A year ago Posted 31st March 1998 Reliable sources within AMD US have confirmed that the company is on top of its yield problem which have dogged its K6 processor for the last six months. But the company is in its so-called "quiet period" meaning that no official will go on the record about the improvements in production. "Yields are now very good," the source said. And as the news came through, rumours again began to circulate on Wall Street that AMD was once more in talks with a possible investor. IBM is the favourite but other names pulled out of the hat include NEC and Fujitsu. These last two are unlikely choices because both are in the process of re-engineering their semiconductor business units. They might, however, see more benefit from manufacturing CPUs rather than memories. News that yields were better caused AMD’s shares to rise by nearly $2 to $26.56 on Wall Street. At the beginning of the week, Jerry Sanders, CEO of AMD, took a wage cut to show his commitment to the company. ®
Today's Korea Herald is reporting that the giant semiconductor company merger between Hyundai and LG faces anti-trust investigation by both the EU and the US. The putative merger, pushed through by the South Korean government in the teeth of relentless LG opposition, was supposed to have been concluded by the first of April this year. Squabbling over money has delayed the deal, but now the Herald is reporting that both the US and the European Union will announce their opposition to the wedding. Hyundai, according to the reports, will pre-empt any investigation by opening itself to investigation by the two large trade blocs. ® RegiStroid A The word 'bovver' is the Londoner's way of saying 'bother'.
Serial numbers may not be so good for your privacy, but they come pretty damn useful when tracking down virus authors. Using Microsoft's highly contentious Global Unique Identifier, two bug hunters have unmasked the web site where -- they claim -- the Melissa email virus originated. Bug hunter Richard Smith, president of Phar Lap Software, has passed this information onto the FBI. He says the virus bears the hallmarks of a virus unleashed on an unsuspecting planet two years ago. Let's hope he's right. The FBI will be trampling all over this unnamed web site with a less than fine tooth comb. The Melissa author could always have framed the web site. We wonder how Microsoft will deal with this unexpected plug for serial numbers. Smith and fellow virus chaser, Swedish PhD student Fredrik Bjorck, used the electronic fingerprint contained in the Melissa-infected Word Macro. This fingerprint is identical to documents posted on the "guilty" web site. ® See also: Melissa virus threatens to bring email to a halt How to say goodbye to those Win98 ID number blues Microsoft caught with pants down over hardware IDs Unique serial number exists in all 25 micron chips
In another blow to Intel's battered pride, AMD announced late yesterday that Compaq will use its mobile K6-2s in Prosignia notebooks. According to AMD, Compaq will use its 380MHz and 350MHz mobile K6-2Ps in a range of its machines. Intel has recently vowed to mop up the low end market for notebooks, with senior executives saying at its Developer Forum last February that it was determined to grab that market space. It is the first time that Compaq has offered AMD mobile processors in the Prosignia line, said our old friend and sparring partner Dana Krelle. AMD claims that in February it had captured 33 per cent of the low-end notebook market with its MPUs. ®
The Register went down this morning for more than four hours. Our ISP says a G703 converter failure is to blame. The Register went down for 15 minutes yesterday afternoon. Our ISP does not know who to blame. The Register went down for 20 hours on Friday night through Saturday afternoon. Our ISP says a sub-contractor is to blame. Our ISP says we should not run our server using Apache and Linux -- "an unsupported operating system" -- so we are to blame. Our ISP is called Wisper. It is a subsidiary of INS, one of Europe's biggest ISPs. Level 3 of America is interested in buying INS, according to a recent report in the Daily Mail. INS says this article is completely "erroneous". We shall have a mirror server up and running Stateside in the next few weeks. In the meantime, The Register would like to apologise to any reader who tried and failed to get onto our site in recent days. INS has asked us to clarify that we are not a customer of Wisper. Rather we are a customer of a customer of Wisper. Friday night and Saturday morning, Wisper's network connection worked fine -- so the down time must be the fault of our hardware -- or software, INS says. We don't agree. INS says we should not depend on a single circuit, which is true enough. That is why we are planning to move to a dual circuit operation. ®
The transcripts of Gates' interview with Sir David Frost, and several speeches that Gates gave in his east coast tour to promote his book, have become available. They provide some fascinating details of the dear leader's current thinking. Frostie, the one-man conglomerate and impresario, is himself the author of 17 books, but none come to mind. It was clear that Frost was spoon-fed the questions by Microsoft PR, including introductory remarks attributed to "friends" which were designed to suggest that Frost had some independence. His introductory remarks included a rerun of Gates not wanting to be US president because he couldn't stand the loss of power, and Gates being part Albert Einstein, part John McEnroe and part General Patton. Gates went to Washington after the interview, while Redmond was burning from the heat of the DoJ case so-to-speak, and gave a number of promotional speeches for his new book for the business schools at Georgetown University and Columbia/NYU. Gates revealed again what an intellectual pseud he is - his favourite reading seems to be lives of the great scientists, and PhD theses, most of which he could not possibly understand as he lacks sufficient basic knowledge of computer science and mathematics. His theme of a digital nervous system, which he said is "about empowerment", should not be parsed since his use of it is as a marketing slogan. Gates' claims, such as IBM, Wang and DEC fell because they didn't see new trends, is a convenient story for a company that wishes to claim it has to innovate to stay in business. Of course reality is different, since Microsoft has succeeded because it has not innovated nor exploited its internal information (we know this from the trial evidence). There wasn't much about the current trial of course, but it was evidently thought necessary for Gates to claim he spent about ten days straight suggesting approaches and being in dialogue, because, you know, we'd like nothing better than to have this behind us." A glance at Gates' diary shows that there was no straight ten-day period when this could have happened between his sales calls on heads of governments or the like. After all, Microsoft's revenue now exceeds the GNP of more than 100 countries. Gates maintained the story that a principle at stake was that Microsoft "put Internet support into Windows" and "there were immense benefits to this" - to Microsoft of course, he forgot to add. He whinged on about "the right to innovate", "the American system", and "how do we make Windows [and Office] better". Of course the unintegrated IE5 was acclaimed "in virtually every one of the reviews", Gates said, which does tend to negate his "Internet support in Windows" argument. A strategic direction that Gates revealed is that Microsoft has come down on the side of handwriting recognition rather than speech recognition - and that this will be built into Windows "in the next two years", although "some time in the next five years dictation will be something that even a typical user will be doing". The problem still remains that people who cannot speak properly, or have a whiney voice, cannot get any benefit from speech recognition. In answer to a question at Columbia, he noted that Microsoft's speech group calls itself the 'wreck a nice beach' [recognise speech] group. It was good to know that the Y2K problem is "mostly just getting the software updated". Gates' personal future plans, which he first set out in two Playboy interviews in which he suggested he would have retired by now to devote himself full-time to philanthropy, have conveniently been forgotten and replaced by a new version - that he was going to wait until he was 60 before he did this, but some opportunities could just not wait so he and his wife set up a foundation. Of course, it might just help Microsoft at this tricky time, but focussing on world health issues is a step better than handing out tax-deductable Microsoft software. Another historical retrofit he used is a variant of the story about the birth of Microsoft. Gates said he and Paul Allen thought " 'Well, let's, you know, make the computer be the kind of thing we could have fun with'. And we had a vision that maybe we could impact the entire world". A modest delusion: but untrue. Equally untrue was how Gates in his 20's and 30's was working on software. The last thing he wrote was for the Radio Shack TRS-80 (affectionately known as the "Trash 80") - hardly "great software", even in those days. Gates still has a curiously distorted view of early Windows - "the graphical interface that people ridiculed". Perhaps that was because many of them had seen what was possible on a Mac. Capitalism was "the matching of buyers and sellers", Gates said. About four or five per cent in the US were "living a web lifestyle" (he said at Georgetown, but "less than 10 per cent" at Columbia). However, in Europe, Internet use was "about a fifth or a sixth as big" as the USA, he claimed, without the benefit of any knowledge on the subject. Then there was the term Gates had created for his new book: "Web workstyle". This became confused with his digital nervous system for the simple reason that there is no systematic thinking in these sales slogans, at least in his Georgetown speech. At Columbia it seemed to be the "web lifestyle" that was "the same concept" as the digital nervous system. In reality, Gates is trying to reduce paperwork, but office automation is not exactly an area for excellence at Microsoft. Indeed, evidence from witnesses during the trial has shown that Microsoft is indeed the shoemaker's shoeless child when it comes to office efficiency, starting with the claimed absence of any accounting system to track expenditure on Internet Explorer, thought to exceed one billion dollars. The development history of IE was revised again: "we did lots of retreats in 1994 and 1995, we kept raising the priority of what had to be done ... and by late 1995 we had made it the top priority". He did add: "There's no doubt that our leadership would have been eclipsed if we had taken [another year or two to] drive Internet standards into our products." Except of course, Microsoft drove its own versions of Internet standards, and the products were not really Microsoft's, but Spyglass' Mosaic. That was the story at Georgetown. At Columbia, "The Internet came along, and it's hard to say exactly what year it is, but sometime in 1993, 1994,1995, it moved out of the university environment, out of the computer science department, and into the world at large." Gates must have been personally one of the last software industry CEOs to become familiar with the Internet. For a long time, it was only possible to access it from a few PCs on the Microsoft campus, controlled by the library - it also required approval from a Microsoft VP to search the Internet. A few questions were allowed at Columbia. In his answers, Gates said Microsoft was "in a position of responsibility". Gates did not tell the truth when he claimed Microsoft had gone from ASCII to Unicode: as is all-too-well known, Microsoft has never favoured standard character sets and uses non-standard characters in an attempt to make it difficult to convert to standard, non-Microsoft character sets. The biggest surprise was when he mentioned "business objects and being able to exchange them across businesses, so you're not just working at bits and bytes level, you're really working at the semantic level." Somebody should remind Gates that Microsoft has declared itself to be resolutely against CORBA, especially since he went on to say that "standards are more important than ever". But there again, Gates lacks the education to be able to distinguish between standards and proprietary protocols like COM (which is not to have any rudimentary cross-platform capability in the future, Microsoft has been suggesting recently). It was also naughty of Gates to claim that "Windows supported all the standards", but he probably hasn't heard of POSIX. He also referred to "our breakthroughs like Cleartype". Still, it was comforting to hear him describe DirectX as the "game API". Gates seems remarkably out-of-touch with the price of PCs, where he saw the market as being "100 million units at still [sic] an average of about $2,500 per unit." He should visit a few stores. In his ECU speech, Gates spoke about scalability, which "is a very key initiative for us ... This year there's going to be a year of major, major advances in PC server scalability." If so, who is doing the work for Microsoft? Could it be HP? Gates also noted the importance of clustering to Microsoft, and boasted about Windows load balancing, but omitted to explain how this was a rather old idea and that Microsoft was late to the party. Transaction processing was going to be added to the operating system. It's beginning to look as if BackOffice will be an obligatory add-on for NT, which would herald a very significant price increase. Gates mentioned how Microsoft was working so that systems could be managed "without going out and visiting each of those individual systems". He must have been reading up about Novell's NDS. Gates' claim that Microsoft had partners that could offer 99.9 per cent guaranteed up time is not going to impress IT directors who want no downtime at all. Microsoft has failed to deliver what Gates has called a digital wallet (the contenders being Microsoft-free Psion and Palm, of course), so he is now advocating what he calls a tablet PC, with a high resolution screen for reading long documents. We doubt that too many people will curl up in bed with even a good tablet. Gates is out of touch with technology and the true desires of the customers to whom Microsoft claims it talks so much. All in all, this may not yet add up to the end of Microsoft as we know it, but there are certainly signs that it is the beginning. ®
Corel is set to announce today what it calls a major strategic alliance with a leading PC vendor -- in other words, just another bundling deal. However, if speculation as to the identity of the PC manufacturer in question is correct, the deal could signify something rather more important for Corel. According a report from the Reuters news agency, Corel recently posted a press conference notice warning that the company will be announcing today that it has signed the vendor to bundle WordPerfect Office. The event notice claims the deal will ensure Corel's personal productivity suite ships with "almost one fifth of all computers shipped annually". That suggest a big name, almost certainly Compaq, the only company whose global marketshare approach 20 per cent. Overall, IBM is larger, but it already owns Lotus SmartSuite, which it can ship for free. Of course, you might wonder why it matters when the agreement is simply a bundling deal. If that's all it is, that argument isn't an unreasonable one. But Corel's recent history suggests that it could be more than that. And if we speculate on a deeper deal than that, the possibilities don't seem all that unlikely. Earlier this year Corel engaged in a series of moves to protect the company from what appeared to be a hostile takeover bid. The suggested bidder was said to be Adobe, itself the subject of a takeover bid from Quark, itself rumoured at the tail end of 1998 to be talking to Corel about some kind of takeover or merger. At the time, Corel's bosses said the moves it had made in the light of the Adobe bid -- they refused to confirm or deny that it was Adobe -- were not done to protect the company from a takeover but to allow the board time to evaluate any such offers. That suggests the company, which not long after announced a rather wider than anticipated loss, might well be receptive to a buyer coming in to rescue it from further trouble. The company's problems, of course, emerged from its mis-guided (with hindsight) plan to adopt Java as the core platform for its applications. That strategy ultimately failed, but the company neatly sidestepped onto the Linux bandwagon, though a Linux hardware development programme, which created the NetWinder Internet access appliance, had to be sold off. More recently, Corel announced it was working on its own flavour of Linux, and with a Linux version of WordPerfect Office in the works, this makes the company an attractive partner for the growing band of hardware vendors keen to target the Linux market. Interestingly, the press notice doesn't mention which version of WordPerfect Office will be bundled, but it's hard to imagine the Linux release being shipped on 20 per cent of the computers to be sold this year -- unless the company's partner has an interesting new OS strategy to announce too... Of course, none of this is proof that Corel's about to be snapped up by, say, Compaq, but having its own application supplier under its belt would be a useful fillip for Compaq, doubly so since it brings in Intel-based OS development, thanks to Linux, too. And what did Compaq tell The Register yesterday? It's pursuing a low-end Linux strategy. Curiouser and curiouser, said Alice... ®
More and more it seems that Gates does not want to be remembered for what he is: rich on monopoly profits (it's around $100 billion now, and he is on track for being the world's first trillionaire). Gates seems to be trying to make his mark as some kind of business strategist, but his new book is both boring and lacks any real insight into business problems. His co-author, Collins Hemingway, was formerly employed by Microsoft, and his narrative ability barely passes that of his former leader. It is dull, and unreadable. Gates' theme that business will change more in the next ten years than it has in the last 50 is the kind of truism that has applied for more than a century. Refrigeration and air transport have been more fundamental to life than computers. Gates' hidden message is always that business needs more computers, with Microsoft software of course. It's as though he's hoping that every speech he makes will save another hundred souls -- except the reverse may be true. How many businesses have been ruined by computerisation? In how many cases have productivity levels dropped as a result of inadequate training and temporary staff being unable to use computers effectively? The FT serialised the book as part of a package deal, because Penguin, the UK publisher, is owned by the FT. It would be surprising if the decision to publish had not been taken by the FT's American-born CEO, since common sense should have warned even the FT editorial staff that this was a book that was more likely to create laughter than revenue. It is quite extraordinary that Gates is so vain that the marketing cost of "his" book will far exceed the income. Although it is said that he will be donating the profits to charity, the cost of his promotional tour is enormous. Any right-thinking shareholders should be clamouring for him to pay attention to the trial, and not indulging himself. Of course, Microsoft PR is trying to use the book tour to distract from the trial and to give Gates a savvy businessman image, rather than the petulance seen in his appalling video depositions. If this dreary book has any message, it is that Microsoft got where it is by not following what Gates suggests in it. It is being said that bookshops are confused as to whether the book should be shelved under psychology (cognitive dissonance) or humour. ®
Big Blue has found it necessary to post a message from the head of its PC division after the world's press reacted badly to CEO Loot Gerstner's statement about losses last week. David M. Thomas, group executive of IBM's Personal System Group (PSG), said: "I do not condone or excuse our business shortfalls, nor should any business owner or executive do so. The PC industry is strong, and we're in it to win." He said there was no question that Big Blue will exit the PC industry. "IBM is committed to not only staying in this business but to leading the industry." He said the PC division of IBM was affected by competitive price wars, the economic downturn in Asia, and investments in more efficient means of distribution, such as Internet and telesales venues." Unlike his boss, Thomas did not blame www.something.com. ®
Sighs of relief all around. The biggest winner of the Ericsson-Qualcomm dispute is probably Qualcomm - it could have ill afforded a protracted legal fight and the deal finally rids the company of its disastrous infrastructure division. But losing the network business entirely weakens Qualcomm's ability to influence third generation digital standard development and probably affects the long term success of the handset division. The trouble is that network technology evolves so rapidly that handset development has to take place concurrently with infrastructure development. But Qualcomm's network business wasn't robust to begin with - it was based mostly on selling base stations at a loss. The digital switch expertise that turned Lucent, Ericsson, Nortel and Nokia into network powerhouses didn't exist at Qualcomm. After the deal was disclosed Wall Street backed the "good riddance" view of the Qualcomm network business enthusiastically, and Qualcomm finally looks like a focused company; licensing income, chipsets and handsets. The benefits Ericsson derives from the deal are a great deal fuzzier and depend largely on how much leverage the company gained in the 3G development. The immediate prospect of folding a loss-making CDMA network division into Ericsson and launching an expensive CDMA phone project should be sounding alarm bells. Let's review some of Ericsson's PR coups from recent months. After announcing the weak fourth quarter results Ericsson deftly moved the spotlight from shrinking handset sales into the future by unveiling the new T28 handset. Then Ericsson used a huge Turkish GSM network deal to mask a simultaneous announcement that its vaunted GSM/TDMA/AMPS phone will be pushed into next year. Most newspapers fell for it; the network deal buried the product delay. More recently, the persistent worries about the delay of T28 and continued market share losses have eroded Ericsson's stock price. Then came the master stroke: Ericsson's first quarter profit warning was timed to be released just days before the Qualcomm deal was announced. Once again, the bad news was lost under upbeat press releases. Pure chance? Or skillful manipulation of Wall Street? Take your pick Ericsson is now milking the CDMA manufacturing angle for all it's worth - but like in some tacky horror movie, the mouldering corpse of Ericsson's 1999 profit picture is going to claw its way to daylight, no matter how deeply the company is trying to bury it. It's not that Ericsson's long term prospects are bad - the year 2000 handset line-up looks bright and a power position in W-CDMA development can be a genuine boost for profits in 2002-2005. But it seems unlikely that Wall Street will swallow the string of bad 1999 quarters Ericsson is poised to feed it. In contrast, Qualcomm's profitability is instantly improved by the deal. The CDMA division Ericsson bought is notoriously weak. We caught a glimpse of just how much synergy exists between GSM and CDMA network technologies after Motorola's attempt to leverage their GSM know-how into launching a CDMA network business - the CDMA specialists Lucent and Nortel pretty much knocked Motorola out. The CDMA handset situation is even more desperate. Nokia and Motorola are now invading the high end CDMA handset market after years of preparation. I doubt that Ericsson's belated CDMA epiphany will result in much headway in either infrastructure or handset markets. Welcome to Spin City The deal has plenty of real impact on the telecoms industry - but as usual, the hype surrounding the agreement spun rapidly out of control. The premise that Qualcomm and Ericsson will now have a genuine interest in collaboration and smooth technology transfer is absurd. Vicious backstabbing, byzantine intrigue and outright slander - following the past relations between these two firms has been like watching a rerun of "I, Claudius" without the orgy scenes. The ink on the agreement wasn't dry when the companies already started to leak conflicting information on the specifics - Qualcomm hinting at high licensing fees and Ericsson insisting that they were able to push the fees down considerably. I wouldn't take the word of either party at face value until all the fine print has been deciphered. The big issue concerning 3G is the one that few are willing to discuss. Is there actually a market? Will the consumers go for handsets that cost around $1,000 and weigh a lot? The Iridium project was based on the premise that people will pay practically anything for a satellite phone, even if you can brain a bull moose with that lump of metal Motorola is currently pushing. Consumers disagreed. The current "creeping" improvement of existing, second generation digital networks is bringing the data transfer rate above 100 kbps with much smaller investments from operators. It's a long way to 2002 when 3G solutions are expected to premier outside of Japan. Valuing a company based on 3G revenues is a risk investors are assuming - not a one way bet. The giddy media reaction to the deal seems slightly unhinged. Wall Street Journal gushed over "short term benefits" like multi-mode phones - apparently ignoring the fact that a GSM/CDMA phone is highly unlikely to surface before the year 2002. GSM/TDMA phones that are much easier to produce are arriving next winter after several years of development. And these two standards are kissing cousins technologically. Combining GSM and CDMA in one unit is a monstrous challenge. The leading companies can build these kinds of models right now. But they couldn't sell them. One consumer poll after another has shown that people are not willing to accept a substantial price or weight premium in multi-mode phones. That's the reason Nokia and Ericsson are debuting GSM/TDMA phones only next winter - it's the earliest they could come up with models that weigh under 150 grams and can be sold in the US market for under $300. I'm not sure just how serious Qualcomm's hints at "expanding to Europe and China" are. In the long run, operating a successful handset business demands synergy benefits with the infrastructure side. Moreover, total lack of brand recognition and distribution networks would make Europe and China fiendishly difficult markets to enter. There are no examples of succesful handset companies that do not have a double digit global market share; Siemens, Philips, Samsung, LG, Denso, etc. are not making any profit from handset manufacturing. The only exception is Alcatel, which seems poised to enter the front rank of mobile phone companies this year, largely at Ericsson's expense. And that success is rooted in Alcatel's 30 per cent market share in the phenomenally growing Mediterranean markets and a solid 10 per cent on China. Meanwhile, Qualcomm would do well to finish 1999 with 10 per cent market share even in its US home base - not a good launch pad for global expansion. Qualcomm's share of the CDMA chipset market in USA has topped 90 per cent and has provided the company with a near-monopoly position with all the associated benefits. That share is set to collapse in 1999 - Nokia and Motorola are not using Qualcomm chipsets for their new CDMA phones. By next autumn Sony is also switching to non-Qualcomm chipsets. The end result will be a rapid market share erosion for Qualcomm. Even more worryingly, leading American CDMA operators are showing signs of embracing Motorola and Nokia as their preferred handset providers. Sprint is elbowing Qualcomm's Q-phone aside and positioning Startac as its top high-end model; Primeco is considering using Nokia 6185 as its leading spring model. And Qualcomm had several years to consolidate its position as the premium CDMA brand before real competition arrived. That wasn't enough. What Qualcomm needs to do is to consolidate its position in the American market; if it yields the $200-plus market segment to Motorola and Nokia it will end up with a very tough competitive position in the future. The product cycle of this company is excruciatingly long. Apparently the new smartphone, pdQ, will have spent three and a half years in development when it finally debuts this summer. Launching a global expansion without a solid home market showing would be the most likely risk for Qualcomm's future success. Similarly, Ericsson's CDMA expansion is a chancy move. Being years behind the competition is bad enough - but Ericsson is also saddled with interminable product cycles. It has not kept up with the GSM phone development of Motorola, Nokia or even Alcatel. So where will the extra zip to catch up in CDMA phones come from? Qualcomm is so small that it can use Ericsson licensing revenues to fund considerable R&D projects. But Ericsson's licensing revenue from GSM and CDMA patents is dwarfed by the sheer size of the company. While Qualcomm's handset profit margins have risen due to lack of stong competition in the US CDMA handset market, Ericsson's margins have been eroded by going head-to-head with the big competitors in the fiercely contested GSM market. I would think that Ericsson's near term risk of profit problems is considerably higher than Qualcomm's. Nokia now faces the prospect of Ericsson's hegemony in determining the future of a third generation standard. W-CDMA's position was strengthened by last week's deal - but it's no longer an ostensibly equal Nokia-Ericsson project. Ericsson probably will get higher licensing revenue thanks to cross-licensing deals with Qualcomm. It is also well positioned to fine-tune the technical details of the standard to suit Ericsson better than other companies. CDMA2000 looks embattled. Now the development of this standard rests mainly on Nortel and Motorola - both of which have considerable investments in W-CDMA as well. Whereas several W-CDMA handset prototypes already exist, CDMA2000 product development seems to be falling behind. For Qualcomm, this may not mean much, since it gets licensing fees from both standards. But North American companies like Motorola and Lucent may fall behind in W-CDMA development if they split their early R&D among two, separate standards, both demanding enormous cash infusions. One big question the Ericsson-Qualcomm deal poses is how it will influence the sales growth of second generation standard GSM and CDMA. That's hard to gauge - now that W-CDMA is free from litigation worries the development of this GSM upgrade standard can be accelerated. Qualcomm is stressing that all 3G standards will be "compatible" with IS-95, the second generation CDMA standard. That may be - but it appears that Ericsson now has the power to determine the actual meaning of "compatible". And these guys could teach a thing or two about hair-splitting to Bill Gates. What is certain is that Ericsson will do its damnedest to keep W-CDMA much easier to overlay on GSM than CDMA. Operators now making choices between GSM and IS-95 will have to weigh the implications. ®
US wires reported late yesterday that Bob Palmer, who ran Digital before it was bought by Compaq, may be invited to be a board member of AMD. Palmer, who had a walk-in wardrobe when he worked at Digital, has natty taste in clothes. But he is widely blamed for selling off Digital's "family silver", leading it to a point where it would inevitably be bought. If the reports are true, however, it would be further evidence of the close ties between AMD and Compaq-Digital, already widely reported here. ®
Terry Shannon, editor of newsletter Shannon knows Compaq is delivering a presentation today at a user group meeting which reveals the future of the Alpha roadmap. Shannon is an analyst and journalist who has tracked Digital and latterly Compaq for years. He has kindly sent us a copy of the presentation he will make to DECUS today, and interesting reading it makes. He says that the Alpha 21264 EV6 delivers blistering speed and the EV67 will be available shortly. He thinks that it will clock at 700MHz for TurboLaser and 767MHz ins for WildFire, attaining an estimated 36Specint95 at 667MHz. By the end of the year, the EV68 will arrive at speeds of around 833MHz, and possibly 1GHz by year end. According to Shannon, the 21364 EV7 Alpha will use the EV6 core plus on-die level two cache and support logic. It will provide what he describes as "glueless" symmetric multiprocessing (SMP), RAMBUS support, lock stepping for NSK. The EV8 21464 Alpha "Arana" technology is expected to arrive towards the end of 2001, in the same time frame as Intel's McKinley architecture is due. It will offer symmetric multithreading and 200SPECint95 at 1200MHz and even 1400MHz. The EV9 and EV10, thinks Shannon, are in advanced stages of development. Shannon says we will see one to four CPU Regatta class ES40s in Q2 of this year, along with the WebBrick DS10, which will have one to two CPUs. Again, in the same timeframe, we will see the VMSstation VS10. Compaq will release a 1-2 CPU Alpha ProLiant called the Catamaran in the middle of this year along with a one to four CPU ProLiant dubbed Schooner. We will see WildFire 16-CPU boxes by year end, as earlier revealed here. Future NT ProLiants will include the eight CPU Python, Diamondback, Cobra, and Copperhead boxes in the first half of next year. ®
Yesterday's Microsoft reorganisation announcement was strangely confused. It refers to two divisions (business and enterprise, and consumer Windows) and three groups (business productivity, developer and consumer and commerce). In addition there is a home and retail products group that is said to "operate outside of the four core business divisions" - but, er, Microsoft has only named two divisions. The only other significant changes to our earlier story (Earlier Story) is that Jim Allchin's responsibilities cover two divisions (to fit the confused Microsoft structure, this should in effect put Allchin in charge of a group). Bob Muglia gets business productivity (applications), Paul Maritz seems happy with a lesser role, looking after developers, while Brad Chase gets the consumer stuff including MSN, shared with Jon DeVaan. The home/retail appendage (games, input devices - mice, keyboards, joysticks - and reference products such as Encarta) will be headed by Robbie Bach and Rick Thompson. Another change is the scrapping of the executive committee and the formation of a business leadership team that will meet Gates and Ballmer once a month. This consists of Jim Allchin, Orlando Ayala, Brad Chase, Jon DeVaan, Bob Herbold, Joachim Kempin, Greg Maffei, Paul Maritz, Mich Matthews, Bob Muglia, Bill Neukom, and Jeff Raikes. There were several promotions, including Mich Matthews, a key player who has been long-overlooked, to be VP, public relations; Greg Maffei, CFO also becomes senior VP finance and administration, with additional responsibilities for procurement (strange that he did not earlier control this) and real estate. Seven additional VPs were created. One mystery has been the delay in the announcement, which was expected two weeks ago. It now seems likely that Gates did not want any such events to overshadow his book promotion activities. Gates took part with Ballmer in the conference call yesterday after the announcement. There was also a need to save up some news to detract from today's status hearing in the trial, when it is expected that a provisional date for resumption will be announced. Judge Jackson is unlikely to see the reorganisation as being related to the trial. It is known that Ballmer has been mulling over these changes for some time, and there is nothing that indicates that they have anything to do with shaping Microsoft for a voluntary or compulsory breakup. Indeed, as our interview with Sun CEO Scott McNealy last week suggested, there is no need to split Microsoft to ensure competition. McNealy's solution is to stop Microsoft's acquisitions, which would be far more practical and effective, since Microsoft's "innovation" results from products that have been acquired or copied from others. Ballmer did bellow away that he would not find any breakup of the company acceptable, but it's not his call. The role of Brad Silverberg, who is believed to be unhappy at Allchin having such broad responsibility in view of Silverberg's work on launching Windows 95, had been uncertain. Ballmer said that Silverberg wanted a more unstructured lifestyle, and is returning from his long leave of absence to assist Ballmer. ®
MS on Trial Bill Gates would not comment on the possibility of successful settlement talks with the DoJ yesterday: "I think that kind of speculation is not valuable," he said, sticking to the script. "We've certainly been interested even before the lawsuit was filed." He then went on again to demand that the "integrity" of Windows be preserved, and that if so, "it would be nice if a settlement could be reached." This time around the DoJ will not be allowed to accept a pathetically weak consent decree, since the 19 states + DC effectively act as a stiffener to stop any sign of weakness. The states can seek different and additional remedies. Comments from state attorney generals have all suggested that what Microsoft is offering falls far short of the minimum necessary concessions. The DoJ has not commented. The Seattle Times carried a report on Sunday that the states wanted Microsoft to be forced to auction Windows and the trademarks. Assuming the story is true, it was quite a coup for the newspaper to get a leaked copy, since Washington state is not one of the 19 states opposing Microsoft. It would be most interesting to know how and why this leak occurred. It is unlikely that Microsoft assisted in any way, since the proposal would be violently opposed at Redmond. The report says that the states' white paper, written by Kevin O'Connor, the Wisconsin attorney general, requires more than Microsoft promising to sin no more. Specific terms mentioned include the prohibition of exclusionary contracts and predatory conduct. Threats by Microsoft to withhold products would not be allowed. Acquisitions would require approval. Modifying software, such as Java, would not be allowed (and it is to be hoped that this would be broadened into preventing Microsoft from subverting standards). The states also want Microsoft to disclose prices, and to stop discriminating in its prices. Microsoft has been looking naked in Euroland (since discrepancies in Euro prices are now apparent even to the arithmetically challenged). Cheaper prices in Spain have contrasted sharply with high prices in Sweden, so Microsoft is now having to harmonise prices. Compulsory licensing is also favoured by the states, although alternatives are apparently considered, including source code disclosure and breaking Microsoft into Babysofts along product lines. ® Complete Register trial coverage
Sources close to IDT said today that it will create a version of its Winchip2 using the 370 pin socket first developed by Intel. That adds IDT to Cyrix and Rise, who have already said they will create their chips in the pin factor. The source said that Cyrix and IDT will produce a "pin for pin" version of the chip, making things simpler for motherboard manufacturers. This means the effective demise of the Super Seven socket but with one exception, AMD. AMD insists it is still committed to the Super Seven Socket and since we first broke the first of the S370s stories at the end of last year, it has insisted it is sticking with it. Except, of course for the K7 which is an Alpha Slot A.... ®
The complex pieces of the jigsaw that makes up the AMD and Compaq relationship are now beginning to slot into place. Bob Palmer joins the AMD board, even though he's still a member of the Compaq board. Jerry Sanders III, CEO of AMD, delivers the company's results in a few days. The results will be bad. Intel's reputation has sunk to the lowest ever, even amongst its loyal OEMs. Compaq's Eckhard Pfeiffer becomes the white knight riding to AMD's rescue. The K7 is an EV technology. AMD and Compaq have always been Texan friends. Eckhard Pfeiffer has had a grudge against Intel since he argued with a European senior VP. Quod erat demonstrandum, as the Cockneys say.
High street betting office giant, William Hill, is to set up a gambling Web site aimed at European and Asian people who want to bet on UK football matches. That’s soccer, to the rest of the world. UK gamblers will not be able to participate though, as the site will be based on the Isle of Man. Despite being part of the UK, the Isle of Man has its own tax laws which will prohibit UK residents from taking part. Graham Sharpe, William Hill media relations manager, said there would be a UK equivalent to the site by August. “We set this site up for the World Cup and now we’re giving it a revamp. It’s aimed at international gamblers who are eligible to settle their bets without paying UK tax. Obviously we are not able to extend that sort of service to UK residents.” The company has a large Far Eastern customer base and will be hoping the football Web site will pull in the Asian punters in droves. Sharpe said William Hill had no idea how big the online gambling market is. “This is still a fledgling market and in truth, no-one has a clue how big it could be. We’re optimistic, though.” Despite its squeaky-clean image, UK gambling has had its fair share of set-backs recently. There has been an increase in the number of allegations of match-rigging at Premiership games, from those that led to Bruce Grobbelar and Justin Fashanu being cleared of such charges. More recently, the Metropolitan Police in London made a number of arrests after the floodlights at Charlton Athletic’s ground were found to have been tampered with. A connection was made by some in the football world between the tampering and overseas betting syndicates. ®
Saucy old scoundrel Lord Bath has exposed himself by putting his amorous autobiography online. The hereditary peer, famed for a string of relationships that he refers to as his "wifelets", has launched the first of six volumes of his life story into cyberspace. From Strictly Private, To Public Exposure kicked off yesterday with the Loins of Longleat. Lord Bath’s writings will be racy and candid, but the 66-year-old has promised to remove all names and material that could prove upsetting to those involved. The aristocrat has collected 40 years of journal jottings to form this modest account of his love life. The full story is expected to run to some three million words, proving that there may be more than a modicum of truth in the line attributed to Mae West – it’s not the men in my life but the life in my men that counts. "My purpose is to reveal myself here on paper, at considerable length until, to the best of my ability, something approaching the entire Me has been laid open to view," he told The Times newspaper. Lord Bath, who lives at his Elizabethan stately home, Longleat House in Wiltshire, is known for his individualism and eccentricity. ®
Tax-free ISA savings accounts will be delayed because they are too complex for some investment houses’ computer systems. Three of the UK’s biggest fund managers, Invesco, Mercury and Save & Prosper, will not launch the replacement to Tessas and Peps on time next month. The individual savings accounts will be between a week and a month late, according to today’s Financial Times. This follows warnings that there can be no repeat of the administrative mess that followed the launch of Peps. Despite potentially missing out on millions of pounds of investment, companies have been told by the Financial Services Authority they must get their administrative systems up to scratch. The three companies all use Rufus system software, supplied by RBAS Trust Bank. Other users of the system plan to begin offering Isas on time. Mercury said the delay was in best interests of investors and that it gave the company the best chance of getting things right first time round. ®
Distributors and customers of AMD have received a bulletin from the company pointing out that Ziff-Davis' 3DWinmark benchmarks using the null driver are inaccurate and misleading. AMD's bone of contention is that processors, such as AMD's K6-III and Pentium III are not compared fairly. Ziff-Davis has now posted a warning on its Winbench Web site. It says: "Don't compare processors using Null Driver Tests" and admits to problems in the benchmark, complicated by "certain attributes" of the Null driver. It continues: "The results from Null driver runs do not accurately reflect the performance differences between different CPU architectures". An Intel representative said: "This is unlikely to be an architectural problem. The cores of both the PII and the Celeron are identical." Ahem... ®
Citrix Systems is targeting more than 1,000 companies in Europe with the expansion of its reseller accreditation programme. The server software vendor has added two levels – Silver and Gold Solutions Provider – to its existing Authorised level in its reseller programme for Northern Europe. It aims to reach over 1,100 Citrix Solutions Network members as well as resellers selling Citrix MetaFrame and WinFrame server-based packages. Florida-based Citrix charges $995 to resellers at the Authorised level. Silver Solution Providers get extra training, technical support and marketing material for $1,495. Gold programme membership costs $4,995, which buys you sales leads, top-level account management and customised marketing programmes. Jane Rimmer, Citrix regional marketing manager Northern Europe, said: "We pledge extra support for our valued channel partners who ultimately support our customers." ®
Two failed UK channel companies will face angry creditors tomorrow morning. Memsolve’s winding up hearing takes place at The Royal Courts of Justice, London, at 10.30. The petition for winding up was presented by the distributor Northamber on 10 February. Administrative receivers from BDO Stoy Hayward were appointed when the Lancashire-based components distributor admitted owing around £4 million. Earlier this month Roldec ceased trading . There will be a creditors’ meeting for the Wolverhampton-based mail order reseller tomorrow in Birmingham at 11am. A liquidator is expected to be appointed. ®
Updated Three sites (now four sites) which all specialise in parody and satire have received writs from a large corporation which lacks a sense of fun. The sites concerned were Segfault, Userfriendly and Bedope. But we have just received an email from a reader that shows another "fun site" has received a writ from m'learned friends. If you point your browser to OTB Online, you will learn that the site has just received a letter from a grand sounding set of m'learned friends called Fross, Zelnick, Lehrman and Zissu, which represents the Jim Henson Company. The owner of OTB Online points out that a simple email would have been enough for it to have ceased and desisted from its no doubt extremely interesting and probably funny parodies and satires. Two of the (original) three sites carry notices saying they have been advised by their lawyers not to disclose which corporation is involved. But speculation is rampant. It's a very worrying development. For example, will Dell bash us because we call it the Great Satan of Hardware, Intel for calling it Chipzilla and Great Stan, Compaq's Pfeiffer because we call him Lord of Haircuts, and AMD because we dub them The Great Satan of Tape Recorders? Or even His Billness, His Dweebness or Bilge Ates? When we discover who this humourless corporation is, you can rest assured it will get ribbed by us… ® RheoStatFact On the first of April in la Belle France, the festival is called Poisson d'Avril. People pin fake fish on people's clothing to alert them to the fact they've been gulled. Gulls, being gannets, don't know the difference, so rend people's clothes in search of fishfood. In the UK, April Fool's Day only lasts until midday, when the couplet kicks in...
You have got to watch Intel all the time. Over the weekend it registered ten online sites and now we discover it has a new one called Art Museum It is a joint venture, says Intel, it sells merchandise and has a Van Gogh picture on the front and the first 10,000 memberships are free. So we joined, being cultural folks. ®
Computacenter has become Microsoft’s first UK Authorised Support Centre (ASC) since the programme’s launch six years ago. IT reseller and services giant Computacenter was appointed a top ranking services partner following what Microsoft described as: "Computacenter’s continued investment and experience in delivering and supporting Microsoft based solutions to its customers." It joins the exclusive club of ASCs in the UK consisting of Compaq, Unisys, Hewlett Packard, ICL Multivendor Computing and Wang. As an ASC, Computacenter will expand its services, backed by Microsoft. "This will lead to reduced downtime due to quicker call resolution, improved trouble-shooting due to faster escalation routes into Microsoft and the highest quality support for both Microsoft and third party products," according to a statement by the Mr. Gates’ company. Analysts congratulated Computacenter on the accreditation. Clive Longbottom, analyst at CSL, said it was good news which moved Computacenter up the pile to help it compete in Europe on a higher level. Richard Holway, analyst at Richard Holway Ltd, commented: "They are the first European ASC that does not come from the ranks of the 'traditional' vendors…This 'elevation to the peerage' should help accelerate Computacenter's already rapidly growing desktop support services business." Computacenter’s shares rose 15p to 600p today.&
We have, so far, pinned our prestigious Vulture badges on only those who deserve it. So we are pleased to announce that in the last 24 hours, there have been three recipients of the pins, which don't get given to anyone. Oh no. Last night we bestowed the pin on Ed Iacobucci, chairman of Citrix, and his charming wife Nancy, in a ceremony at the Old Monk pub in Maddox Street, Mayfair. Ed, the Father of OS/2, proudly sported his pin, as did Nancy. And this morning, we bestowed the Order of the Vulture on Sukh Rayat, general manager of component distributor Flashpoint, for services to the industry. Previous winners of the award include Rana Mainee from AMD, David Lin from Rise, Robert Stead from AMD, and Steve Tobak from Cyrix. They become knights and dames of the Order of the Vulture, through proudly wearing their pins. So far, one Intel representative has received a pin.. ®
Analysis When you buy a company -- any company -- legacy comes with it. So when Compaq bought Tandem, it was obvious it was making a strategic move into the telecomms space. Tandem, after all, supplied most of the automatic till machines (ATM), with its 99.9999 per cent non stop computing Himalaya model. With Tandem, it also bought operating systems, chip legacies, and a very committed team of sales people who had the rewards to boot. But no sooner, in hindsight, had Compaq bought Tandem with its MIPS legacy than it also bought Digital, with its Alpha legacy. It quickly cut down its chip legacy from two 64-bit processors to one -- the Alpha microprocessor. It still had (or has) to deal with the future Intel IA-64 legacy but that's a different story. What's far more interesting in this mix is where that leaves the software -- in particular the NT versus Unix battle. Bill Gates, CEO of Microsoft, was adamant that Unix could not succeed in the corporate marketplace in 1987, when we first met him. At that time, he was carrying around a set of slideware (courtesy of Harvard Graphics as we remember), which showed there were no tick boxes in favour of the Unix OS. For him, it was Microsoft Presentation Manager (in conjunction, of course with IBM Presentation Manager on the OS/2 platform), all the way. This was in spite of the fact that most of Microsoft's internal systems were run on the Xenix OS. Gates, as well as Eckhard Pfeiffer, Compaq's CEO, struck a ten year agreement in 1996 to cooperate on Microsoft Windows development on the x.86 platform. How quickly things change. When we first met Eckhard Pfeiffer in 1990, he was then general manager of Compaq in Europe, but Rod Canion, the then CEO, was deposed, and replaced with Pfeiffer. At that time, Compaq talked about industry standards and that has not stopped from when the company was set up in 1984 until now. Compaq is committed to industry standards and that means, in shorthand, what the industry, at large, wants at any given time. When Compaq was just another name for Compatible Quality (hence ComPaq), and it was copying IBM PCs, the industry standard was the x.86 processor and the PC platform. That's how Compaq started to make its money. It ran DOS, it was first to introduce the 386 chip on a portable, and it would run any software that was that "industry standard". Fifteen long years on, the tide has turned against the rise and rise of the PC, and we are faced with Internet proliferation everywhere. Whether it be NT, Linux, D/UX (Tru64) or Tandem Unix (or even Novell Netware) Compaq's software needs have changed with the times too. It now has a very large corporate base, most of which are stuck with heteregeneous systems. Luckily, Compaq has been able to adapt with the times because of the very large amount of cash it has in its coffers. So the industry standard of yesteryear (x.86 and IBM Presentation Manager), has yielded to the industry standard of today. Putting three very large companies into one easily identifiable pot is not easy, but if you check out the software, Pfeiffer's strategy is not very hard to define. It has the telcos with Tandem, it has the enterprise with Digital, and it has the desktop with Compaq. Whichever OS is suitable, Compaq can supply. Unlike Big Blue, which spawned it, Compaq was (and is) a flexible company, able to change its mind about industry standards overnight. Its Achilles Heel, however, is that it has to produce a consistent message across all software platforms. And that's far from an easy task. Imagine, if you were a user of ATMs, that a sales person was trying to sell you ProSignias as well. Or if you were a Little Blue (Digital) site, that VMS is still the name of the game. Or if you were a large PC corporate reseller, that WinNT is the way to go. These are the balls that Compaq is juggling and the plates that Eckhard Pfeiffer is spinning… And they're all software related. ®
For obvious reasons, The Register is rarely invited to Microsoft gigs, whether here or anywhere else. Don't worry, Bill Gates, your UK agency Text 100 can screw things up all by itself. We were not invited to the launch of IE5 for some reason but met many Brit journalists who were and went there. We're still trying to figure out why we're not invited. It can't be Graham Lea, can it? When they got back to their various offices and tried to make IE5 work, for some reason there was an enormous problem. And Text 100's reaction? We hear the spin dottores and meisters were told to tell the hacks: "Oh, it works, but the versions for UK journalists didn't". Ahem... ®
The Melissa Virus is the sexually transmitted disease for our virtual sex times. It was a masterstroke by the author to post the virus on an alt.sex newsgroup.