Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Transmeta

Transmeta Goes Embedded 155

quinticent writes: "An article at CNet talks about Transmeta's entrance into the embedded market. CEO Mark Allen is quoted as saying, "By this time next year, it could equal the notebook market." Wow, when can I get my hands on a cheap embedded Transmeta system to play around with?"
This discussion has been archived. No new comments can be posted.

Transmeta Goes Embedded

Comments Filter:
  • embedded is SWEET (Score:5, Informative)

    by TechnoVooDooDaddy ( 470187 ) on Wednesday October 03, 2001 @08:16AM (#2382840) Homepage
    I've been playing around creating embedded ethernet devices... temperature sensors, stuff to control my homebrew beer rig..

    check out the TINI stuff at
    Dallas Semiconductor [dallassemiconductor.com]

    $50-60 will get you a board you can play around with and put on you home ethernet lan..

  • cheap? (Score:1, Interesting)

    by Anonymous Coward
    Wow, when can I get my hands on a cheap embedded Transmeta system to play around with?

    Wow, that's a good question... considering I seem to remember hearing Transmeta say, in the very beginning, that they wanted to make inexpensive, powerful, efficient laptops, etc., available for everyone.

    Where are those computers? The Transmeta laptops certainly aren't any less expensive than your typical Intel/AMD machines. This is unfortunate.
  • Dreams coming true? (Score:4, Interesting)

    by richie2000 ( 159732 ) <rickard.olsson@gmail.com> on Wednesday October 03, 2001 @08:18AM (#2382860) Homepage Journal
    The current high-end intel-powered laptops are just silly - who needs a 1.2GHz Mobile Pentium III in a laptop anyway? Is really Powerpoint XP that bloated? I don't think so. I've just sold off my old Toshiba PII@266 laptop but when I try to find a cheap, light-weight system with a decent screen, Bluetooth, 802b.11, Ethernet and modem built in I come up empty-handed. The models with the bells and whistles are also oversized with CPU so the price goes up and battery life goes down.

    Transmeta may provide the solution to this equation.

    • I for one need my Laptop. I use it at work while docked, use it at home and use it at remote locations.

      Compiling code, monitoring systems, working on spreadsheets, logged into Oracle applications and spotlighting several huge oracle databases.

      I wouldn't want to do this on a crappy 266.

      If you bought a 1.2ghz laptop you would see most come with dual nics, modem, infrared, 802b.11 support built in or for another 50 bucks. Not to push dell but i have 2 batteries, dvdrom, 2internal 10/100 nics, built in v.90 modem, and a large lcd screen at 1600x1200 with a fast video card that works great in hi-res/high color on my 21 inch monitor. I need the screen real-estate since i'm monitoring many systems or utilizing many windows. Hell i even have dual display when i'm at home which is great. That is the beauty of a powerfull laptop.

      There are a ton of users who have to take there work home with them, and a laptop is the easiest way. Just because people are mobile doesn't make there work or productivity less computationaly stressfull and thus there is no reason for not having a comparible workstation for use.
      • I'm not saying you don't need a laptop, I'm saying maybe you could make do with an 700 MHz CPU instead of a 1200 MHZ one. If I could trade those CPU cycles into an extra hour of battery life and a cheaper box and still get the other high-end extras, I'd be sold instantly.

        However, something (insert favourite conspiracy theory here) keeps driving 'innovation' to keep those gadgets out of my price range. Sure, I could get a used high-end or new low-end laptop but then I'd have to add several PC Cards to get the functionality that I'd prefer to have builtin from the start.

        It's possible that Transmeta will change that since their CPU focuses more on conserving energy and money than generating high numbers on MIPS (Marketing Innovations Per Second) tests and MHz tables.

      • On the other hand, there's an argument to be made that many people don't need 1.2GHz on their desktop or laptop. When a 400MHz K6-II went down on me, I switched back to an old 200MHz K6 (supposed to be 233MHz, but my crappy Azza motherboard doesn't have the right multipliers).

        I was able to run all my normal stuff (I'm a Math/CS grad student) mostly painlessly, including Gnucash and Mozilla. Mozilla was a little slow, but surprisingly it wasn't that bad. I suppose someone running the latest MS Office or MS Dev might have more problems, from what I'm told.

        -Paul Komarek
    • The models with the bells and whistles are also oversized with CPU so the price goes up and battery life goes down.

      I think most modern notebooks come with CPUs that step down to save power (not as advanced as Transmeta's --- but they certainly won't use as much power as an 800mhz cpu 100% of the time).
    • Is really Powerpoint XP that bloated? I don't think so.

      I haven't seen the current version... but for every program that comes from Microsoft, I can only shake my head at the version requirements. I used to use an office suite that fit into 38K of RAM! (not all at once, mind you, but it all fit on a 180K floppy :)

      Sure, today's programs do more, but I'd be happy with a *cheap* laptop that approximated the performance of that C64, or maybe an early 286. Monochrome screen, etc. -- priced to compete with today's models, maybe < $50?

      I know, I'm dreaming...
    • I've just sold off my old Toshiba PII@266 laptop but when I try to find a cheap, light-weight system with a decent screen, Bluetooth, 802b.11, Ethernet and modem built in I come up empty-handed. The models with the bells and whistles are also oversized with CPU so the price goes up and battery life goes down
      Sony's C1 line of picturebook has a 600mhz crusoe, bluetooth, ethernet, usb, firewire and pcmcia. There are companies importing it to the US now, such as dynamism.com [dynamism.com] (Note: I have no idea who they are, they're just the only company I found in the first few pages of a google search.)

      Or, if you can forego bluetooth, you can do what I did and get an ibook. It's got a wonderful screen, builtin ethernet, modem, usb and firewire, and with the antenna apple thoughtfully built in to the case around the screen, my airport card gets wonderful range. And, with DVD, 20gb drive, and airport, it only cost me $1900 or so.

      Plus, with the ibook's keyboard, I had no problems rearranging the keys. It would have sucked to have several keys wrong on a machine because of a pointing device between the H and G keys not allowing those to be moved (short of taking a dremel tool to I and D.)

      Yes, I'm a dvorak slut. ;=)

      • Sony's C1 line of picturebook has a 600mhz crusoe, bluetooth, ethernet, usb, firewire and pcmcia.

        There you go. I'd have thought it'd take a little longer for someone to build a laptop/notebook after my specs, but I'm not complaining. :-)

        This kinda validates my point a bit: A step down in CPU power do not necessarily mean having to step down in features. But most laptop and notebook manufacturers act like this is the law or something. Are we really driving that curve or is the manufacturers doing it for us?

    • Let's see ...

      musicians - software sythesis and sequencing software

      engineers - real-time data gathering/processing in the field

      students - practice games of quake

      the list goes on ...

      Oh, wait - was your question rhetorical?
      • Oh, wait - was your question rhetorical?

        Not really. I mean, I've had this old beaten up Toshiba Sat Pro for like four years now - it was a 64MB, 4GB PII@266 when I got it and it had 160MB/12GB when I sold it last week. It ran Red Hat 7.1 with KDE no sweat and even Win 2k Pro with no major problems (apart from what you'd expect from MS). I had Office 2k and all that, no big deal. Way back when I was consulting for Sendit/MIBU, I dual-booted into an NT 4 Server Enterprise Edition with an SQL Server, MSMQ and loads of other crap for documentation and testing and it ran jeest fine, thankyouverymuch.

        So why would a musician need 4-6 times the CPU? An engineer in the field would be better off with a ruggedized sub-notebook, regardless of the CPU - he'd want battery power and durability, not 1200MHz of powersucking bestiality. OK, the Quaking student is a point, but what student can cough up 2 grand for a laptop like that anyway?

        • Practical example for musician: I'm planning on building a system to run Reason, Logic Audio Gold, SoundForge and Premiere, plus some as-yet-undetermined others. The first two packages have to compensate for a certain amount of MIDI lag that can be minimized with a faster processor. SoundForge can get pretty processor intensive working on large files, and Premiere just eats up processor time rendering previews.

          In addition, I'm adding an external multi-channel audio recording device and an AV-quality time synchronizer.

          All this so to that the machine will actually play the note I hit when I hit the keyboard or will be able to keep up with audio being digitized at 48KHz across eight channels with no skips. (If I had tons of money, I could get a ProTools system, but it is not at all portable.)

          Ideally, all of this would run from a notebook, but there just isn't one fast enough yet. This is a fairly reasonable software loadout (my opinion) for a good multimedia rig.

          For normal purposes (word processing, browsing, even some design work) I don't need an ultra-fast processor. My 700MHz machine does just fine. But it sure isn't portable.

          All that aside, speed is good. The less time I spend waiting for the machine to boot up and sort itself out, the better.
          • (If I had tons of money, I could get a ProTools system, but it is not at all portable.)

            Are you aware of programs like Samplitude? (www.sekd.com) It's in the same league as ProTools, a full DAW - multichannel mixing/processing/sampling in one app. At least, if price was your worry.

            --
    • The short battery life is probably down to the "decent screen" and other components (DVD, Harddisk) and little to do with the CPU if you compare the latest from Intel / Transmeta.

      That being said, I'm typing this on a Dell Inspiron 8100 [dell.com] which I can get 9 hours of work done between recharges... I do have 2 batteries in it, but I think that this is still quite good.

      (Running Slackware 8.0, 2.4.6 kernel with APM on, with Windows 2000 under VMWare)

  • Is this because.. (Score:2, Interesting)

    by Anonymous Coward
    They have pretty much failed to capture any significant chunk of the notebook market?

    Even in the server arena those new low power 0.13u pentium III tualatin's compete very nicely in terms of performance and power consumption.

    So has the hype finally been laid to rest?
  • by Shivetya ( 243324 ) on Wednesday October 03, 2001 @08:21AM (#2382877) Homepage Journal
    Eventually if you poke enough times Chipzilla will wake, and thats exactly what has happened. Both AMD and Transmeta have to look at other markets to sustain their profits. AMD has flash memory, and Transmeta is now realizing it needs to enter a market other than the laptop market. I would not be surprised to see Transmeta leave the notebook market completely within a few years.

    Simply put, their advantage over Intel in the notebook market is nearly gone, Intel simply did what they constantly do, get off their butts and release the products they should have already had out there, and at prices where they should have been.

    If anything, Transmeta's best contribution to the notebook market was in forcing Intel to release superior products are reasonable prices.

  • by jonr ( 1130 )
    Big deal. ARM has been doing this for years. If you want a really powerful and cheap and power efficient CPU, check out the ARM line. And I'm not only talking about the StrongARM, although it is nice. ARM's are everywhere where you need good CPU power, but want to run it off a pair of AA batteries... Would Transmeta get all this hoopla if Linus didn't work there?
    • I agree that ARMs are good... thats what runs some of those fancy cell phones. The question is, will Transmeta be better? I think I'll let other guinnie pigs find out for me, before i just go diving right in.
    • by tmark ( 230091 ) on Wednesday October 03, 2001 @08:32AM (#2382932)
      Would Transmeta get all this hoopla if Linus didn't work there?

      I am pretty certain they wouldn't get as much hoopla here if Linus didn't work there. Linus could run a company selling gas chambers to neo-Nazi groups and the crowd here would be talking up how the code determining gas mixtures was open-source.


      • How many people can claim to have taken on both Microsoft and Intel??

        What's not to like?

    • by tjb ( 226873 )
      The biggest problem with the ARM core is cost.

      Buying the chips might be cost-effective if all you need is a basic ARM, but licensing an ARM core to incorporate into a custom design requires a $0.50 royalty on every every chip sold. This is fine for low-volume devices, but when you plan on selling 10 million chips, its cheaper to fork over the million $ for a MIPS design and sell the chips royalty free. Until ARM changes their licensing agreement they will be more or less locked out of the high-volume custom chip market.

      Tim
    • Well it depends on what kind of embedded device you're designing, but in many cases the MIPS [mips.com] processor fits the bill better than ARM. MIPS processors are designed by many different vendors, and each has their own target market. The MIPS architecture has seen a massive growth in the embedded market in the last few years and there are some impressive processors out there from the low end [nec.co.jp] to the extremely high-end [pmc-sierra.com].

      StrongARM is well positioned for handheld devcies, but I'd say the MIPS architecture is the most promising in the embedded space over all.

      - j
    • yes ARM arch is nice and well designed ISA

      and yes transmeta have not got a hope in this area
      (anyone with any experance in building real embeded systems knows you cant afford 16MB for code morphing or have the fact that it might not execute in 25 cycles )

      but really Intel with StrongARM2 (aka Xscale) has the ropes nobody can licence this off intel so ARM's bussiness model goes out the window
      and TI who just use ARM to get their DSP going well
      (lets face it a phone is just a DSP and so is a MP3 player (watch out empeg))

      both of these dont licence so the only thing that ARM has left is ARM7TDMI and derivertives + makeing up new extenions to arch + macrocells

      yes ARM is nice but with Intel killing ST and the rest I dont see where any more revenue comes from for ARM the company

      + the fact that all WinCE XP will be Only ARM
      (suspect that Intel cuddled up to MS and got them to see reason much like symbian)

      just watch out for MIPS
      (I saw a dual 64bit 1GHz on one chip which was nice)
      and for SOC parts there is always the ex strongARM design team http://www.alchemysemi.com/ [alchemysemi.com]

      which seem to be doing very well with more on chip and lower power than StrongARM + can alter clock speed

      lets face it MIPS is one of the oldies but goodies

      regards

      john jones
    • Depends on the embedded device, but some of them can't use the ARM, MIPS, or PPC because of commercial offerings NOT being there for those architechtures.

      ARM's very nice, but it's not the end-all-be-all solution to embedded systems.
  • Embedded? (Score:4, Insightful)

    by zCyl ( 14362 ) on Wednesday October 03, 2001 @08:22AM (#2382885)
    I realize transmeta is a poster child company for the Slashdot community, but they really haven't made a notable contribution to the industry since their creation. I can't help but consider this another pie-in-the-sky agenda for them to attempt, because I can't come up with any reason for the embedded market to inflate that significantly that quickly. Yes, technologically we CAN embed technology in a lot of items, but what benefit do the consumers gain from this that makes it worth their money?

    When it comes down to it, especially in a recession, average consumers aren't going to buy something that doesn't make their lives better.
    • Re:Embedded? (Score:2, Insightful)

      by shibut ( 208631 )
      The issue here is at what price. If it was a 5 cent overhead I think you would agree that most people would buy LAN-ready appliances. Not to mention blue-tooth ready cell phones. The problem is that, for example, the blue tooth chip set is way over $5 even which by the time it gets to the consumer means increasing the cost of their cellphone by 50-100%. This ruins the attractiveness.

      Similarly, with a Transmeta processor, if you could for example put multiple sensors in your house that would wirelessly transmit humidity and temp readings to a central unit that would adjust the thermostat for, say $50, I think many people would do this. In the future these sensors could be embedded in the AC vents, for example, PROVIDED they are very cheap (not sure what an acceptable price is but I'm assuming $1 should do it).

      I think the issue here is that transmeta will not enable a $1-$5 solution so their "penetration" into the embedded market depends on the existence of complementary simpler chipsets (e.g., in the temp case transmeta could be embedded in the central unit).
      • Re:Embedded? (Score:4, Insightful)

        by tmark ( 230091 ) on Wednesday October 03, 2001 @08:36AM (#2382958)
        Similarly, with a Transmeta processor, if you could for example put multiple sensors in your house that would wirelessly transmit humidity and
        temp readings to a central unit that would adjust the thermostat for,


        So why would such simple-minded code need the code-morphing technology that defines Transmeta ? As far as I can tell, code-morphing only makes sense in higher-end computing devices; application in the embedded market where device requirements are far simpler seems to nullify any advantage Transmeta might have. In fact, the overhead this technology almost certainly imposes, along with the overheads associated with the bevy of patents surround Transmeta's (here superfluous) technologies would probably place such chips at a marked disadvantage.
    • Re:Embedded? (Score:3, Insightful)

      by bluebomber ( 155733 )
      ...because I can't come up with any reason for the embedded market to inflate that significantly that quickly



      Uhhh... you don't *work* in the embedded market, do you? I write embedded software. The market is already huge. Don't think about where you *might* want to put microprocessors, think about where they already are: phones (cell/cordless), tvs, set-top boxes, microwave ovens, vcrs/dvds, stereo equipment, ... the list goes on and on. Motorola, TI, and Intel are all big players in this market (among many others). As for transmeta being a player? The embeded market moves slower than that...

      • Uhhh... you don't *work* in the embedded market, do you?

        I don't work in that market either but I wonder if it looks bigger to you because you're close to it?

        • Phones = these don't need transmeta chips
        • tvs = these don't need transmeta chips
        • set top boxes = don't have one
        • microwave ovens = doesn't need good chip, small market
        • vcrs/dvds = doesn't need good chip, small market
        • stereo equipment = small market


        Some of these markets are small because the equipment lasts for 15 years. (Stereo equipment, microwave ovens, etc.) I can't help but think that as they move to embedded systems they have decided to stop trying be cutting edge in favor of moving to cheap systems in an attempt to stay alive.

        It's too bad (according to me) because the competition was moving the whole chip business forward.
        • They don't need big chips the way they work today. But tomorrow they might need more power.

          For instance, we just got a new microwave oven where I live. It has a lot of neat features like frying capabilities, steaming vegetables etc. but it's a pain to use the because you need to look it up in manual and then use the features. (For instence using the automatic thaw and fry function for french fries means you need to know to use the "Crisp 2" mode. Not hard when you know, but it's not something you could figure out without reading the manual.)

          If instead this oven had a better controlling chip it could have an menu as well. That would increase the usability of the device by a lot. (It already has a rather good LCD display.)

          The same goes for TV's, VCR etc. Just compare the menues on a typical VCR/TV with those on a TiVo. Quite apparently you can make a better product if you have more power to back it.
        • Phones: Need mucho-CPU for encryption, data-handling, games, etc.

          TVs: On-screen menus (Tivo/Replay built-in), Wake-on-Word (watching captions on multiple channels). Potentially decoding some Flash/Shockwave-type protocol for easy field-upgradability.

          Set-top boxes: Potentially computer replacements, need enough CPU to work with a few million polygons (eventually) for high-end graphics.

          Microwave ovens: How about instead of entering the desired time, if you could tell the microwave what type of food you were cooking and it could tell when it was done via surface temperature and oven-humidity? This is already possible for simple foods like popcorn and baked-potatoes.

          DVDs/CDs: Same requirements as a TV, except also the ability to decode higher MPEG levels for future expansion. Also to support decent wireless networking so you can configure play-lists on a device with an interface (handheld, or desktop computer.)

          Stereo: 5.1 Theatre sound decoding, etc. Audio filters.

          Now, a lot of that doesn't need a general-purpose CPU. You could do it with a smaller CPU and a DSP, but DSPs cost more to develop for (if you're getting their full potential from them you've got to know a lot more about the chip and do a little ASM). Transmeta also has a possible advantage here in emulating CPUs, giving you a faster version of whatever you were using before, removing any development time that would be spent porting.

      • The Z-80A? A couple of years ago I was looking at embedded systems and the Big Name Chips were the 386/486, StrongArm, and Z80A. IIRC, the Z80A was a 32 bit Z80. The Z80 was what the Sinclair computer ran on, and some calculators.
        • Re:Y'all still using (Score:5, Informative)

          by gorilla ( 36491 ) on Wednesday October 03, 2001 @09:28AM (#2383188)
          The Z80A was actually the chip that the Sinclair systems had, a higher rated version of the Z80. It was actually rated for 4Mhz, but Sinclair underclocked it to either 3.25Mhz (ZX80 and ZX81) or 3.5Mhz (Spectrum) because it saved on a seperate clock chrystal. The original Z80 was a 2.5Mhz chip. Later Z80 variants were the Z80B (6Mhz) and Z80H (8Mhz), and the Hitachi HD64180 added a MMU to support paged memory.

          The Zilog Z280 was a 16 bit CPU, also adding a MMU, but mainly increasing the address space to 16Mb, with an internal clock quad, so it would run at 16Mhz on a 4Mhz bus.

          Later still, there was the Z8000, which was also basically 16 bit, but had instructions to use its registers as 8,16,32, or 64 bit registers, and the Z80000, which was 32 bits internal version, but with the same basic instruction set.

          Zilog's current product line include a number of variations on the above chips, designed for embedded systems with things like on chip ethernet, UARTS, real time clock, etc.

          • Sort of right. The 16bit Z8000 (Olivetti M1 in 1985?) came before the Z180 and all that though - it was Zilogs failed answer to the 68000 and 8086. It was too complex. Later they tried a Z800, but it was too late for that one.

            The Z180, Z280 and Z380 are their current Z80 based chips. Also there is some kind of eZ80 which looks like a variant of the Z380. The Z380 is the top end, and runs at over 40MHz, perhaps 100's of MHz, but I don't know because the website is confusing.

            Great choice for lots of applications though, especially if you have legacy Z80 code to support. The additional hardware support, as you mention, really makes the processor.

    • Re:Embedded? (Score:3, Informative)

      by sql*kitten ( 1359 )
      I can't come up with any reason for the embedded market to inflate that significantly that quickly.

      Plus, a typical embedded CPU is a Z80, a 6502 or a 68000 or StrongARM at most. There's absolutely no need for a PC-class processor for embedded tasks.

      But cost of the CPU isn't the only reason. The Z80 (for example) is very well understood, it and the software running on it can be made very reliable. Applications for the Z80 can be written with minimal memory, because of its 8-bit simplicity, using a 32-bit processor would be using a sledgehammer to crack a nut. The embedded software community are quite conservative, and will good reason, consumer electronics (other than PCs) just can't crash without serious economic consequences. If you did need a 32-bit processor, it's likely to be a SPARC or an i386 (or the aformentioned 68000) for the reason that they're cheap, reliable and well-understood.

      The embedded market is already large, but it is dominated by a few entrenched players who compete aggressively with each other. Good luck to Transmeta, but I don't rate their chances.
      • Plus, a typical embedded CPU is a Z80, a 6502 or a 68000 or StrongARM at most. There's absolutely no need for a PC-class processor for embedded tasks.

        Really? You want aircraft, ships, spacecraft, air control towers, radars, etc. running off a 68k or SA1100 at most?

        I've built many systems where an entire PC is shoved into a 19" rack and is considered embedded because it performed a set of defined tasks and is not used for general computing. One system analyzed radar jammer signals and consisted of one 486 host with 20+ 32-bit transputers plugged into the 486's expansion slots. There is quite a market for "big" embedded systems.

        -tim
    • Re:Embedded? (Score:4, Informative)

      by iso ( 87585 ) <slashNO@SPAMwarpzero.info> on Wednesday October 03, 2001 @09:49AM (#2383304) Homepage
      I can't come up with any reason for the embedded market to inflate that significantly that quickly.

      This is so funny. I've noticed that the vast majority of people on slashdot don't even understand what constitutes the embedded market. Most people, if they think of it at all, consider "embedded" to mean handhelds and Tivos. Consider: routers, switches, DSL modems, DVD players, microwaves, robotic control, almost any modern car; basically anything electronic that does more than a simple task and isn't a desktop computer probably has an embedded processor in it.

      I've got news for you: the embedded market already outnumbers the laptop market! It is estimated that for every personal computer (not just laptops) there are eight to ten embedded computers. When Transmeta mentioned that they expect this market to be bigger than the market for latops they were referring to their chips penetrating that market (ie, sales of their embedded chips will outnumber sales of their laptop chips).

      If you're dealing with just PCs and servers, you're missinga an entire world of computers. And unlike the PC market, the embedded market is actually growing.

      - j
      • It is estimated that for every personal computer (not just laptops) there are eight to ten embedded computers.

        Also embededded systems often have multiple processors. For example, one to run the display, one to do each real time operation. That means if the display is being hammered, the real time ops aren't affected.

      • I'm glad to see that someone mentioned this.

        Items like thermostats, and even toothbrushes (see the Braun 3D electric jobby). There are lots of low power and low spunk MCU's like the Atmel's AVR's (actually a fair kick to them), and PICs, TI's MSP's, Intel 8051, Motorola 68HC16...etc all the way to highend embedded PowerPC's.

        What else? Radar, industrial/process control, X10 appliances, test equipment, audio equipment, robotics control, noise cancellation systems, ovens, humidifiers, Christmas tree lights, radios...etc.
      • that is a very good point
      • >>This is so funny. I've noticed that the vast majority of people on slashdot don't even understand what constitutes the embedded market.

        I totally agree. Having spent a fair amount of time working with coldfire, kahlua, and various mips cores my perspective grew substantially.

        I don't read slashdot too often for exactly this reason. Too many people speak authoritatively when it's clear they only have a vague idea what they are talking about.
        This site seems to be fueled by fringe issues and the people who love being an expert in them. Having worked in the semiconductor industry in a lab and FAE role, I have some knowledge of the biz and have to laugh when I read some of these posts.

        Ask any design engineer and they'll tell you that their favorite job is to design cpus but then try to hire them into a company that does only that to watch their expression slowly change into discomfort as they try to wriggle away. The simple fact is that 100 cpus are designed for ten that get to market and of those, only one will actually make any money.

        Transmeta had an interesting idea and if they could have executed in a reasonable time frame, they might have gotten some market share. At this point, unless they are very careful, they will miss the portable embedded market cycle power cost curve as well.
  • customization (Score:2, Interesting)

    by shibut ( 208631 )
    My understanding is that embedded systems need more customization than your typical laptop. That's why companies like Tensilica can sell their design cores. Having a chip that relies so much on sw can be a boon for customization, provided that it is open enough (I'm assuming Linus took care of that). However, one of the reasons that so far companies have customized their chips is that they wanted to minimize the sw part for efficiency reasons. I wonder if this doesn't contradict the vision of cheap embedded systems for all...
  • "By this time next year, it could equal the notebook market," Mark Allen, Transmeta's CEO, said of the company's prospects in the market for embedded chips.

    I wonder here whether he was talking about the notebook market for Transmeta's chips or whether about the notebook market at large. Note also here that their entrance into this market really amounts to nothing more substantial than 'marketing to' makers of equipment that uses embedded chips. When Transmeta first started trading there was so much buzz around it that the marketing seemed plausible. Now, after a precipitous drop in share price (yes, I own TMTA), and numerous fallen-through-supposed-deals with computer companies, and TMTA looking like it might be delisted in 6 months, this kind of hyperbole makes me suspicious. Why, if this market is going to be so huge, was TMTA *not* marketing to these guys in the first place ? What's wrong with their leadership that they did not initially target a market that seems like it fits TMTA so much better ?
    • "By this time next year, it could equal the notebook market," Mark Allen, Transmeta's CEO, said of the company's prospects in the market for embedded chips.

      $280mm of cash and near cash
      burn rate of $70mm per quarter
      capital markets dried up entirely
      macro environment going down the toilet

      Transmetta will not be around in its current form a "this time next year".
      • Transmeta was never a going concern. They had crap for technology and fools for marketing. Late to market and off target with poor controls. feh.

  • by wowbagger ( 69688 ) on Wednesday October 03, 2001 @08:28AM (#2382918) Homepage Journal
    I've been eagerly awaiting one of the Single Board Computer (SBC) venders to make a device with a Crusoe on it. I work for a company that makes embedded devices, and currently we are using a Pentium/MMX based system at 200 MHz. Even that CPU board sucks down about 25 to 50 watts, which is not good when you are trying to make a system that needs to run from a lighter socket (ca. 120 watts maximum, more like 80 under real conditions.) Plus, you have to get rid of that heat, which adds to the cooling burden, and the @#()$& CPU fans keep crapping out on us. This is on an industrial SBC costing about US$700.

    I'd love to go to a Crusoe based system, pick up a few MIPS, and cut a few watts out of the power budget. Add to this the fact that the Crusoe has the North Bridge built-in, which reduces the size of the board, and you have a great win all around for us embedded types.

    However, unlike the laptop market, x86 compatibility isn't as great a deal for us embedded weenies, and therefor the StrongArm XScale is an attractive option too. Transmeta had better make this a very compelling option.
    • Actually, Advantech already makes one. PCM-9370F [advantech.com]. It's a 3.5" form factor, and after talking with one of their reps, the PCM-9370 is $725 and has a 2 week lead time. That's comparable to the Intel and NS Geode boards they have. I'm not sure if it's on their online store yet (that's why I had to ask a rep). They give a typical power consumption of 10.7W @ 533MHz with 64MB of RAM.

      -RevRigel

    • If you're that concerned about power, why aren't you using an actual embedded processor?! It never ceases to amaze me how many people insist on using x86 based processors in embedded systems when there are lot of other processors, such as MIPS, PowerPC, StrongARM and SuperH, that are specifically designed for these types of circumstances!

      Christ! We're already locked into the x86 architecture monopoly for personal computers and servers, why do we need to bring this into the embedded market where it was never meant to be used?! Use the right tool for the job. Embedded systems are a lot more than an old PC in a little box.

      - j
      • by wowbagger ( 69688 ) on Wednesday October 03, 2001 @10:07AM (#2383397) Homepage Journal
        First of all:
        Sometimes you have to run other people's code, and you may not have control over what they targeted.

        Second of all:
        The cost of a ready built X86 CPU board is a fraction of the cost of a ready build board with the other chips.

        Third of all:
        For a low-volume house, the cost of building your own CPU board is rediculous.

        Forth of all:
        Most of your "designed for embedded" CPUs don't have an FPU. The embedded PowerPC, the StrongArm, and (IIRC) the SuperH and embedded MIPS don't have FPUs. If you need number crunching power, you need to go elsewhere.

        Unfortunately, some of us work under all those constraints (also, time to market, availability of second sources, and such).

        I suggest that if you are going to tell somebody with decades of professional embedded experience how to do his job, you should try to have a little yourself.
        • Wow, looks like I hit a nerve! Well I work in the embedded market too, and I've experienced exactly the same engineering pressures as you have. Perhaps the x86 processor really does fit your system requirements, but if you're going to use it, accepts its limitations: don't go complaining about 20 watts power consumption.

          What I stated above is still perfect valid: obviously power consumption isn't that big a deal for you or you'd be using a better processor.

          - j
        • by Anonymous Coward
          For the FPU, at least IBM already announced that the FPU would be an optional add-on on their embedded 440 CPU. And at 1 multiply-acccumulate per clock with clock around 500MHz it's quite a good one.

          Note that the Motorola 82xx chips also have an FPU: a 82xx, is basically a CPU with host bridge and an interrupt controller, connect PCI peripherals on one side, memory on the other and you're done. Yes, I do PCB myself and I know that it's not that simple, but getting midrange performance at low power levels is rather easy.

          I'll concede you that the FPU on the 82xx is not top of the line, but it depends on what your needs are. Besides that, the problem with these FPU is that you should avoid divides, which are slow and block the pipeline for too long. On the other hand, the fused multiply-add does wonders for polynomial evaluation and filtering (when you factor in power consumption, they are actually excellent) which should be the bulk of processing in embedded systems.

          Finally, Motorola also recently announced the MPC5xxx series which include floating point and are truly embedded devices, at least if they continue in the vein of the MPC5xx series. Their clock frequency is rather low, however. But it may be sufficient for many applications.

          FYI, I'm using old 603E in a data acquisition system and they work fine for what I do. The top CPU usage is around 25%, worst case (most of the time I'm well below 5%), but at least I don't need a huge power supply, fans on the processor (despite the fact that the system is installed in an observatory at high altitude, hence with rarefied air) and have much less disturbance than with an Intel or AMD chips whose power consumption changes by 50 watts in a fraction of a microsecond when they enter or exit power saving states. This last point is especially relevant since there is a lot of extremely sensitive analog electronics in this system and around it (did you ever chase the nanovolt, I'd bet not ?).

          • Actually, I do (well, I personally don't, but folks on my team do). My gizmo does RF from DC to 3GHz, down to -120dBm, so noise is a concern. I'm guessing you do image processing, and are looking at a CCD, so you can at least bandwidth limit your signal. Wish I had that luxury.

            Could be worse. My brother worked on CCD gear that would be fried by an ESD of 35 volts. LHe2 cooled, fun stuff.
        • >> First of all:
          Sometimes you have to run other people's code, and you may not have control over what they targeted.

          And your point???

          It's called linking baby, get with the 90's. If the old code sucks, build a lex for it and bring it forward. fuck this legacy shit in the gsa.

          The crusoe sucks from just about all perspectives... Did any of the talent stay there ater jan99?? Did someone recruit them all to another company that had a clue? hmmmmm.... The x86 is for fools who don't know how to chose their embedded dev environment.

          I.E., if your're here, i'm sorry and i completely understand that sometimes the boss-man sticks the ungreased pole up your ass because he understands a spreadsheet and little else but still, get a clue..... anyone doing embedded work on an x86 has got a serious legacy problem (like working for a telco or govt).

          Get with the plan and do your embedded work on a mips or ppc core where shit's so fast it makes your eyes bleed at 1.5v.
          if you can give me even one reason why x86 is better, i'll eat these words. Surprise me. ;)

          alan
          • Obviously, if I have access to the source, I can recompile. What I meant by "running other people's code" is "running other people's binaries for which you have no source", an unfortunate occurrance that sometimes happens to real engineers but that posers like yourself have never encountered.

            As I said in some earlier posts, I really dislike kiddies who think they actually know something. Try growing up, learning to express yourself without being obscene, and getting some real-world experience. Engineering is about making do with what you have, not living in some fantasy perfect world.
    • by Anonymous Coward
      ICP makes 3.5" size board with TM3200 400MHz and I'm sure it's much less than $700 you are paying. Go to part no WAFER-6820.
    • The Lippert CoolRoadRunner ][ has a 300mhz cyrix mediagx, isa/pci, eide, cf, vga, audio, serial, parallell, and 10/100 ethernet and pulls less than 10 watts. It isn't cheap at $560 a pop though.

      burris
    • >> I've been eagerly awaiting one of the Single Board Computer (SBC) venders to make a device with a Crusoe on it

      WTF are you talking about?? You must not be an embedded developer. The crusoe is irrelevant to just about everyone....

      So, let me ask, what exactly are you waiting for???

      it certainly not a job as an embedded programmer. feh, crusor for fools and market wishers...
  • Cool, so in a year I can get a cheep curso chip to build my mp3 player for my car with. >:)= Soundas fun

    my 2 cent plus 2 more
  • by serps ( 517783 ) on Wednesday October 03, 2001 @08:37AM (#2382961) Homepage

    When you look at the specs of (or the hype, if you hold that opinion) the crusoe chips, it boggles the mind why they didn't push this this sooner. 'Runs on one AA battery', goes the sales pitch - perfect for the embedded systems which need a half-decent chip but need to conserve battery life (i.e. practically all of them)

    A friend of mine was so hoping to use a Crusoe in his embedded web browser tablet thesis project back in 2000. Pity it was released too late for him :)

    • by Anonymous Coward
      Transmeta has a low-power product which they want to sell at a premium that solves the x86 compatibility problem. If they can stay alive long enough, they have an architecture whose performance they hope to scale and compete on pure performance.

      The embedded market needs low-power, but is a low-premium, high volume market where people are willing to rewrite applications to the chip. Therefore it lacks the premiums that Transmeta wants, and lacks the requirement for x86 compatibility which is their greatest strength.

      Similarly, and ironically, Linux doesn't help Transmeta any either. Linux is a highly portable OS with a lot of useful software which is readily ported along with it. This undercuts Transmeta's interest in marketing to people who need x86 compatibility.

      Therefore until they can achieve competitive x86 performance (which was their original goal), their best target market is long-lasting Windows notebooks. Because they are tied to end-user Windows, they need to run on x86. Because they are notebooks, battery life is a requirement. By luck the business market can pay top dollar, so they can get a good premium.
  • by Junks Jerzey ( 54586 ) on Wednesday October 03, 2001 @08:41AM (#2382976)
    Maybe you don't blink at paying an extra $100 for a fancier CPU for your home machine, but in embedded systems price is a huge part of the deal. If you can shave $1 off of a CPU then that that's a big difference when you're selling a $50 appliance. Transmeta's chips are even *more* expensive than offerings from other companies, so it isn't at all clear how they're going to go up against processors like the 8051, ARM, MIPS, and so on.
  • Business Strategy (Score:4, Interesting)

    by BadBlood ( 134525 ) on Wednesday October 03, 2001 @08:58AM (#2383038)
    As a stock holder of TMTA (ugh, lately :), I am a little concerned at their changing strategy. Generally speaking, changing business strategy so early in the game for Trasmeta speaks of trouble. It is especially difficult to come to market and try to play catch up with a product that's similar to what already exists, i.e. notebook cpu market. It is even more difficult to switch business strategies and succeed.

    Having stated that, my opinion is that this company will end up succeeding. I have trouble believing that such talented individuals (Ditzel, Torvalds, Taylor, etc.) will end up falling on their faces. As a stock-holder (I bought in @$3), my worst-case scenario is that their technology gets bought by one of the bigger players (AMD, Intel). Best-case scenario is that they finally find a product that utilizes the competitive advantages of the Crusoe processor - whatever that may end up being.
    • When I read your post I had to go read the article because from what you wrote it sounded like Transmeta was going out of the notebook market. They aren't (for anyone else reading)... I don't see how their "change in strategy" is a big deal or such a change - from the article it sounds like it will compliment their current offerings. Also they need something for all those chip designers to do :).
    • your best case is that they get taken over because they are not going to make any money when up against
      ARM (intel& TI), MIPS (NEC, IDT, PMC) and PowerPC (IBM & MOT)

      really its nothing more than fancy VLIW machine

      in terms of selling it they have not got a chance outside of the low power x86 market
      (where legacy apps need x86 otherwise we would use something else witness MS useing ARM and only ARM for the new WinCE XP)

      get real they are screwed because they tried to use the best fabs to get one over on intel (TSMC fabs) and failed

      regards

      john jones

    • Sell your stock now. This is actually the SECOND time Transmeta has changed business plans.

      First they wanted to compete in the Notebook market and failed.
      So then they tried to compete in the Server market and failed. (Who in their right mind would use a Crusoe in a friggin server? To save 5 degrees in temp? Gimme a break.)
      Lastly they are trying to compete in the Embedded market, and will fail.

      Sell your stock now, while you still can. Transmeta has no chance to survive, make your time.
  • First it was ultraportables, then portables, then servers, now embedded?
    Transmeta's abject lack of a business plan or any semblance of forward-looking pretty much guarantees that Transmeta can look forward to being embedded in the garbage heap of dot coms.
    No doubt it was sexy technology, just not sexy or business savvy enough to overcome the big boys. Too bad.
  • where to get one: (Score:3, Informative)

    by gtx ( 204552 ) on Wednesday October 03, 2001 @09:20AM (#2383118) Homepage
    Wow, when can I get my hands on a cheap embedded Transmeta system to play around with?
    try here:
    Crusoe-based PC/104-expandable single board computer [linuxdevices.com]
  • If this gets applied, it would make cyber clothes more of a reality, since the last thing I want to do is have huge battery packs in my pocketess.

    Besides, It's cool for the winter...

  • Embedded Linux Journal [linuxjournal.com]

    Omega Engineering [omega.com]

    Computer Boards [computerboards.com]

    Your comment violated the postercomment compression filter. Comment aborted

  • Poor Transmeta (Score:3, Interesting)

    by ZigMonty ( 524212 ) <slashdot&zigmonty,postinbox,com> on Wednesday October 03, 2001 @09:42AM (#2383274)
    Ever since I first read about them in Spectrum, I admired them. Here is a company that has a very original idea and the CS behind it is very cool.

    IMHO, the embedded market isn't as big as they claim. And if it is, why would anyone want x86 compatibility? The only use I can see for their current design is what it is doing now, Low power x86 compatible PC chip.

    Personally, I would love to see how well the Crusoe goes running native code. The whole super-RISC VLIW architecture sounds a lot better than the x86 or even the PPC. Maybe they should try native BIOS AND a native Linux kernel. In fact, have everything possible native and leave the whole x86 'code morphing' bit for binaries you don't have the source for.

    Another idea, since they are emulating/binary translating the code anyway, how big a job do you think it would be to have multiple virtual machines? Naturally they'd all need their own RAM but RAM is cheap. It could give dual booting linux with windows a new meaning!

    Also, how easy would it be to add other architectures? I can see a PDA that not only runs Windows CE software but PalmOS stuff as well. Just some ideas, any of them feasible?

    • Maybe Transmeta will be writing a Z80/8051/6809/etc 8-bit code morphing engine for their processors. Then they could easily say that they had some kind of mega-beast of an embedded chip! Their processor is 256-bit VLIW, which is a tad overkill to emulate in hardware an 8-bit processor, but if that means a 500MHz Transmeta CPU equals a 10GHz Z80... :)

      The reason that the VLIW architecture is not suitable on its own is problably due to non-standard address/data widths going into the core of the processor. This is fine when you are doing the entire system, but if you need to interface with 64-bit memory, etc, then you want standard word sizes. Hence 32-bit processors, not 38-bit or 44-bit processors even those could turn out to me more convenient for the particular ISA that processor might implement.


  • Transmeta realises its done nothing in notebook market and makes desperate dive for embedded market. ARM and TI described as "not worried".

    Every little piece of Transmeta news gets broadcast as bible.
  • I know it's fun to talk on /. about how cool PC, chips and gadgets are. It's nice to have new stuff, and it's nice to have cool stuff. "A cell phone that will play my MP3 files with a garage door opener and streaming wireless video? Cool!"

    But I work in the embedded systems sector. It's the plumbing of world industry, and it's not cool. There are a lot of factors when it comes to embedded systems, but the number one factor is "cheap." Systems must be built cheap. (I'm ignoring reliability and other factors for the sake of the argument.) That means 486's and Pentiums, not the bleeding edge of 2GHz cool chips. Code is written in assembly, C and uncool languages, not Java and Perl. OS's include VxWorks, pSOS, LynxOS and DOS. Not cool ones like OS X, XP, etc.

    My point is that it's nice for Transmeta to get into the embedded market, and I hope that they do well, but the "that would be cool" posters don't have a clue about how to make it happen, and neither will statements by Transmeta CEO Mark Allen, like "by this time next year, it could equal the notebook market." This industry is dominated by players that have been there for 30+ years, and if Transmeta wants to make a mark--as it seems they must if they're going to stay alive--they better have a pretty compelling offering in the works.

  • Can a Transmeta chip emulate a Java VM and run Java byte code natively. Or is Java byte code not low level enough.
    • The folks at Transmeta have already done this. They even apparently implemented a version of Doom, with the main loop in picoJava code (Sun's Java processor core) and the rest in x86 code. The Transmeta chip "morphed" both instruction sets into its native VLIW instructions, and seamlessly executed both.

      This was demo'd at the Crusoe launch, but was not intended as a product. I don't know the current status of any Transmeta products related Java - does anyone else?

  • What about ZFLinux? (Score:3, Informative)

    by certsoft ( 442059 ) on Wednesday October 03, 2001 @11:13AM (#2383739) Homepage
    Next year, the San Jose, Calif.-based company will come out with an inexpensive system-on-a-chip that fuses a processor, a chipset and a graphics chip as well as a new high-performance version of Crusoe, said Dave Ditzel, Transmeta's chief technology officer.

    Sounds like they are trying to get into the same area as the ZFLinux chip. [zflinux.com]
    Has anyone seen any power consumption comparisons between the two?

  • by Anonymous Coward
    Around a $1.30.
    As high $50 last November.
  • Flaming testes!

"An ounce of prevention is worth a ton of code." -- an anonymous programmer

Working...