8th > June > 1999 Archive

The Register breaking news

Pfeiffer positions Alpha against Merced

A year ago From The Register a year back Compaq has kicked sand in the face of its partner Intel and will ship its high volume servers using the Alpha platform and Digital's 64-bit Unix. The company claims that will make the fast chip an industry standard, positioning it squarely against Intel's IA64 Merced platform. Eckhard Pfeiffer, CEO of Compaq, made the announcement during a conference call today and spelt out details of an alliance it has made with Microsoft to ensure interoperability between Unix and Windows NT. Pfeiffer said: "I've made a commitment to our customers that we'll continue to support Alpha. Compaq will use Alpha and Digital Unix in its volume platform and establish them as an industry standard. Industry standards lie behind two of the most important developments, which are NT and the Internet." Pfeiffer's comments were amplified by a senior vice president of the company, John Rose. He said: "Our strategy is to drive technology into standards. We have a rich history of driving industry standards." He said: "In 1992 Digital introduced 64-bit computing. The Alpha promises a five year headstart and is coupled with 64-bit Unix performance. Along with Intel, Samsung and AMD, Compaq will drive the Alpha as the industry standard for 64-bit computing." He said that Compaq would ensure customers had a high level of integration between the NT and Unix and was working with Microsoft to ensure interoperability between both. "Compaq is the leader in NT systems and our market share overall ranges between 40 to 70 per cent, including SAP, Baan, Peoplesoft and many others," he said. Compaq will invest in middleware to ensure NT and Unix worked together and will package both services and solutions. "We'll package services and solutions," he said. "Our customers can depend on the support of over 2,000 Microsoft certified network engineers and thousands of Unix specialists. This year, IDC projects that 72 per cent of enterprise appliances will be shrink wrapped. Active Answers is our next generation of enterprise solutions. The online library of tools will provide multivendors solutions." Intel reacted coolly to the news. A representative said: "Compaq is a very important customers and we work closely with them on the whole of its product line. We don't see this as a race against Alpha, we make Alpha processors and we give support to them through that." He added that Compaq was working closely with Intel on its IA-64 initiative. ®
Mike Magee, 08 Jun 1999
The Register breaking news

Judge beats up Intel in Intergraph case

Intel may have been let off the hook by the Federal Trade Commission (FTC) but its antitrust woes are not at an end. Alabama judge Edwin Wilson yesterday denied the chip giant a summary judgement to end the case between it and Intergraph. Intergraph is pursuing its case against Intel because it claims it uses monopolistic ways …
Mike Magee, 08 Jun 1999
The Register breaking news

Battle royal breaks out over Russian chip claim

A bitter argument has broken out over a claim that Russian technologists were the first to develop superscalar microprocessors, later adopted by Intel. Mark Smotherman, writing from a US university, said: "I would like to correct, for the record, the statement in Intel uses Russia military technologies" by Andrei Fatkullin in which he says: "Superscalar architecture was invented in Russia." "The first superscalar design was the IBM ACS-1 supercomputer,designed in Menlo Park, California, in the mid-1960's at IBM's Advanced Computing Systems by a team that included John Cocke. In fact, the vision for a computer that decoded and issued multiple instructions per cycle was due to Cocke. "The ACS-1, which was cancelled in 1968, would have issued up to seven instructions per cycle: three index register instructions, three floating-point register instructions, and one branch. The floating-point register instructions could be issued out-of-order from an eight-entry "contender stack" that was scanned on each cycle. "Each floating-point register had a "backup register" so that a new value could be loaded into the backup register while dependencies still existed with the previous value. (This is a simple form of register renaming). "There was a condition register with 24 individually-addressed bits so that independent compare instructions could be executed in parallel. Moreover, each instruction included a skip flag so that the instruction would be executed or not based on a current skip condition, which was specified as a logical operation between two condition bits. (This is a simple form of predication). " Branches were implemented using two types of instructions: a "branch" instruction that evaluated a logical operation between two condition bits and placed that result along with a branch target address into a branch table; and, an "exit" instruction that selected the first tabled branch that evaluated to true.(This is a form of multiway branching.) "Simulations of the ACS-1 on the innermost loops of a scientific code such as Newtonian Diffusion showed a sustained instructions-per-cycle value of 1.8. On this code, which contained branches within the body of the innermost loop, a 100 MHz ACS-1 would have been 76 times faster than the CDC 6600 and 22 times faster than the IBM S/360 Model 91. " The ACS-1 was described in the "Western literature" in 1971 (see reference 3)." However, Alexei Pylkin, aprogrammer-researcher at the Supercomputer Software Department, Russian Academy of Sciences Novosibirsk, Akademgorodok, has refuted Smotherman's argument. He said: "I'd like to note that ACS-1 was experemental design while Elbrus-1 was a mainstream computer. Serial production of Elbrus-1 started in 1977 but research on superscalar architecture was done in Russia in the same time frame as in IBM. "Experimental ACS-1 had no successors. The second Western superscalar processor has appeared only in 1992. Serially produced Elbrus-1 had successors - serially produced Elbrus-2 (1984), microprocessor El-90 (1990, prototypes)." Smotherman produced references to support his argument, which follow: 1. John Cocke, "The search for performance in scientific processors," Communications of the ACM, vol. 31, no. 3, March 1988, pp. 250-253. 2. Emerson Pugh, Lyle Johnson, and John Palmer. IBM's 360 and Early 370 Systems. Cambridge, MA: MIT Press, 1991. [see section 7.5, which covers ACS] 3. Herb Schorr, "Design principles for a high-performance system," Proceedings of the Symposium on Computers and Automata, New York, New York, April 1971, pp. 165-192. 4. Mark Smotherman, "IBM Advanced Computing Systems -- A Secret 1960's Supercomputer Project," available online. ®
Mike Magee, 08 Jun 1999
The Register breaking news

Via revenues soar on PC-133 wave

Taiwanese chipset firm Via said today that it sold NT$648.7 million of its kit in the period up to May 1999. That represents a rise in revenues of 218 per cent compared to the same period last year. On a sequential basis, claimed Via, its sales rose by 65 per cent from January to May 1999. Net sales in thousands of NT dollars amounted to 648,694 in May 1999, compared to 204,181 in May 1998. January through May 1999 sales amounted to 2,896,829, compared to 1,756,079 compared to the same period last year. Again, sales are in thousands of NT dollars. Meanwhile, Rambus stocks rose on Wall Street, as analysts sought to bolster its long term position. Intel UK, yesterday, maintained that Rambus was still its preferred future memory technology. It is waiting for its delayed Camino 820 chipset to arrive and is hoping that there will be enough RIMMs to make everything fit together. However, it must be concerned that Mosel Vitelic, the biggest Taiwanese memory technology company, pointed to technical problems with Direct Rambus when it announced support for PC-133 SDRAM at Computex last week. According to Mosel Vitelic, PC OEMs are demanding PC-133 rather than Direct Rambus solutions, a point Intel has so far not addressed. Last week, an Intel representative at Computex confirmed yields on Rambus RIMMs would not rise until the end of the year. Meanwhile, memory company Apacer, an Acer subsidiary, also made the same point and said it was pushing PC-133 hard. ® Computex 99 coverage
Mike Magee, 08 Jun 1999
The Register breaking news

Merced tape out resembles Desperation Derby

The question whether Merced has taped out or not is now becoming something of a burning issue, with Intel still refusing to give details on where it is on the 64-bit chip platform. Yesterday, we reported that architects were divided over the viability of the platform, with one British scientist claiming that Merced will run slower and hotter than IA-32 architecture. Earlier this year, Stephen Smith, who runs the Merced programme at Intel in the US, said silicon samples would be ready in June. But only two months ago, Craig Barrett, CEO of Intel, seemed to suggest that such samples will now be late. The official Intel stance from the PR department is that the platform is still on target, with Merced sampling in "mid 1999". Earlier this year, insider reports reached our ears that a tapeout had been ordered by May 27th. Meanwhile, Henrik Klagges, an IT analyst at Strategy Partners, has confirmed there is considerable disquiet about Merced's speed in the wider community. He said: "A substantial number of people (including me) have said so, last not least in comp.arch, but also in public talks (e.g. DECUS) and the SAP user group etc. I even have a public bet outstanding for a bottle of champagne that when Merced ships, it will be slower than x86 on SPECint." ®
Mike Magee, 08 Jun 1999
The Register breaking news

MS economist gets fail mark from his old prof

MS on Trial Government economist witness Franklin Fisher did a better job in rebuttal than he did in his primary evidence, but he still showed that he doesn't know nearly enough about the industry. And Microsoft's cross-examination was designed to test how little he knew, rather than to counter his arguments. It was nothing but ritual silliness, a game played by highly-paid witnesses, to claim that black was white, or vice versa, depending on which side they were on. In some circles, it might have been called "bearing false witness". Most of the weaknesses in the Microsoft testimony would have been more convincingly criticised by somebody from the industry, but because of some Spanish customs in the witnessing business in antitrust cases, economists have it stitched up. Since there was no written direct testimony, Fisher was first examined by David Boies, the DoJ's special trial counsel, and already the cult figure in the case, thanks to some useful one liners on the courthouse steps, and some humour in court. Fisher hadn't been allowed by court rules to see written testimony by economist Richard Schmalensee for Microsoft before he was cross-examined, so Boies dwelt on what Fisher had found out about his former student's errors. The MS student's goofs It turned out there were quite a few mistakes, and that today, Schmalensee would not pass Fisher's course in antitrust economics. The big flaw with Schmalensee's argument had already been flagged: his refusal to accept that a market definition was needed. The danger of a definition of the operating systems market for Microsoft would be that it could be proved to be a monopoly with almost any definition. Fisher put the boot into Schmalensee for confusing entry into the applications software market as being easy (it is) with entry into the PC operating systems software market (it's almost impossible). There's no "demand substitutability" said Fisher, triumphantly - and he was right. And there's no "supply substitutability" either, he protested, because of the applications barrier. Right again, and Fisher had some reasons for this. First of all, Microsoft's customers didn't think there was an alternative to Windows. Microsoft could also change the spec of Windows and put up the price (it's already done that Prof, but never mind). Then Microsoft is not constrained by non-Microsoft operating systems. Next, Microsoft's share of the market is high and has been high for many years. Reason five was the applications barrier, and reason six, a damning one, is that Microsoft's conduct can only be explained by postulating that it behaved as it had because it was protecting its monopoly. This was all good stuff. Schmalensee, said Fisher, had been non-relevant, looking into competition generally, and comparing it with other industries. It was silly to refuse to define a market, and then to say that Netscape and Java both were in the OS market. Because Microsoft has monopoly power, which it abused, it took action against Netscape and Java, which it would not have done if they weren't a threat to its monopoly. Arguing for the danger of an unknown threat "out there" sounded rather like whistling for a wind. Fisher also thought that Microsoft might believe it would hasten the day of judgement if it charged high prices now, and did not like looking at possible future constraints as an analytically correct way to examine the nature of any constraints. This was in fact a repeat of his fourth argument, but at least Fisher was scoring points with what appeared to be rehearsed responses. Predatory pricing? Microsoft's losses from its predatory pricing would be recouped, Fisher said, and the time it might take Microsoft to recoup these losses did not relate to whether Microsoft had monopoly power. In any event, he thought, Microsoft would get its money back quickly. No new firm had entered the PC operating systems business and established a 5 per cent share since 1991, Fisher agreed, which meant that in such a profitable market there was an extreme entry barrier. Schmalensee was swatted down by Fisher for a number of misleading and non-relevant statements. No evidence of Microsoft controlling software distribution was presented, because it is not relevant to the principal claim of the case, Fisher noted. Unlike Schmalensee, he thought that "whether Microsoft has monopoly power over the OS" is relevant, in contrast with his opponent's claim: Schmalensee had failed to demonstrate the non-relevance, and had not understood the issue of predatory behaviour, his former mentor said. It was no excuse in a winner-take-all world for Microsoft to be allowed to do anything it wanted to reach that position: only competition on merits was permissible. Fisher launched into an analysis of predatory acts, and how anti-competitive actions like giving away an expensive product free - stand-alone IE - only made sense if it were done to protect a monopoly. He warmed to the issue: "Microsoft gave away Internet Explorer. It was to be forever free. Microsoft's documents describe it correctly as "a no-revenue product." This was a product which Microsoft not only gave away for free, but basically bribed people to take, giving preferred places on the desktop for which a charge could have been made. But beyond that, Microsoft also spent hundreds of millions of dollars on the development of this no-revenue product, and then gave away the technology. That is not a profitable act, except for the protection of the operating system's monopoly, Fisher opined. Allchin used as club Michael Lacovara, for Microsoft, did his best to stop the pillage by jumping up occasionally to object, but to no avail. He became very twitchy when Boies started on a line of questioning about the "advantages to consumers from the combination of Windows 98 and IE", and how that varied from having Windows 95 and IE separately. Boies dug out Allchin's trial testimony admission, to help, and asked Fisher to comment: Boies [on 1 February 1999 to Jim Allchin of Microsoft] : The question was: "And you agree that you can get those benefits either by buying Windows 98 or by having purchased an original retail version of Windows 95 to which you added IE 4, either downloaded or bought from retail or gotten in some other way; correct? Allchin: I do. [Fisher commented that the consumer benefits, such as they are, that come with the integration, the welding, the selling together, the bundling of IE - in this case IE4 and Windows operating system - don't require any of those actions. The consumer gets the same benefits if the two things come separately. In that event, there is no reason why Microsoft shouldn't offer them separately and let consumers decide.] Boies: And it was entirely practical for you to deliver these two pieces of functionality, if that's what you want to call them, separately; correct, sir? Allchin: Yes. That's the way we do development. We develop lots of different pieces concurrently, separately. [Fisher said that this suggests extremely strongly that not only are there no particular consumer benefits from getting the things together, as opposed to getting them separately, but there are no particular advantages.] Boies: It's just a distribution vehicle, correct, sir? Allchin: It's the same code out of Windows. Boies: It's the same code and all we're talking about are different distribution vehicles, in your words; correct, sir? Allchin: Yes. That's what I said, yes." Boies had trapped Allchin, and destroyed any effective argument that Microsoft might produce to defend its welding operation. Judge Jackson called "time", and the court took a break. But it was scarcely noticed that the judge asked Bill Neukom, Microsoft's head lawyer, to approach the bench for an off-the-record discussion. We have no desire to be reincarnated as a fly, but oh to have been a fly on the bench at that moment. Was the judge suggesting that Microsoft should throw in the towel? Or was he just enquiring why Neukom looked so tense? When Microsoft gave away its browser to Apple (in fact insisted that Apple took it), it made a grave mistake in that it could not have increased Microsoft's ancillary revenue, Fisher said. Nor did it help that Microsoft, according to Schmalensee, kept track of operating system sales "by hand on sheets of paper", and so could not calculate ancillary revenues or track the business: it was just protecting its monopoly. This is the kind of accounting that would normally attract attention from the IRS, since it is hard to believe that this is truly what happens. It will be interesting to see if Microsoft's auditors qualify Microsoft's accounts as a result. Fisher mentioned Gates' Newsweek article gaffe where he had said that the PC market will not be easily displaced - and agreed with him. Joachim Kempin was quoted to Fisher saying that Microsoft did not have to worry about OS competitors, effectively because of the applications barrier. The procedure used to price Windows 98 was a perfect example of monopoly pricing. Boies led Fisher into pointing out where Schmalensee had made "a big mistake" in his arithmetic that attempted to show that Windows was underpriced. After correcting Schmalensee's error, Fisher derived the actual price being charged at present. The biggest dispute concerned which data set for browser use should be used as a basis for the statistical games. The problem was between the actual usage of browsers from a large data set (the Adknowledge data), or a small survey of the browser used, according to heads of households (the MDC data). It was not hard to see that the former was better, and that the latter, used by Microsoft of course, was both erroneous from a statistical design point of view, and gave an imperfect view of browser distribution. Damned lies and statistics Unfortunately for Microsoft, it turned out that Judge Jackson had had at least one statistics course, and was quite at home with the idea that Microsoft's sample size was too small to have statistical validity, because of the confidence level at 95 per cent probability. Some previously unconsidered data from Zona was introduced by Boies. It suggested that, contrary to what Microsoft had tried to show with its data (that Netscape was gaining market share last year), in fact Netscape's browser share was decreasing amongst enterprises, and that corporations were not supporting Netscape in its ambition to dominate that market sector. Consequently, Microsoft lawyer John Warden's claim that the DoJ case was in trouble as a result of the AOL merger was ill-founded. Fisher's view was that it didn't have any bearing on Microsoft's monopoly. Fisher (undoubtedly as a result of a DoJ briefing) drew attention to errors in a Schmalensee exhibit that exaggerated Netscape's browser market share by 9 per cent, because contrary to Schmalensee's claims, Netscape's browser was not loaded on Packard Bell's consumer PCs, but only on one business line. It seemed that this was to be Fisher's revenge for having been set up. Microsoft witness Brad Chase's wild statements about the number of users that he thought had downloaded Netscape were dismissed as "foolish". As the direct examination of Fisher drew to a conclusion, Boies ensured that some key comments were added to the record. Fisher said that the case was not about Netscape, but about Microsoft's protection of its monopoly in operating systems, and the extent to which Microsoft succeeded in preventing a platform threat from materialising. The issue of where consumers stood in all this was also examined. Users were affected because their choice was limited, Fisher said. He also corrected a misquotation of his remarks about whether consumers had been injured already. He maintained that they had not, but some media did not trouble to include the reason for this, and reported the opposite of what he intended. The zero pricing of IE does not harm consumers at present, but after the predatory period, there would be harm. The central proposition of antitrust policy is that competition is good for consumers, and harm arises when competition is removed. So far as innovation was concerned, Fisher had some strong criticism of Microsoft's actions: "Microsoft has been giving out very, very strong signals that innovation is fine and they will cooperate with it. They will even assist it, if what you're doing is producing simply complements for Microsoft products. But they will take very, very aggressive action against you if what you're doing is producing innovations that might lead to something that threatens their operating system monopoly." So true. ® Complete Register trial coverage
Graham Lea, 08 Jun 1999
The Register breaking news

Did Compaq and MS collude to ambush DoJ witness?

MS on Trial Franklin Fisher, the DoJ's economics expert witness, was made one of the victims of an elaborate sting operation by Microsoft, it would appear. It seems that Netscape was probably manipulated into paying Compaq to load its browser on new models, as part of a plan at least partly devised by Microsoft. Fisher was made to look particularly ill-informed when in evidence given in early January he did not know that Compaq had announced it was putting the Netscape browser on every new model of desktop. Consequently, Fisher had used earlier data provided by Netscape. It now transpires that Compaq first announced that it would offer the Netscape browser at a very important juncture in the trial. The transcript shows that Fisher said in his rebuttal last week: "In fact, the Compaq arrangement to put Netscape on the desktop was announced on January the 5th, the day that I took the stand, and therefore was not part of the information I could possibly have had before I was on the stand. The deal was consummated in the middle of the trial. "More substantively than that, what this does not tell you is that Netscape is actually paying Compaq in order to get its browser on the desktop. It was paying them advertising, something supposed to be worth over $700,000. "Now, there isn't any doubt, I suppose, that if Netscape were willing to pay sufficient money, it could, in fact, get OEM's to put it on the desktop. That would not mean that it is not severely disadvantaged. That's called raising rivals' costs. And it is, at least, misleading to suggest that gee, this is just happening." We suspect that the following occurred. First, Microsoft wanted to discredit Fisher's evidence; to neutralise with FUD as far as possible the harm done by some testimony from Compaq witnesses that reflected very badly on Microsoft; and to show what tough competition Microsoft faced when the number one PC maker "decided" to offer Navigator. So when it came about, either by Microsoft's design or by chance, that Netscape offered Compaq $700,000 to get its browser on new Compaq models, Compaq decided to go ahead. What we don't know with certainty is the degree of collusion between Microsoft and Compaq. There could have been none, but the accident of a deal being announced the day that Fisher's evidence started is too great to make that a serious possibility. With Netscape very busy with plans for the AOL merger, it is less probable that the initiative came from Netscape, although it is possible. On the available evidence at the moment, we favour a scenario where Microsoft suggested to Compaq that it approached Netscape with the offer of putting up its browser for some advertising money. Netscape would agree; Compaq would not upset Microsoft because the idea came from Microsoft; and Compaq would benefit financially. The beauty was that Netscape would pay for it all, and no under-the-counter payments from Microsoft to Compaq would ever be discovered. Clearly the DoJ knows more than it has fed to Fisher so far, and we may discover more from the redirect testimony. There is no possibility that Fisher worked this out for himself. Re-reading the transcript of Fisher's primary cross-examination in January supports the hypothesis of Microsoft's intervention in several ways. Lacovara said in his cross-examination of Fisher on 13 January 1999: "Let me suggest to you that there is at least another source of information that you may not know of, which presumably is the PC OEMs themselves. And so I decided yesterday to go to the web site of Compaq Computer and to see exactly what they ship on desktops today. Now, you understand that Compaq is the largest OEM by far ... It accounts for roughly 20 percent of the units shipped in this country ... let me offer into evidence ... pages from the Compaq web site that I located and then my colleague here printed under my direction yesterday. You will see the date 1/12/99 [the previous day] in the lower right-hand corner." It is beyond reason that Microsoft or its legal team could have stumbled into this. It was orchestrated in some way, as is seen from Lacovara's statement that: "I asked someone to go to CompUSA to verify that where Netscape Communicator appears in the model number, the software is actually on the desktop and not somewhere else in the machine." This seems to confirm it was all a fix, because Microsoft made the mistake of going over the top with the "evidence". Lacovara disingenuously asked Fisher if he'd thought of looking at OEM web sites to determine if Netscape was used. This was of course a silly suggestion, since Netscape knew precisely what agreements it had. Microsoft had set out to exploit the time interval between Netscape's evidence in October, and Fisher's evidence in January. It was a trick, tantamount to misleading the court. Fisher had to admit that he hadn't thought of this. Microsoft's problem is that it lacks finesse in undertaking dirty tricks, and barges in with brute force, tripping itself in the process. Nor was it a good idea for Lacovara to address Boies as "David" for the only time in the long examination, after Boies (smelling the rat, no doubt) asked him how the Compaq data was obtained: "I went through all the Compaq products, David, and downloaded every one that says Communicator is included." It seemed to confirm that it was all a set-up. No wonder when we called Compaq in the US about this in January, they were tight-lipped and couldn't say when they had started offering Netscape. You could feel the embarrassment. ® Complete Register trial coverage
Graham Lea, 08 Jun 1999
The Register breaking news

AMD K7 veils lifted

Our friends over at JC's have put together a cold collation of what is known about AMD's up-and-coming K7. The information on the site includes motherboard and other information about chipsets. It's a useful aide memoire to what we're about to see and our information is that high premiums are already being paid for available boards by Taiwanese hardware manufacturers. AMD has been quite cunning about its plans, causing the maximum anticipation while giving out the meagrest of details. You can have no doubt whatever that Intel will do its level best to scupper AMD's launch plans. Coppermine only being the half of it. ®
Mike Magee, 08 Jun 1999
The Register breaking news

MS shines light on ShadowFactor

Microsoft yesterday announced its latest acquistion, Canadian software developer ShadowFactor Software. ShadowFactor makes software that allows online games players to communicate by audio rather than by typing in text -- always a tricky process when you're trying to nail some SOB with your Railgun at the same time. Gaming veterans may recall Bungie Software offered just such a feature with its first-person shooter Marathon way back in 1994, but this is the first time a company has offered it in an Internet context. And with online multiplayer games increasingly taking on more co-operative, team-based gameplay, as in Epic's Unreal Tournament, for example, better player-to-player communications are going to be a godsend. ShadowFactor's technology, known as BattleCom, will be integrated into DirectX 8.0, itself due to ship Q3 2000. In the meantime, it will presumably remain a standalone product. Terms of the acquisition were not revealed. ®
Tony Smith, 08 Jun 1999
The Register breaking news

Forget the K7, how about the K8?

Moscape is collaborating with AMD to create a circuit analysis tool for the K8, it said. According to the Moscape release, its patented "assertion based" technology will allow chip designers to increase quality and performance before a design is taped out. Bruce Gieseke, K8's circuit design manager, said that AMD has chosen Moscape because it needs to enforce a good methodology for complex chip designs. "In a large and complex design task such as a high-end microprocessor, we cannot leave anything to chance. Besides needing the best static timing analysis, logic verification, and physical implementation tools, we need to verify the electrical integrity of the entire design," said Gieseke. "Our intent is to integrate CircuitScope into our design methodology, and to continue to work with Moscape to fix bugs and continue to enhance its functionality." ®
Mike Magee, 08 Jun 1999
The Register breaking news

Race on for AMD domain names…

It seems like there's a mad rush to jump on AMD's bandwagon before the K7 cart pulls out of the station. Our friends over at WWW.AMDZONE.COM have confirmed that they have registered WWW.SLOTA.COM. And two days ago we reported that an individual has registered the domain WWW.ADVANCEDMICRODEVICES.COM. But another reader has just pointed out that WWW.ADVANCED-MICRO-DEVICES.COM is still available for registration. Wonder how long that will last... ®
Mike Magee, 08 Jun 1999
The Register breaking news

Oracle buys Thinking Machines

Oracle has bought Thinking Machines for an undisclosed sum, the company said yesterday. Thinking Machines' Darwin software uses artificial intelligence techniques to trawl customer purchase activity logs and use that data to predict future buying trends. That clearly sits neatly alongside Oracle's strategy of using e-commerce to flog more database management software. Darwin will ultimately be integrated into Oracle's suite of software, but for now it will continue to be sold as a standalone add-on. ®
Team Register, 08 Jun 1999
The Register breaking news

BT blocks toll-free access to LineOne

The technical loopholes that allowed some LineOne subscribers to log onto BTInternet's toll-free access service without paying the usual monthly subscription has been closed. A spokesman for BTInternet said technicians resolved the problem on Saturday. He also denied allegations circulated by some Net users that BTInternet is unable to prevent LineOne customers from doing this again in the future. "This simply is not true," said a spokesman for BTInternet. "Because of technical problems some LineOne customers could log onto BTInternet but that was only for a matter of hours," he confirmed. "But this was fixed on Saturday." LineOne is a joint venture between BT and United News & Media. Other than a few "teething troubles" BTInternet said it was pleased with the take-up of the offer, although it would not confirm how many people used the service. "The whole thing worked -- we're very pleased with it," said the spokesman. BTInternet's decision to offer toll-free access at weekends for subscribers of its £11.75 a month unlimited price plan was confirmed last week. ®
Tim Richardson, 08 Jun 1999
The Register breaking news

Quantum intros Fireball Plus KX

Storage company Quantum said it is shipping its Fireball Plus KX, with volume production slated for the end of this month. The Fireball Plus KX comes at 6.8Gb per disk, running at 7,200 revs per minute. The product has been qualified by six of the top 10 OEMs, claimed Quantum. HP will use the drives. The hard drives use giant magneto resistive (GMR) heads, developed by partner Matsushita-Kotobuki. According to Quantum, the product uses its high speed Ultra ATA/66 interface, which has Intel support. Average seek time is 8.5 milliseconds, and there will be drives offered at 6.8Gb, 10.2Gb, 13.6Gb, 20.5Gb and 27.3Gb. ®
Mike Magee, 08 Jun 1999
The Register breaking news

MS ‘Office Servers’ to rival Web servers – Ballmer

With the official launch of Office 2000 yesterday Microsoft put some more flesh on the "knowledge worker" strategy it announced last month. Office 2k is heavily Web-enabled and tellingly, Microsoft president Steve Ballmer said he hoped that "Office Servers" acting as Internet collaboration servers would become almost as common as Web servers. Well might you think: 'Aha!' Microsoft has officially embraced Internet standards and file formats, and therefore ought to be expected to finally abandon its attempts to insinuate Microsoft standards and formats onto the Internet. But we didn't believe that anyway, did we? The company's "Web Store" plan (See story) is intended to allow a broad range of unstructured data to be hosted and accessed via the Internet by our knowledge workers, and Ballmer's use of the term Office Servers makes it abundantly clear that Microsoft intends to use its massively dominant share of the productivity applications market to gain a lock on Web-based collaboration. Ballmer's dream that Office Servers could come close to Web servers in penetration meanwhile suggests that the old plan to get the Microsoft formats out there (as expressed by Bill himself in subpoenaed documents circa 1994-95) is by no means dead. Another component of the plan that emerged yesterday seems to have covered application hosting; Microsoft signed deals with Verio and Concentric Network to provide hosting services. According to Concentric, which offers Internet services to small and medium businesses, the deal involves it offering "comprehensive support for Microsoft Office Server Extensions (OSE)." This will essentially allow Concentric to extend its "ConcentricHost" NT-based applications hosting service. So effectively Microsoft is pushing into the Application Service Provision arena with Office 2k. Microsoft currently has over 90 per cent of the productivity applications market, with Ballmer claiming 120 million users. Some 15 million copies of Office 2k have been sold to businesses already - now it's your turn. ®
John Lettice, 08 Jun 1999
The Register breaking news

Intel opens Unix developer's site

Our latest trawl of domain names registered by Intel has pointed us to www.udigwg.org, a site aimed at Unix developers seeking to create solutions for IA-32 and IA-64 servers. According to the blurb, "The UNIX* Developer's Interface Guide for Intel-based Servers is a specification that will focus on specific areas affecting the development of common hardware interfaces for the UNIX* operating system and peripheral devices for 32-bit and IA-64 based server systems. The content is being developed by a working group made up of leading UNIX* OSVs, BIOSs vendors and IHVs. The guide contains a description of system and component architecture from existing industry standards and future technology trends to provide a context for programmers and system engineers. By adhering to the UDIG specification, companies developing UNIX* products aimed at Intel-based server platforms will benefit from reduced porting costs and access to a much broader selection of compatible software and hardware devices." There are other companies involved in the site but they haven't got round to putting their logos up there yet. ®
Mike Magee, 08 Jun 1999
The Register breaking news

40 per cent of business software installed illegally

Nearly 40 per cent of all business software installed during 1998 was done so illegally, the Business Software Alliance (BSA) and the Software and Information Industry Association (SIIA) have reported. In a survey conducted jointly by both trade organisations, an estimated 615 million new applications were installed across the globe last year, 231 million of them unlicensed. That figure was up 2.5 million copies on 1997's figure, and represents a cost to the software industry of $11 billion. Unsurprisingly, countries will little or no intellectual property protection legislation showed some of the highest piracy rates, most notably Vietnam (97 per cent), China (95 per cent) and Indonesia (92 per cent), though in value terms the Far East caused the industry less fiscal damage than it has in the past. The region was second only to Eastern Europe in terms of the quantity of illegal copies being used. In value terms, North America, the UK, Germany and France all generated a high financial impact, even though the level of piracy was fairly low. That tells you all you need to know about just how high prices are in the West. The level of piracy in Western Europe fell by three per cent over 1997's figures, suggesting that software companies are starting to price their products at a point where fewer users feel they need to pirate software. ®
Tony Smith, 08 Jun 1999
The Register breaking news

Amazon readies digital music delivery service

Amazon.com is to move into the downloadable music market today. It dipped its toe in the water earlier this year with one-off free singles, but this is the first time the online store has offered a specific department for downloadable music. The new site offers tracks from 14 artists, many of them fairly famous, which makes a change from the usually roster of unknowns found on many online music archives. Right now, Amazon is offering free tracks as a teaser to prompt sales of albums on CD, but its move clearly paves the way for full-scale downloads later this year -- the company is predominantly using Liquid Audio's technology, which not only limits piracy but allows Amazon to start charging for songs as and when it wants to. That's not likely to happen before the major labels begin offering albums digitally since they're the ones with the most popular -- ie. the most lucrative -- artists. That, in turn, won't happen until the music industry led Secure Digital Music Initiative (SDMI) gets its act together and releases its specifications for compression and copyright protection mechanisms later this year. ®
Tony Smith, 08 Jun 1999
The Register breaking news

US hackers picket for Mitnick…

Hackers in the US took a time out from trashing government sites to voice their support for gaol-bird Kevin Mitnick. Courtrooms in the US were picketed by slogan-waving youths eager to make a point about the detention of their hacker martyr. Mitnick has spent the last four years in custody awaiting trial after being arrested in 1995 for a string of offences. Although he coughed up earlier this year he is still awaiting sentence. "We think Kevin Mitnick's suffering has gone on way too long," read a statement on a free Kevin Site. "It's clear that the federal prosecutors feel otherwise, as they want to keep him imprisoned for yet another year and then add another three years of 'supervised release' which would forbid Kevin from even talking about computers." Elsewhere, convicted German hacker Kim Schmitz is in talks with British venture capital companies to raise cash to fund his data protection company. He needs £25 million to give Dataprotect a good start in business life. Schmitz, the man who gave Chancellor Kohl a negative credit reference after breaking into a top German bank, is still serving a two year suspended sentence. ®
Tim Richardson, 08 Jun 1999
The Register breaking news

The day Bill Gates screamed IBM's house down

MS on Trial Bill Gates went so ballistic in a phone call to a fellow IBM exec that Garry Norris could hear the screaming from across the room, the DoJ witness told the court yesterday. The awesome sound of Gates screaming about SmartSuite and OS/2 is regrettably lost to posterity, but Norris, who kept a diary of IBM's negotiations with Microsoft was able to record the subject matter. "He was complaining about SmartSuite, the audit and competing with OS/2," said Norris, and the date of the call, July 24 1995, was important. IBM had bought Lotus the previous month, and a week before had said it would make SmartSuite its standard productivity software. Prior to the purchase of Lotus IBM had been shipping SmartSuite, and had been coming under severe pressure from Microsoft to desist. SmartSuite and OS/2 were regarded as competing software by Microsoft, and according to Norris IBM was being charged more for Microsoft licences because it was shipping competing software. IBM was also undergoing an audit to determine how much it owed Microsoft because of under-reporting of Windows sales, and all three of these matters were linked into the main event, the arguments over the terms of the deal IBM would get for shipping Windows 95 later that year. The timing, the volume and the subject of Gates' phone call make it clear that he was inextricably bound up in the whole matter. He'd mailed his OEM sales chief Joachim Kempin in March 95 asking if SmartSuite "should become an issue in our global relationship with IBM," and Kempin had responded: "I am willing to do whatever it takes to kick them out." This is the sort of email exchange that makes Bill so forgetful when he's on-camera. He's clearly suggesting the possibility of global linkage, screwing IBM in other areas until it dropped SmartSuite, but isn't actually saying 'go do it.' Kempin's understanding of his meaning does however seem firm, and he says he'll use the PC relationship (the part of the MS-IBM relationship he owned and owns) to apply pressure. Norris says Kempin suggested IBM drop SmartSuite for six months in exchange for settlement of the audit. An August email from Norris' boss, Tony Santelli, confirms this: "Joachim offered to accept a single payment and close all outstanding audits." He suggested IBM not bundle Lotus SmartSuite on our system for a minimum of six months to one year." That didn't happen, and Microsoft ran IBM's negotiations over a Win95 licence right up until the product launch. According to Norris, this inflicted massive damage on IBM, and the company realised it would have to deal. As a Microsoft exec told Norris, Windows was "the only game in town." ® Complete Register trial coverage
John Lettice, 08 Jun 1999
The Register breaking news

Net study is a no-show

A major socio-economic study into how the Internet affects people is to be launched today by the Center for Communication Policy at UCLA. It will find out how people react to the Internet and how it alters their behaviour. It will examine everything from the use of e-mail and shopping habits to how it affects personal behaviour and political prejudices. "On June 8 you will be able to click here to learn more about the Center's new, groundbreaking study about the impact the Internet is having on the social structure of America and around the world," reads the introductory bumph on the UCLA site. Well, it is 8 June and not a dicky-bird and the link that does exist is ruptured. So here's a free piece of socio-enlightening data courtesy of The Register to get the researchers on their way. 1. Broken links hack people off. Big time. 2. Claiming to launch something on such-and-such a day then have nothing to show for it could drive people to flip and become violent. 3. To pre-suppose that everyone is working on laid-back Californian time is an insult to the wider Net community around the rest of the world. As well as putting the habits of US Net users under the microscope, the study will also track Net usage in Italy and Singapore, with the possibility of other countries getting involved at a later date. "The Internet is transforming how we work, how we play and how we learn," Jeffrey Cole, the study's principal investigator and director of the UCLA Center for Communication Policy is quoted as saying. "There isn't anything that's not going to be affected by this technology." AOL, Microsoft, Disney and Sony are just some of the companies that have stumped up the cash for this annual benchmark of Net behaviour. ®
Tim Richardson, 08 Jun 1999
The Register breaking news

Rambus Meltdown a sorry tale of fudge, mudge and kludge

The aptly named "Rambus Meltdown" er, Plugfest, was hosted by Intel in a Silicon Valley hotel Tuesday and Wednesday of last week. Revision B0 (B-Zero) silicon has just started sampling, but it will not go to production as planned. Intel calls pre-production silicon A0, A1, A2 until it is certain that the next rev will really work - then it calls that rev B0. Well, you guessed it, Camino rev B0 didn't quite make it. Now Intel is hoping that Rev B1 (sampling in July) will be the real thing. But Intel has not yet officially changed its already thrice delayed launch date. Intel has instead enhanced the schedule to "late in September," rather than just "September". At the meltdown, Compaq and Dell each slaved over three unique board desings each with the aid of Intel and Rambus personnel, working out the details and trying to get them to run up to speed with RIMMs from several different DRAM suppliers. IBM unenthusiastically showed up with only one board design. No other OEMs were present. All of the boards are six layers - now accepted as the minimum required to get 400MHz running. The challenge now is to get Rambus running at 400MHz on the same board with a 133MHz CPU bus and a 133MHz AGP4x bus. It seems doubtful that even Intel can pull this one off with its Vancouver OEM board. If rev B1 does not fix these minor annoyances, AGP will be the first to go - dropping the platform spec back to AGP2x (66MHz). Then if it cannot be made to work at 133/800, the RDRAM speed will be dropped to 712MHz. If that doesn't work, then "late September" might as well become "Next Millennium". A couple months ago, the Intel proposed solution was to drop Rambus to 600MHz instead of 800MHz and to shrink the maximum system configuration from a rather slim 256MB to a paltry 192MB. Surprisingly, no one seemed to complain about the DRAM capacity issue, but after loads of benchmarking, the OEMs came back yawning over the performance figures. Couldn't even match the BX at 100MHz. It was then that Intel started shifting gears to define the 712MHz and 533MHz speed grades. 533MHz offers bandwidth equal to PC133, but worse latency. Who would want this? 712MHz is a rather odd speed derived as follows: 133.3*8/3*2. Synchronous? Yes. Logical? No. 666MHz (133*5) would have been a more logical choice, but a little too Satanic for Intel marketing. It is amazing what lengths Intel will go in an attempt to avoid the already obvious associations. Now, with a veritable smorgasbord of Rambus speed grades, CPU bus speeds and AGP speeds to choose from, Intel is hoping that somebody will be able to ship something this year. The question remains - Will anybody want to buy it? ® * Rambus rose on Wall Street yesterday. See also: What the hell is Rambus and Camino all about?
Mike Magee, 08 Jun 1999
The Register breaking news

Court ruling vapes Intel's Pentium patents

Friday's ruling against Intel by Alabama district judge Edwin Nelson (Earlier Story) threatens the entire Pentium family of processors, If Intergraph is to be believed. Judge Nelson ruled that Intel does not have the right to use "Clipper" technology, and according to Intergraph this means that the entire Pentium family infringes its patents. Intel is naturally appealing the ruling, which was made as part of Intergraph's antitrust suit against the chip giant. This isn't actually scheduled to go to trial until February 14 next year, but in the meanwhile Intergraph seems to be getting the best of the skirmishing. The patent issue is one of numerous areas of dispute, and is fiendishly complicated, dating back over 20 years. Intel had argued that it had the rights to use Clipper technology for cache memory via a 1976 corss-licensing agreement with NatSemi. But Clipper's original developer was Fairchild. Both Intergraph and NatSemi bought into Fairchild in 1987. Intergraph was involved in talks to buy Fairchild's Advanced Microprocessor Division in September of that year, and said it had reached an agreement to transfer the division's assets, including Clipper patents, to Intergraph. But NatSemi bought a controlling stake in Fairchild the next month. According to Intergraph, at this point the division's assets went to Intergraph. According to the judge's ruling, this is correct. "The undisputed facts establish that NSC [NatSemi] had no legal authority to grant a license, as the patents at issue belonged not to NSC but to a legally distinct corporation - Fairchild," he says. "Intel thus never received a license from any entity with the power to grant one." In reply Intel argues that the judge is setting a new precedent for cross-licensing deals by saying that companies have to get their subsidiaries' permission before implementing them. That might be what he's saying, but we're not so sure. If there was an irrevocable deal to transfer Clipper to Intergraph before NatSemi bought Fairchild, then Clipper was not an asset that could be included in the NatSemi deal, and hence not covered by the 1976 cross-licence. We await the appeal with interest. ®
John Lettice, 08 Jun 1999
The Register breaking news

InQuest to host PC platform conference

Market research firm InQuest is hosting a conference in July which promises to be interesting. Platform99, to be held in San Jose on July 21-22, will cover platform strategies, semiconductors and system integration. InQuest says that will include the current hot debate about memory technologies, bus standards, chip architectures, microprocessor futures, host processing versus hardware acceleration of 3D geometry, real time MPEG2 encode, 3D sound and comms. Bert McComas, director of the conference, said that Platform99 will add perspective to anyone who has attended either the Intel Developer Forum (IDF) and WinHEC. He said the forum will help attendees to evaluate risks, opportunities, technology trends and strategic directions. There will be dozens of briefings from the major vendors as well as material from independent analysts. Sounds interesting to us and we plan to attend. For more information, and pricing information, go to Platform99 ®
Mike Magee, 08 Jun 1999
The Register breaking news

Coppermine delay could give AMD aid

Intel is not confirming or denying that its Coppermine process will be delayed, because it won't speak about unannounced products. But the word on the street is that CuMine, is, indeed, on the back burner. According to various reports, Intel will release 550MHz and 600MHz parts at the end of the year, and a 666MHz part in the first quarter of next year. Problems, problems. Compromises, compromises. But the delay, if true, will give succour to chip rival AMD. Intel's strategy was to introduce CuMine early and queer AMD's pitch. ® See also Coppermine taped out and real copper on way New high end Intel systems coming your way Future Xeon roadmap revealed Coppermine could finish AMD off for good Katmai was resurrected because of AMD threat Secrets of Intel's IA-64 revealed
Mike Magee, 08 Jun 1999
The Register breaking news

Disney moots Infoseek buyout

Disney wants to increase its stake in Internet portal Infoseek from 43 per cent to 100 per cent, the company has said. Such a deal would require the approval of both companies' shareholders, said Disney, but it admitted it was discussing such an arrangement with the Portal to see whether it would be a viable way to proceed. What makes the discussions interesting is Disney's comment that it is also looking at issuing a new class of stock for the combination of Infoseek and its own Internet businesses. That neatly ties in with earlier speculation that Disney intends to spin off its Internet operations to cash in on the inflated prices Net shares have been sold at over the past nine months or so. Other media companies, such as Ziff-Davis, have been equally keen to spin off their online divisions, again with the prospect of profitable share trading enriching the company coffers. Here in the UK, Bath-based Future Holding's upcoming sell-off looks set to value the company at £525-600 milion, despite the fact it lost £9.5 million last year. Meanwhile, Disney already has a portal site, the Go Network, that's co-owned by Infoseek, so buying up the latter and merging it with Go would neatly tie up all the lose ends. The cost would presumably be amortised through the money Disney would make through the sell-off of the combined operation. ®
Tony Smith, 08 Jun 1999
The Register breaking news

3Com faces insider trading lawsuit

3Com has been hit by a class action which alleges the company and unnamed "officers and directors" violated the US Securities Exchanges Act of 1934 by using company money to buy shares in order to boost the stock's value and thus allow the same "insiders" to sell their own shares for an even greater return. The suit, filed in the US District Court for Northern California, claims $130.4 million was spend buying back 4.3 million 3Com shares during the company's first and second quarters of the 1999 financial year. It also claims that 3Com staff made false and misleading claims about the company's performance in order to talk up the share price. And indeed between 1 September 1998 and 32 December 1998, 3Com's share price more than doubled, rising from $23 1/8 to $51 1/8. From November 1998, says the suit, 3Com insiders began selling 4.2 million of their own shares at prices as high as $48.69 per share to net, according to the complaint, $189 million. During February 1999, 3Com's stock fell back to $30 9/16 when it emerged Intel was trimming 3Com's share of the network card market. The suit claims 3Com executives said the company's business plan was intact causing its share price to rally slightly. Unfortunately, at the end of Q2 99, 3Com admitted its sales of modems and network cards were well down and that it was unlikely to match previous forecasts by some margin. By 3 March, the company's share price hit a low of $22 3/4. The action was filed by Atlanta-based law firm Chitwood & Harley, which is now asking 3Com common-stock holders who purchased stock between 22 September 1998 and 2 March 1999 to support the suit. 3Com had yet to respond to the suit at the time of posting. ®
Tony Smith, 08 Jun 1999
The Register breaking news

Pawn star checks online pals for pulse

Chess may be a cerebral pursuit of strategy and tact, but Microsoft has just managed to make it plumb new depths of dreariness. It wants Net users to play chess champion Garry Kasparov. The catch is that this group of Net chess stallions will have 24 hours to send in their suggested moves. (d4) Kasparov with then have 24 hours to reply. (d5) Then the Net people will get another go. (c4) Then Kasparov (c6) Netties. (Nf3) Kasparov. (Nf6) Netties. (Nc3) Kasparov. (dxc4) And so it goes on and on and on and on (BTW you've just witnessed the opening salvos of the "Slav Defence") for the whole summer. Kasparov -- the man who lost to Big Blur's chess-playing computer a couple of years back and then whinged it was too competitive -- reckons this stunt will pull people to the "sport" like bees round a honey pot. Of course they will, Garry, of course they will. ®
Tim Richardson, 08 Jun 1999
The Register breaking news

ISP hopes to woo users with share option offer

A group of London financiers has developed a risk-free way of cashing-in on the success of Net shares by promising to give its users a stake in its ISP when it starts operating next month. Themutual.net is the UK's newest subscription-free ISP and plans to "seize the lion's share of the ISP market, toppling giants like Dixon's Freeserve". The ISP is offering 50 per cent of themutual.net to its customers by issuing 2.1 billion units, each convertible to a share, on a first-come-first-served basis. Cheshire-based Telinco will provide the network infrastructure for the service. Customers will be able to realise the value of their equity if the company is sold or floated in the future. With its Net users doubling up as shareholders, the brains behind this new service believe it will help create a set of loyal users that will help build the ISP into a powerful brand. It's a new spin to an old idea but it does leave themutual.net open to carpetbaggers who may just register their interest in the hope of a future pay out. Clive Poulton-Sinclair, executive chairman of the themutual.net, said: "We've borrowed the age-old concept of a co-operative and applied it to the very latest technology business. "It is a simple formula: our customers help us make money and we help them do the same." The company aims to recruit 100,000 customers by the time it launches in July and is not shy about it prospects. "We are going to turn the industry on its head," said City veteran Poulton-Sinclair. "Less than a year ago, people were paying ISPs around £15 a month just to buy a service. "We charge nothing and let our people share in our success. It is a partnership that creates value," he said. Today's announcement by themutual.net coincides with the launch of MSN's new subscription-free service. ®
Tim Richardson, 08 Jun 1999
The Register breaking news

Turing's thoughts on Merced

Updated (The IA-64 articles in the last few days have spawned some interesting correspondence. If you'd like to contribute your thoughts, why not post them in the thread we've opened on our forum? Ed.) One of our Big Blue readers has penned some thoughts on how Alan Turing might have regarded the IA-64 architecture. Turing (1912-1954), for those readers who may not have encountered him, helped found computer science and in the Second World War cracked Nazi Germany's Enigma cipher. Here's what our IBM reader has to say: "Almost nobody has related Intel's proposal with the theoretical works of Turing. "The problem is that the only way to predict what a program will do is to run it (that's a consequence of the Turing theorem). A classic processor executes the program code and knows what happens with the current data (loops, jumps, etc). "It can infer locality relationship from its instruction flow and make on-the-fly optimizations. Meanwhile, the compiler of an non-optimizing VLIW processor such as the IA-64 can only look at the code beforehand and make assumptions about what kind of data that code will process. "And the Turing theorem predicts that the IA-64 pre-compiled optimization will always be inferior to a good on-the-fly RISC processor. At best, the precompiled optimization will yield an optimal path for a given set of assumptions, which may not be true on every run, since run-time data can defeat these assumptions. "I'd really like someone to look closely at the IA-64 architecture in the light of the Turing theorem. I understand that Turing's tumultuous sexual life make him a questionable reference in some Yankee circles, but the oh-so-British Register shouldn't have the same qualms. :-)" There's an Alan Turing home page here. ®
Mike Magee, 08 Jun 1999