Forgot your password?
typodupeerror
Transmeta Businesses Hardware

Transmeta Mulls Exit From Processor Market 202

Posted by timothy
from the they've-already-made-it-up-in-volume dept.
chill writes "C-Net is reporting that CPU upstart Transmeta, once the employer of Linus Torvalds and maker of 'Code Morphing' processors, is contemplating leaving the chip manufacturing business. Already their IP licensing revenue exceeds that of their microprocessor sales, though both are dwarfed by their recurring quarterly losses."
This discussion has been archived. No new comments can be posted.

Transmeta Mulls Exit From Processor Market

Comments Filter:
  • by ackthpt (218170) * on Wednesday January 05, 2005 @01:09PM (#11264613) Homepage Journal
    Already their IP licensing revenue exceeds that of their microprocessor sales, though both are dwarfed by their recurring quarterly losses."

    And yet they're going to the CES in Lost Wages. (Booth 36235, LVCC)

    [Hello! My name is ARTHUR SWIFT] "Hi, these are our microproceesor products, which cost more to make than we sell them for. We're thinking about breaking into the game console market next. Losing money seems to be working for the X Box!"

    • by Anonymous Coward
      They never came out with a consumer motherboard.
      Geeks can't easily test products they can't easily buy.
      Note (shriek, actually) to proc makers:
      If you invested enough to make the cpu, offer a damned mobo at an accessable price!
  • by wiredog (43288) on Wednesday January 05, 2005 @01:12PM (#11264641) Journal
    they'll become a pure dot-com in an attempt to improve matters.

    They're doomed.

    • Re:So, basically (Score:3, Insightful)

      by Qzukk (229616)
      Yeah, its kind of sad to see that these days the money really is in being one of the pure-IP companies we all hate.

      At least Transmeta is doing this all above board with actual public licensing of their technologies instead of just sinking unsuspecting companies with lawsuits fired by submarine patents years after the technology has settled into use.
    • The only other thing I see happening is that they get sucked up by either Intel, AMD and possibly VIA.

      Intel and AMD will buy it just so Transmeta can't sell their IP to their competitor. VIA might buy it to strengthen their EPIA line, which ironicially is more successful then Transmeta's offering. I also remember seeing rumors a long time ago that Nvidia was going to buy it and start making processors. Although I very much doubt that.

      Regardless, I don't see them selling processors too much longer. Crusoe
  • by OtLa (844129) on Wednesday January 05, 2005 @01:14PM (#11264656)
    "...Transmeta is reporting a further reduction in power requirements by 44% and sees the laptop and sub-laptop markert as the primary markets for their new CPU. Intel and AMD claim to be catching up with the Transmeta chips in terms of power requirements..." Yup, that worked out well.
  • by JessLeah (625838) on Wednesday January 05, 2005 @01:15PM (#11264667)
    Virtually every field nowadays seems to be ruthlessly dominated by one or two (if you're lucky, three) titanic competitors. Trying to break into an existing market is tantamount to financial suicide. Not because newcomers have bad ideas or make bad products-- but because the "mindshare" of the unwashed masses is so stuck on the existing titans..

    McDonald's and Burger King for burgers.

    Coke and Pepsi for cola.

    Nike and Reebok for sneakers.

    Microsoft and .... well, Microsoft for operating systems.

    Dell and HP/Compaq for x86 computers.

    ATI and nVidia for graphics cards.

    And... Intel and AMD for x86 CPUs...
    • by Kenja (541830) on Wednesday January 05, 2005 @01:21PM (#11264732)
      "Trying to break into an existing market is tantamount to financial suicide."

      No, trying to break into an existing market with a sub par product is financial suicide. Face it, Transmeta dosn't make anything that people want. Their much vaulted code morphing has never been used, so they have a CPU that can emulate x86 poorly. Where is the value? Why should I buy a system that uses this CPU when for the same price I can get another that works better? Via has them beat in terms of price and wattage, Intel and AMD have them beat in terms of price and performance, in the embedded market the PowerPC and ARM series are better in every way. Let me put it this way, if it was Intel who had released the Transmeta CPU would you still think it was worth while?

      • ### Face it, Transmeta dosn't make anything that people want.

        I know quite a lot of people who would have loved to have a low power CPU that is x86 compatible in their desktop computers, the throuble is that Transmeta failed completly to sell their stuff. You simply couldn't by a mainboard with a Transmeta CPU, the only stuff that got ever released were some sub-notebooks in japan.

        That transmeta CPU also fall a bit short when it comes to speed is of course another issue, but simply because you couldn't buy
        • "I know quite a lot of people who would have loved to have a low power CPU that is x86 compatible in their desktop computers"

          Then they should buy one. Get a P4-M, Athlon Mobile or Via C3 system. All of which are low power, low cost and run circles around what Transmeta has been trying to sell.

          • ARGH! It's the PENTIUM M, NOT the Pentium 4-M that's the fast low power one! Say "Centrino" if you can't figure it out...
        • by swv3752 (187722) <swv3752.hotmail@com> on Wednesday January 05, 2005 @01:51PM (#11265019) Homepage Journal
          There were a few products that had transmeta CPU's. The problem was that there was a very small window when there was a significant adavantage to choose a Crusoe. Now there is not one. On the Low end there is a the Via C3, that is about as efficient as a Crusoe. On the high end there is the Intel Pentium M. A bit more power hungary but also better performance.
      • Actually people would want their stuff as a replacment for VIA in their home theatre boxes and other stuff, basically the market Via has had dominate for years now. Transmeta simply overlooked the importance of self builders and system builders as the first step of getting into the consumer chain.
      • by hung_himself (774451) on Wednesday January 05, 2005 @02:56PM (#11266132)
        Low level specs are great but I have a real 1Ghz transmeta chip on a sub 2 pound Sharp notebook (OK 2.5 with big battery). It lasts 9 hours like it says (tested on a long flight) and runs more than fast enough with XP for Powerpoint, Word and not bad even with a pig like CorelDraw. The power cord accidentally got unplugged while it was connected to the network once and it still had half the battery left after nearly a week on intermittent sleep

        The price performance thing is pretty meaningless as long as it is fast enough to do what you need. Not everyone uses their laptop as a primary machine, or for video processing. My main need was something that didn't weigh like a lead brick and could let me do real work on a long flight or a meeting without having to plug in somewhere. The Sharp/Transmeta does that admirably.

        As for the Centrino - it may be great - I don't know but I wouldn't go by spec-sheet alone (Xeons are the fastest chip right?). I'm curious if anyone here has *real world* experience with the Centrino based Sony? My understanding is that it has about half the battery life of the Sharp from the user reviews and I certainly don't discount that this might be because of different power management schemes that don't relate to the chip. But as a end-user consumer the Sharp notebook was a lot cheaper than the Sony last I checked and is far from being a sub-par product.
    • I almost made it. (Score:5, Insightful)

      by jellomizer (103300) * on Wednesday January 05, 2005 @01:25PM (#11264769)
      Eating a Wendys burger with a can of RC Cola, Wearing British Knighs Snearkers, Running OS X, on a Power PC Processor, but I have a nVidia graphics card... Damn! I guess I am just a Puppet of the Man!

      Actually they are 3 Mega Corps but the #3 is usually far behind, but still close enough to get good Profit.
      • Eating a Wendys burger with a can of RC Cola, Wearing British Knighs Snearkers, Running OS X, on a Power PC Processor, but I have a nVidia graphics card... Damn! I guess I am just a Puppet of the Man!

        Yeah, grow your own damn food. :)

      • Ijit. Use an XGI graphics card. ;-)

        There's actually a BUNCH of big players in graphics. Intel (the 800lb gorilla of GPUs, FWIW), ATi, nV, S3, Matrox, 3dLabs. Whew. Intel is good in low-end, ATi and nV are good in mid to high-end 3D, S3 is good in... nothing, but their GPUs make Intel's look good, Matrox is good in high-end 2D, and 3dLabs is good in workstation 3D.
      • "The RC Cola brand was acquired in October 2000 by London-based Cadbury Schweppes plc. Today, RC Cola continues under the ownership of Plano, Texas-based Dr Pepper/Seven Up, Inc., the largest beverage subsidiary of Cadbury Schweppes plc."

        I haven't seen an RC Cola in stores forever.
    • by ackthpt (218170) * on Wednesday January 05, 2005 @01:28PM (#11264793) Homepage Journal
      Virtually every field nowadays seems to be ruthlessly dominated by one or two (if you're lucky, three) titanic competitors. Trying to break into an existing market is tantamount to financial suicide. Not because newcomers have bad ideas or make bad products-- but because the "mindshare" of the unwashed masses is so stuck on the existing titans..

      You overlook a once tried-and-true strategy, which doesn't seem to have happened in this case:

      Devise some clever new bit of technology

      Burn venture capital (or even your own money if you're confident) waving it under the big noses in the industry.

      Sell out

      Logically you'd expect Intel, IBM or AMD to snatch them up as some sort of IP asset or leverage against a competitor, but Intel's scrambling against AMD, which hasn't exactly had lots of money to burn on other fronts, which left IBM who probably will pick up the ashes, unless Microsoft does and uses it for their Windows Processor ...

      (Please note, I did not include

      ...
      and

      Profit!!!
      above. Thanks.)

      • Logically you'd expect Intel, IBM or AMD to snatch them up as some sort of IP asset or leverage against a competitor, but Intel's scrambling against AMD, which hasn't exactly had lots of money to burn on other fronts, which left IBM who probably will pick up the ashes, unless Microsoft does and uses it for their Windows Processor ...

        Naw, they'll just snarf up as many of the good engineers as they can, which is what they've been doing all along. Cheaper in the short run, and more valuable in the long run.
    • It's not so much that all of these dominate companies in their respective fields are the only ones making money, or even control over 95% on the market. Possibly with the exception of computer related things at least. Even still, for the small competitors, the point is not to take on the giants but to exist and make money amoung the giants. Last I heard, Wendy's wasn't doing too bad. And I still see a lot of Jolt Cola around, so some people must like it enough to buy it. Even Microsoft has to start wor
      • There is a major difference between the computer industry and the fast food.

        The computer industry heavily relies on patents for enforcement. The fast food industry has no such leverage.

        There was a competitor from China to Intel and AMD already. U5 wiped the floor in the SX 486 segment and had most of the features which provided AMD with the winning hand against Pentium more then 5 years later. First, it was a superscalar CPU. The x86 instruction set was emulated and translated into a risc-like internal

    • Makes you wonder if the two companies could be in collusion to get around monopoly laws, eh?
    • There just wasn't any demand for another slow low power x86 clone. The "code morphing" was nearly useless and failed to deliver what was originaly promised.

      More a case of too much hype too little substance.
    • McDonald's and Burger King for burgers.

      I stick with Chez Malc. It's been years since I ate at one of those places. Whilst they're sticking with their trans fatty acids, I'm using free-range beef for a fuller flavour!

      Coke and Pepsi for cola.

      Nah, I just stick with water plus washing up liquid or bleach if I want to wash my floors. I always found that brown stuff did a good job cleaning but left things sticky. Dunno why anybody would want to put it in there stomachs.

      Nike and Reebok for sneakers.

      I
    • Your markets are totally stuffed. It's not "burgers", it's "fast food", and there are zillions of viable competitors, from Taco Bell on down to the local burrito place.

      Likewise, the market isn't "cola", it's "non-alcoholic drinks", which has tons of competition.

      Sneakers might be the right market, but there seemed to be plenty of competition last time I was in a shoe store.

      For OSes you are right on the mark, of course.

      Again, "x86 computers" is the wrong market, the right market is "desktop computers". In
      • Well, ATI and nVidia have the consumer gaming market for sure, but there is more to the 'high-performance' end than those.

        Some of the old kings of consumer graphics cards are still around. Matrox for instance has some decent consumer gaming cards like the Parhelia [matrox.com]. They also have many other high end cards specifically aimed at workstations. My ATI card runs two monitors fine. What if you want to drive 3 or 4 monitors? Get yourself a Matrox [matrox.com] card.

        There are a number of other video manufacturers around

    • Ayuh, Nike and Reebok have got the market cornered. Nobody gives a damn about the small [adidas.com], unknown [converse.com], irrelevant [newbalance.com] players [puma.com] in the sneaker market.
    • Financially speaking, Wendy's does better than Burger King. In addition, Subway is the world's second largest fast food franchise, not Burger King.
    • but because the "mindshare" of the unwashed masses is so stuck on the existing titans..

      Not disagreeing with you as branding is amazingly powerful, but there is more to it than that. Those big companies also have a lot of other advantages besides brand. They have among other things:
      • Extensive distribution channels which are VERY expensive to replicate
      • Knowledge of the market and competitive environment as well as infrastructure to use this information
      • Economies of scale due to their large production vo
    • I think it's more true for the computer world than the other things you mention. Wendy's and Arby's are just as prevelant as McD's and BK (Last I heard McD's was losing cash too fast from throwing up too MANY locations). As for cola, perhaps, but I know people drinking less of that syrup water these days anyway. Nike and Reebok? Neither I nor anyone I know have bought Nike's in years. Altho I suppose I partially did with a purchase of Chuck Taylor lo's.

      There are plenty of fairly well known "off brand" pr

    • None of these companies are unassailable. Examples:

      Nintendo & Sega used to own the console market. Now it's Sony, with MS and Nintendo a distant second and third.

      Whilst ATI and nVidia currently dominate video cards, others dominated before them (e.g. Hercules, Matrox, 3dfx etc).

      Sony used to be _the_ brand for portable music. With the Walkman, then the Discman and MiniDisc, they owned the space. Along came Apple...

      You list Dell and HP/Compaq for PCs. Used to be IBM and Compaq. Dell are very new entra
      • Nintendo & Sega used to own the console market.

        Atari were the 1000lb gorilla before that, so it's happened more than once. In fact, I considered Sega as the "new Atari" in the sense that what had happened to Atari (once seemingly unbreakable company reduced to everything failing).
    • Your examples aren't really particularly valid:

      Mc'Ds and BK are titans, but there are plenty of smaller successful and profit making fast food restaurants (from the larger ones like Wendy's to small only-one-location mom 'n' pop burger stores). Hardly financial suicide to start a fast food restaurant.

      Coke and Pepsi for cola - there are plenty of other soft drink firms who profitably make own-brand colas.

      Nike and Reebok - again, several other profitable makers exist.

      Dell and HP - plenty of whitebox maker
    • McDonald's and Burger King for burgers.

      I don't know anyone who eats at one of these places more than once every 3-6 months at this point. They also share the market with dozens upon dozens of other national, regional and local chains. Here in the Boston area, I would guess that they Kelly's Roast Beef (a local fast-food roast beef and fried seafood chain) on Rt 1 does more business than any two McDonalds in the area. The place is gigantic, and always packed. Regionally, we tend to have more family-style c
    • You also have

      Wendys, Fudruckers, Sonic, Jack in the Box

      Nike/Reebok? How about Nuke and Addidas man :)

      Microsoft and IBM and HP and Sun

      Dell/HP for computers.. true.. no one else likes those margins - however there are still plenty of mom & pops doing fine.

      ATI, Nvidia - sure if your talking gaming...

      Intel and AMD for x86, but you can also choose from hundreds of other types of cpus.. Your choices don't have to be so limited when you open up your own horizons.
  • by cheezemonkhai (638797) on Wednesday January 05, 2005 @01:15PM (#11264672) Homepage
    When I was researching building myself a little mini-itx box for playing dvd's and doing PVR I considered transmeta and via CPUs and boards.

    The via sort of has reasonable support in linux, however the transmeta seems not to be very open about giving drivers etc away.

    In the end I gave up and just used a long lead from the already present old server (Was doing firewall 7 routing etc) to the TV.

    The idea of a cool & quiet little PC to do that was great, but unless you get prices less than an pc with a quieting kit and good support under linux (and windows) then it's not going to work.

    To beat the incumbant you have to out perform and ouotprice it. Transmeta's problem is that AMD was already giving this a good go and people just don't want to use the unproven.
    • The VIA EPIA boards are actually VERY good value for money, and I think you should reconsider buying one of these boards. They are perfectly fine with a small external power unit as well (I use a 53 watt power supply, which powers the 7200 RPM HDD and DVD player as well. It cost me 139 euro's back then for a 933 CPU and board. The same price will get you a 600 MHz FANLESS CPU and board including 2 x LAN, which you can just leave in open air. Try buying a PC for that kind of money :). Actually, I am going to
  • by Anonymous Coward on Wednesday January 05, 2005 @01:16PM (#11264676)
    sure they lose a little on every sale, but they ought to be able to make up for it in volume.
    • by crow (16139)
      That can actually work.

      If you are only losing money because of the development costs, then you can make it up on volume as you have more sales to amortize the development costs over. If you're still losing money even without your fixed costs, though, you're completely hosed.
  • by sphealey (2855) on Wednesday January 05, 2005 @01:16PM (#11264678)
    Taking on Intel and AMD head-on is always an unlikely path to success. Still, the next big thing in processors has to come from somewhere, and if you can get enough funding to keep it running for 5 years it would be fun to try!

    One key though: your first release would have to be tremendously successful right out of the gate, if not in sales at least in buzz. Transmeta's first releases were, well, who knows. So I guess they weren't successful.

    Next move: sell to Intel for $50 million. Sorry investors! At least you gave Linus a place to work for a few years.

    sPh
    • If you look back, what, 15 years and just say, "Taking on Intel is always an unlikely path to success" people would have believed you. AMD is proof-positive that it is possible to break into a "saturated" market with the right combination of engineering, marketing, and manufacturing.
      • I don't disagree in general. Correct me if I am wrong though but AMD was one of the companies that was licensed by Intel to manufacture 8086s in the era of "second sourcing". They added this knowledge to their existing fab business, so they didn't come to the CPU market cold.

        sPh

      • If you look back, what, 15 years and just say, "Taking on Intel is always an unlikely path to success" people would have believed you.

        AMD has been a publically-traded company since 1972.

        It was the IBM PC, several years later, that put Intel above the competition. That luck could have easily gone to Motorola, AMD, TI, Zilog, Atmel, etc. instead. Look at a few 20-year-old circuit boards. It's just as likely to use AMD chips in it somewhere as Intel chips, or both.

    • "Still, the next big thing in processors has to come from somewhere, and if you can get enough funding to keep it running for 5 years it would be fun to try!"
      My guess would be.... IBM.
      If you look at the new Power 5 and Cell chip they could be the next big thing. At least microsoft seems to think so since they are going for a PowerPC for the next X-Box.
      I often wonder how the world might have been different if IBM had decided to really go for the PC market back when it made the first IBM PC. What if instead o
  • by Anonymous Coward on Wednesday January 05, 2005 @01:18PM (#11264706)
    TM always avoided industry standard benchmarks. That is always a bad sign.

    Instead, they always had some loser from marketing spout about efficency, blah, blah, efficency. I still have those emails and they are very funny.

    While I worked at a major OEM developing blade servers, we evaluated their processors and the performance was very weak. De-clocking existing proven designs was a better alternative.

    As is often the case with weak products, non-disclosure agreements precluded benchmark publication and disclosure of evaluation results.

    RIP TM.
  • Well (Score:5, Funny)

    by cca93014 (466820) on Wednesday January 05, 2005 @01:19PM (#11264710) Homepage
    Quarterly losses? Whew! At least they arent sufferent from monthly losses. Or even weekly ones.
  • Are there not companies which specialize in manufacturing the chip designs of other client companies? Why couldn't Transmeta design their chips, license their intellectual property, and then have their chips manufactured by one of these fabrication companies. I am no expert on the microprocessor industry, but I have heard it said that a large percentage of operating costs go towards running those expensive fabrication plants.
    • I thought they did that from the start.

      The problem is that they don't have the sales volume to amortize the development costs such that they can make a profit.

      Of course, you hear about yield problems, but I'm not sure that's really anything that Transmeta has any control over.
    • by bitmason (191759) on Wednesday January 05, 2005 @01:25PM (#11264764) Homepage
      The article summary is misleading. Transmeta already outsources manufacturing. I believe they've used both IBM and TSMC in the past--not sure who they're using at the moment. The article says that Transmeta is considering getting out of the chip design business and just license their IP. This is presumably patents, etc. around code morphing and other techniques that they've developed. It would presumably also represent a significant scaling back of the company.
    • Why outsource the manufacturing, why not take the same route as ARM.

      They just design the CPU core and then licence it to anyone and everyone. Intel licenses ARM cores (XScale is 'compatible with ARM 5TE'), you've probably got an ARM based processor in your phone and Palm Pilot these days too....

      It means that ARM can concentrate on what they do best - design decent low power CPU cores, and leave it to everyone else to actually make them and put them in things.
  • So Long, Transmeta. (Score:3, Interesting)

    by Anonymous Coward on Wednesday January 05, 2005 @01:20PM (#11264721)
    Tis' a shame that Transmeta had to get out of the chip making business. However, it appears that their presence was not in vain.

    From the article: The company emerged in 2000 with a promise to bring energy-efficient processors to notebooks. The company's low-energy push spurred Intel to cut the energy consumption in its own chips.The company emerged in 2000 with a promise to bring energy-efficient processors to notebooks.

    At least they had a long-lasting impact on Chipzilla. I never had to buy any Transmeta-powered products, but I know others who did. One tongue-in-cheek reason was to "root for the underdog."

    The only hope now is that they don't get vilified for focusing more on the revenue-generating but much maligned IP territory.
  • Power Requirements (Score:3, Insightful)

    by rpozz (249652) on Wednesday January 05, 2005 @01:21PM (#11264724)
    I think one of the problems they had was that a CPU with low power requirements is only particularly useful in embedded (ie handheld) devices, and thus x86 compatibility is not that useful. The embedded market was/is already heavily controlled by ARM-based CPUs to begin with.
    • They could have made a niche for themselves in the embedded market. There's plenty of room for fast, low-power x86 despite the ARM's and Power PC's. The problem is that they didn't go after it.

      The embedded market requires processor chips that have integrated peripherals - serial ports, ethernet, digital I/O, along with glue logic so that low speed flash memory and I/O can be easily attached to it. Transmeta went for the laptop market and only paid token attention to the embedded market.

      Dispite the l
    • If you're talking consumer electronics devices that don't have to deal with any x86 software, then yeah, ARM's the one that dominated that market. The overall situation, however, is a different story. You're going to find that people that can afford to roll their own SOC based solutions will tend to pick a favorite, being something like an even mix of Dragonball, MIPS, and ARM based designs. If the players in question (which is the bulk of the embedded systems industry) are forced to use off the shelf co
  • by Will Fisher (731585) on Wednesday January 05, 2005 @01:25PM (#11264766)
    Its a shame. I (heart) my transmeta based notebook [pcplus.co.uk], (i got it for £500 3 months ago) its so tiny. I may not be able to play back video very well, but its under a kilogram and incredibly small. Perfect for webbrowsing and email on the move :(

    Who will make processors for these kind of notebooks now?
  • pathetic.. (Score:3, Interesting)

    by torpor (458) <jayv@s y n t h.net> on Wednesday January 05, 2005 @01:27PM (#11264781) Homepage Journal
    .. they haven't even tried.

    it seems to me that transmeta should get -serious- about what the real issue with sillicon-business is: getting the product *used*.

    as a more-than-casual observer of the sillicon markets, but being consequently, admittedly, ignorant of transmeta's "consumer" stance, i can't help but feel that transmeta are still in the 'precocious spoiled brat', rather than 'serious competitive contender against golliath', stage of 'tech biz' development ..

    obviously, what they needed to do was conqure small-run manufacturing, and get the 'last-gap' hardware issues solved, while fostering their development cults. they didn't do this, instead just forever 'being defeated in the Desktop war'.

    we -need- more bold new CPU and silicon vendors, people. if only a handful of people in the world can print and manufacture silicon, that's sad..

    if, after their cut-up, whatevers left of Transmetas' engineering team get enough of a reboot, maybe we'll see them focusing on chips for devices, rather than chips for general-purpose computing (in weird ways).

    as a developer, if i could have 10,000 transmeta cpu's, all in good low-power/high-performance ratio, on 10,000 motherboards, with 10,000 power-supplies and invoices for 10,000 cases/assembly, i would write some bad-ass software, which would put those 10k cpu's to *use*. (i like to think i do this for a living..)

    but i never got the sense that transmeta gave a rats about *actual* devices, preferring to over-general-purpose-ize their engineering efforts, so that everything was *expensive*.

    (10k worth of 8051's, some batteries and leds==90's::10k ARM/PPC/TM-core ass'lies, some batteries, LCD, and a radio==2000's)

    in sum: transmeta didn't think small enough.
  • by MtViewGuy (197597) on Wednesday January 05, 2005 @01:28PM (#11264797)
    I think the biggest reason why Transmeta can't compete in the x86 CPU market is the marketing success of Intel's Centrino mobile processing technology, with lower-power motherboard chipsets and the low-power Pentium-M CPU's.

    Why bother with a company with a relatively short track record compared with Intel's long track record?
  • by gelfling (6534) on Wednesday January 05, 2005 @01:31PM (#11264823) Homepage Journal
    Transmeta came to the fore with a promise that sounded almost as dramatic as the Scientific Revolution. The problem with that is, you have to be right.

    Too bad really, because it's just one more indicator that the era of significant investment in new technology is looping ever shorter. The day when a company would invest in Xerox machine development for 20 years like Halloid did is I think, gone. Now you have to show a tiny incremental improvement right away and the hell with quantum leaps.

    And large oligopolies are in the best position to do that. Show minimal improvement with maximal crash and burn to upstarts. Didn't the Transmeta guys learn anything from Bill Gates??
  • The Crusoe Chip (Score:5, Informative)

    by elecngnr (843285) on Wednesday January 05, 2005 @01:32PM (#11264830)
    I remember back a few years when their Crusoe chips were touted as the next great development in chips. IEEE Spectrum had a big article that really pumped them up. Here is the abstract from that:

    Abstract:

    It took Transmeta engineers $100 million, five years of secret toil, and a little magic to create fast low-power chips that turn into x86s in a microsecond. Transmeta Corporation's Crusoe chips look nothing like Intel's Pentium processors. They do not even have a logic gate in common. They are smaller, consume between one-third and one-thirtieth the power (depending on the application), and implement none of the same instructions in hardware. However the Crusoe microprocessors can run the same software that runs on IBM PC-compatible personal computers with Pentium chips-for instance, Microsoft Windows or versions of Unix, along with their software applications. The paper describes the development of the Crusoe chips

    All that development and hype, yet now they are getting out of the market. Seems they should have been well positioned to dominate in the handheld and portable market. Bad business practices? The EE Times also has a good article on this. [eet.com]

    • You make it sound like the Crusoe is all hype, but it isn't like that. When it was released, it did fill a niche: a market for low-power, x86 compatible CPUs. Unfortunately, Intel jumped into that market soon after with Centrino and Pentium M. These are selling like hot cakes in Notebooks; these sales could all have gone to Transmeta.

      Of course, with their opportunity snatched away in front of them, Transmeta lost both the initiative and likely some of its investors, and now they're probably falling behind
  • Suit: "Hey, why are you guys just sitting around? Don't you have processors to build?"

    Lazy yokel 1: "Nope. They told us we don't have to do that any more. They said we have IP now."

    Lazy yokel 2: "Yeah. IP."

    Suit: "We sell processors! What the hell are we going to do if we don't make processors!?!"

    (pause)

    Lazy yokel 1: "Wanna sue someone?"
    • The flaw in your little drama play is that once the suit concludes they are an IP company, all the build-it yokels will be fired. The suit will likely instead have a little drama play about yelling at the janitor for not keeping the bathroom clean ... after all, an IP company only has an office.
  • by BillsPetMonkey (654200) on Wednesday January 05, 2005 @01:44PM (#11264947)
    will be stranded ...
  • by Animats (122034) on Wednesday January 05, 2005 @01:55PM (#11265056) Homepage
    "Code morphing", which is a form of emulation, was interesting, but not all that promising as a way to emulate one well-understood CPU architecture. AMD does some code modification when instructions are loaded into the instruction cache; they expand all the instructions up to a fixed size, like a RISC machine.

    "Code morphing" would have been more useful if the instruction set to be emulated was less well matched to a hardware implementation. The VAX instruction set comes to mind. That instruction set was hard to make run fast. Individual instructions had too many sequential steps. DEC struggled with that for years. But few need a fast VAX any more.

    The only reason that Transmeta had any success at all was that they built a chip with good on-chip subsystem-level power management. That's something which Intel and AMD had previously not considered too important, having focused on desktops first and laptops second. But it's not hard to do, and Intel then started doing it.

    • it's actually not that bad of an idea. x86 has what, 16 registers, while just about all modern designs use a tomasulu's OoO machine to use more than that. VLIWish things are probably the future for a bunch of reasons and code morphing is about the only way I can think of to make it work; the only other way is the stuff that the itanium did, which will also be in future processors, but I don't think that OoO will really go away. This would let x86 have its 16 registers and the core could rewrite the code
      • Actually, AMD expands instructions at cache load time, while Intel expands them when they enter the execution pipeline.

        After the demise of the Inanium, it will be a while before anybody tries VLIW again.

        • >>AMD expands instructions at cache load time,

          kinda, it kinda depends what type of instruction they are. the athlon's integer core has 3 decoders that are in the normal order, after fetch, before the reservation stations

          >>while Intel expands them when they enter the execution pipeline.

          intel now uses a trace cache (as long as they don't kill that and go back to the M for everything) so things are decoding __in the cache__ (well, in the trace cache), but their front end is basically the same,
  • Java processor? (Score:5, Insightful)

    by Quixote (154172) on Wednesday January 05, 2005 @02:02PM (#11265144) Homepage Journal
    I always wondered: with their "code morphing" technology, why didn't they turn it into a Java processor, with the ability to execute Java code natively? Yeah, I am aware of Sun's past efforts in this direction; but imagine if you will: Java apps running natively at the same time as Linux apps. Even if the processor is 3x slower than a x86, Java running natively would be comparable to Java running under VM on an x86 (please, I don't want to start a flamewar about Java's speed here).
    • This is exactly what people already do.

      Well, maybe a bit of explaning here. Most modern java VMs use "dynamic binary translation" to translate the java to the host instruction set. They are able to do many of the same kind of optimizations a compiler does at the same time.

      This is the same thing that Transmeta does with their code morphing, but the underneath architecture is not exposed. And instead of Java bytecodes (which is good at code size, but rather poor for untranslated performance due to the

    • So, what's to say that running JAVA on the TM *wasn't* tried?

      Maybe it didn't offer enough benefit (as compared to JIT) to matter...

      Ratboy.
  • Transmeta (Score:5, Insightful)

    by MemoryDragon (544441) on Wednesday January 05, 2005 @02:07PM (#11265215)
    Sad to say that, but Transmeta marketing is to blame. Face it, Transmeta could have had the market via now has with their C3 and Intel partially holds with their Pentium-M Centrino. The main problem was, that Transmeta went from day 1 to the manufacturers only, and were leaving the early adopters hardcore system builders out. Add to that an emerging home theatre pc market which Transmeta failed to cover that way (nobody bought at the early stages a HTPC from a manufacturer and most people still dont do due to DRM and other nastyness)


    and a VIA which just said to the people, we are not fast, but they can handle the stuff you want to do with your HTPCs self made routers, firewalls, fileservers (you name it), we are cheap you can buy our stuff from the next vendor on the net and we will support you, and Transmeta was on a losing ground.


    On one hand there was ARM which only sold cores and they did need less power, on the other hand there was VIA with the mentality you can buy our stuff even as a private person, and on the Notebook computer segment there were the Heavyweights Intel and AMD crushing Transmeta left and right.


    So where did Transmeta stand there, basically nowhere because they refused people (and there were thousands who wanted to buy that stuff at an affordable price) the hardware, by selling only reference designs and not having others selling decent boards to an affordable price. Add to that that in Europa and other markets you basically could not get the stuff and that interested people were complaining in forums about that situation for years and you have a company doomed from day 1.


    Now they want to concentrate on the core selling business, I wish them good luck they will need it, between a very good ARM on one side and VIA which still also sells boards to people if they need them on the other side and an Intel with a very good low to medium power solution on the server/notebook corner of things. Also IBM is in the business or at least other companies selling cores on the based PowerPC design.


    Guess it is time to say to Transmeta, goodbye it was nice knowing you. (Hopefully not but there is a high chance)
    • Re:Transmeta (Score:5, Interesting)

      by Simulant (528590) on Wednesday January 05, 2005 @02:30PM (#11265626) Journal

      I agree. And I'm an ex-employee, assuming that counts for anything.

      They should have been GIVING away small form factor reference designs. They ones they did sell weren't all that great, geared mainly to laptop vendors... and way over priced.

      That said, people still drool over my small, Crusoe based, laptop; especially after they've watched the SECOND movie with out changing batteries. It's the only laptop I own which I never fear running out of juice on. With a couple of batteries, I can fly just about anywhere in the world with out having to recharge. If only it had a bit more cpu power....
    • Re:Transmeta (Score:3, Insightful)

      by bani (467531)
      and a VIA which just said to the people, we are not fast, but they can handle the stuff you want to do with your HTPCs self made routers, firewalls, fileservers (you name it), we are cheap you can buy our stuff from the next vendor on the net and we will support you, and Transmeta was on a losing ground.

      thats precisely the problem though. transmeta wasnt cheap. they priced themselves out of the hobbyist market and aimed squarely at laptop manufacturers. their developer support was also very poor. via's is
  • by jd (1658) <imipak AT yahoo DOT com> on Wednesday January 05, 2005 @02:09PM (#11265241) Homepage Journal
    Inmos also died a death. Cyrix was bought out. Dunno what happened to IIT. Motorola quit CPUs. It's rarely been because the ideas were bad (well, other than the MediaGX), but because the ideas exceeded either their ability to produce, or their management's willingness to take risks.


    Transmeta's code morphing was never exploited. They had several former SGI chip specialists, but made no real progress on the graphics front. They had Linus Torvalds on board, but didn't invest enough to make their initial Linux offering stable. Only a few manufacturers were allowed to sell Transmeta products - it was next to impossible to buy the CPU itself. And their QC failed badly on the initial Crusoe chip which had numerous bugs.


    These weren't the fault of the engineers, or the design. These were political errors. Personally, I think Transmeta would do better to stay in the chip market and kick out their top managers. (Better still, sell the managers to SCO. May as well make some money out of it.)


    Transmeta's main legacy, to date, has been to force Intel and AMD to cut back on their global warming efforts. Chips are much more efficient, especially on mobile products. Revolutionizing the attitudes in the top 2 manufacturers is no mean feat. I think people should damn well be impressed by that.


    After the Crusoe was announced, IBM open-sourced their own code-morphing software (DAISY) but also did nothing with it. Another opportunity wasted.


    So, yes, I'm not best-pleased with this decision, but it may have been doomed from the start by the attitudes involved. Sad, but familiar. (Also, not unusual in projects funded by Paul Allen. I only hope Rutan can shake off the curse.)

  • You know we wouldn't have gotten it without them!
  • Just ask AMD how hard it is to chip away at the processor market. It's not enough to be a bit better or a bit cheaper, you have to be head and shoulders above whatever Intel does.
  • ...Master getting out of the lock business?

Prediction is very difficult, especially of the future. - Niels Bohr

Working...