some typical MS cockup, corporate users of Office 2000 "Select Customer" version SR-1 running under Windows 2000 are finding the office suite has been rendered useless by the integrated copy protection scheme. After a certain date, which I still can't ascertain, the registration wizard pops up, and the typical inclination of folks who've already registered is to click the "register later" option. But with this glitch, by the time you click "register later" the 50th time, Office 2000 will fail to load. Microsoft's quick fix for this business-halting bug is to set the computer date back two years! Such remedies hint at a time-limit on SR-1 installations that has gone amuck. If Corporate America has half a brain in it's collective head, it will slowly steer as far clear of Microsoft products as possible. On top of the abysmal security, the onerous copy protection is now costing businesses real money. And MS isn't about to stop, or be more careful about implementing copy protection. They don't need to as long as almost all businesses use Office for their daily work. Years ago, as IBM's OS/2 was dying a slow death, Microsoft really was the only other game in town. Now, that is no longer true. Linux and Star Office may lack the glitz and density of Windows and MS Office, but they are stable, secure, functional and far cheaper to implement and maintain. I can only imagine that corporations that had been sitting on the fence are now starting to lean towards Linux, and it's complete lack of spyware, registration routines, and copy protection. PS, There is another patch for Windows NT, 2000, XP out from Microsoft.... severity: extreme. You can get your latest MS patch fix here. But you've got to ask yourself, "do you feel lucky punk?" Or is this going to install another copy protection nightmare while it purportedly fixes a glaring security problem? Your call. April 15th Canterwood Mystery According to the esteemed Mike Magee of The Inquirer, there is no "bug" in the Canterwood 800MHz version of the Pentium 4. Intel allegedly made that bit up to cover other goings on with the debut of the new version of the chip. Magee posits that production and marketing issues dashed the debut. Indeed, it would not be beyond Intel's repertoire to "debut" a chip that isn't available in enough quantity to fill a thimble. It is no small coincidence that AMD will be releasing their much hyped, much anticipated "Opteron" 64-bit processor one week from today. The end result for Intel is a very un-savvy CPU debut. April 15th "Canterwood" vs. "Opteron" Yesterday Intel released the "Canterwood" Pentium 4 with a quad-pumped 800MHz front side bus frequency. Along with it they introduced Intel's new 875P chipset. The reviews posted on the net yesterday were impressive, despite the fact that the 3 GHz Canterwood only beat the "Barton" version of AMD's Athlon 3000+ by fairly small margins in many tests. But what some analysts are wondering is how AMD's upcoming Opteron processor, to be released on April 22nd, will stack up against the Canterwood P4. It seemed like a big win for Intel when they announced the Canterwood just a week before AMD's Opteron release. But then things started to unravel for Intel. By yesterday afternoon here in the States, reports started to appear on the web that there was a severe shortage of Canterwood P4 chips, a severe lack of properly rated memory modules for 800MHz operation, and a limited number of 875P motherboards. Then a rumor began to circulate that there was some kind of problem with the 800MHz bus, but it was not clear if the problem was with the chipset, the P4 chip itself, or with memory modules. This rumor turned out to be true, and Intel said it was a problem with "a small number of processors". Immediate comparisons were made with previous botched Intel processor debuts. To me this is an obvious example of the marketers over-ruling the engineers, forcing a half-baked product to market too soon to try to trump AMD's Opteron release. Unfortunately for Intel, this paper debut seems to have backfired in a very bad way. It's doubly unfortunate because the Canterwood and 875P seem like a really excellent bit of technology, but technology that isn't quite ready for a public debut. All I can say at this point is, I will be very disappointed if AMD tries the same kind of nonsense with the Opteron release. They have set the April 22nd debut date in stone for the last several months, and I expect them to stick to it, or suffer the slings and arrows of hack reporters like myself. April 12th Opteron is Go! AMD's new 64-bit/32-bit hybrid chip designed for servers, and known as the "Opteron", is ready for release. The chip is starting to show up in the supply channel in both tray and boxed versions. Currently, only versions of Linux can run the Opteron in 64-bit mode, but versions of Windows will be available for the new chip within months. Now we get to see how well the Opteron will be adopted by corporations as compared with the dramatically more expensive (and probably much more sluggish) Intel Itanium, known in the industry as the Itanic due to it's tragic fate thus far. My guess is that the Opteron's fate will be far rosier than that of the still sinking Intel Itanic. Later this year AMD will introduce a version of the chip designed for home use, which will be called the Athlon-64. By then, a version of Windows will be ready to go along with the new consumer-level chip. Of more concern to potential buyers of the chip will be finding a good motherboard to go with them. Currently the only chipset that works with the Opteron is the AMD 8111/8131/8151 chipset. Expect more 64-bit chipsets from the usual suspects over the next several months. April 10th Microsoft Making Friends, Not Money Those of you who know Microsoft's modus operandi will undoubtedly find the headline of this piece questionable. Excuse my sarcasm, but I couldn't resist. A number of stories out recently have me wondering if Microsoft has any real friends left, other than the US Dept of Justice. The first piece of news was another in a long series of security bulletins from Microsoft that warn users of newly discovered flaws in Microsoft Proxy Server 2.0 and Internet Security and Acceleration (ISA) Server 2000, and another problem in MS Virtual Machine. How long will companies just keep blindly applying patches before they begin to truly distrust MS software? The second piece of news that can't be good news for Microsoft is that the state of Oregon has a pending bill that would require state agencies to consider open-source software when purchasing for their computer network software. Microsoft has gone to great lengths every time a foreign country or US State announces that they are "looking at open source software" to dump millions of dollars of proprietary software on the upstarts, in order to make sure they stick with Microsoft. So I suspect that State legislators in Oregon will be getting some phone calls in the next day or two, with lots of free software in the offing. Can Bill afford to do this with every State and Country that threatens "open-source"? The final news story that has to be irking Bill recently is the gripe piece by Ed Foster at InfoWorld. Mr. Foster has a little expose' on Microsoft's much-hated "License 6.0" volume licensing program for businesses. According to the article, businesses that were duped into purchasing the volume licenses will have to keep paying the 3 years of payments for "software updates", even if they sell off the computers that they purchased the volume licenses for. Here is a quote from an IT consultant: "In other words, you become obligated to pay for the years that remain on the SA agreement covering those computers, but that agreement is terminated and Microsoft provides no further upgrades under it. The acquiring party just gets the rights to the current versions you already had on those computers... It will always come down to 'Let's work something out,' " said one. "Translated, that means, 'You have to give Microsoft a lot more money." Many analysts wrote warning articles about the onerous nature of the "Licensing 6.0" scheme before it was even implemented, so any company that was dumb enough to buy the damn things only have themselves to blame for not doing more research before forking over the cash. One has to wonder how many negative stories MS can weather without damage. The number of bad-news articles surfacing on Microsoft far exceed negative stories about other corporations, and eventually, you would think this would erode corporate confidence in Microsoft. But that hasn't happened yet. And yet, the constant 'water torture' dripping must surely be having an effect on IT managers and system administrators everywhere. April 9th 400MHz Bus Athlons Despite the fact that NForce2 and VIA KT400A chipsets and motherboards are basically new, it looks like the 333MHz Athlon bus was a very short-lived, stop-gap measure. AMD is about to come out with 400MHz CPUs, and chipset companies are coming out with new chipsets that officially support the new 400MHz bus speed. Of course current NForce2 and KT400A chipsets have "unofficial" 400MHz bus support, so the new chipsets are probably just new steppings of the same chips. This increase in bus speed also means that PC2700 DDR DRAM will have to be replaced by PC 3200 or PC3500 memory, which are designed to run at 400MHz and beyond. But based on overclocking tests with chips like the new Barton 2500+, some current Athlons can already be run at between 190 and 200MHz on the bus. So this change looks like a simple rejiggering of multipliers and timing settings, to achieve stability with higher bus speeds. Stupid Security Awards I don't know if you all are feeling a little uneasy about the brave new world being offered us by security experts who say that terrorists are everywhere waiting to get you, but some people have actually started handing out awards for the most stupid and pointless security measures put in place. You can find the "winners" of the stupid securities awards here. "Privacy International" has been mulling over thousands of submissions, and has picked the most egregious violators of common sense and logic. Indeed, it would appear that the security measures themselves are far more likely to cause problems than terrorists, who can now sit back on some tropical island sipping cocktails, and laugh at how everyone is jumping through hoops, even in the complete absence of any threat. I especially liked the sign at an airport with a cartoon of a bomb with a red circle and slash over it. That should keep the terrorists out. April 8th New GeForce Card Near? The word is that NVidia has pushed forward the release of their next video card, after the last one failed miserably. In my memory, this is the first time a major video card maker has had to virtually pull a product before it was even released, and certainly is the first time such a well-hyped video card failed before launch. But if NVidia does make it to market with the new card this month, it will be a logistical miracle. They will have turned their entire manufacturing process on a dime, and shifted into overdrive to get the new cards out in record time. And just in time, as ATI just released the Radeon 9800 Pro. The new ATI card just showed up at distributors web sites yesterday, with an ETA of next week sometime. Quantities look extremely limited. I expect that the new NVidia cards will be very scarce as well, at least for another month or two. There is a lot of scuttlebutt about what the new NVidia card will be called, but I doubt that any of the names I've heard are correct. Typically, marketers fret and argue over the number that goes after the name (like Radeon '9800'), so they may be arguing over whether to give the card an even higher arbitrary number than ATI's best card. GeForce FX 9900? 10,000?, GeForce FX2? Who knows, but you can bet they are still arguing about the meaningless number that comes after the silly name. Opteron Software Support Strong After weeks of speculation about how much support the 64-bit Opteron processor from AMD would get at launch, we now find that the support from the software industry is exceptionally strong. Many versions of Linux are already ready, and a version of Windows is on the way. Apache HTTP server is available already, and there is plenty of server support upcoming. If the application support is half as strong at launch as the development tools and server application support, then the Opteron will have far better support at launch than Intel's Itanium... Ouch, that's got to hurt. Intel and VIA End Legal Battle Intel and VIA have been battling in court for years over various patents each company owns. It's the usual IP nonsense that has become a standard operating procedure for so many IT firms. The only guaranteed winners are the attorneys. Well, the court fight is over, with both sides shaking hands and acting like adults for a real change. Very few IP wars end this way, and it has me wondering if battle-torn companies are getting tired of forking out so much cash to overpaid legal teams. After all, a simple hand shake can save millions of dollars in legal fees, so you have to wonder why such outcomes are so rare in IT land. April 7th
The Inquirer has a really unlikely rumor about Microsoft trying to pay id software to make the upcoming game "Doom III" into an "X-box-only" game title. The rumor is rumored to have appeared in Wired magazine, but the online version sure doesn't have it. So maybe this is just a rumor about a rumor, rather than a real rumor. With rumors, it's so hard to tell. SARS Threatens Your Next Computer The much-feared, and not-so-deadly Severe Acute Respiratory Syndrome (wait... isn't this a disease, rather than another syndrome?), or SARS, is affecting computer parts supply chains all over the far East. And because virtually all PC hardware comes from that part of the world (wait... isn't that a National Security Risk?), you can bet that getting computer parts is going to get tougher as long as this respiratory infection continues to spread. So if you are getting ready to put together a new system, you might want to make sure the parts you need are actually available. You can't get very far if the motherboard you want is held up in Taiwan because of uncertainties about the spread of the virus. April 4th Is Bill Gates Dead? No, despite repeated attempts by Russian psychic assassination teams hooked up to brain wave amplifiers (I read about that in a comic book). But a recent spoof report on the internet which said that Bill Gates had been shot in LA sent the Korean stock market into a tail spin. Considering the number of spoof sites on the internet that talk about Bill Gates death (e.g., click here), and the fact that a movie has been made on the subject, I'm amazed that anyone believes these rumors anymore, let alone an entire country's stock market. At this point, even if Bill Gates were shot, reporters would assume it was a joke for at least a few days. I guess they aren't that web savvy in Korea yet. April 2nd Microsoft Readies Win-64 for Opteron Release The good news for AMD today is that Microsoft will announce on April 8th that they do have a 64-bit version of Windows ready for the launch of the AMD Opteron this Spring. The operating system is code-named "Anvil", and will work with all AMD 64-bit processors (Opteron and Athlon 64). According to Microsoft, the new operating system performs very well even when running older 32-bit applications and games, and even better when running new 64-bit programs. The real question now is how many software companies will be able to get 64-bit versions of their software up and running before the release? Probably not many, but you can expect 64-bit versions of many applications to slowly |
A company known as