Forgot your password?
typodupeerror
Portables Hardware

Toshiba Launches First Cell-based Laptop 172

Posted by timothy
from the my-laptop-has-been-cell-based-forever dept.
MojoKid writes "On Tuesday, Toshiba launched the Qosmio G55-802, the first laptop available with the Cell CPU. Yes, think PS3 technology, developed jointly by Toshiba, Sony, and IBM. However, in particular, the Cell CPU is not about gaming, but about the multimedia experience. Taking the load away from the Intel CPU, the Cell processor performs gesture control, face navigation, transcoding and upscaling to HD. Interestingly (and necessary, with 4 GB of RAM), the system comes with 64-bit Vista installed by default, but 32-bit Vista ships as an option as well." However, semi-relatedly, if you'd prefer your Cells run open-source code, 1i1' blu3 writes "IBM's put up an open source project downloads page for the Cell processor — APIs, toolkits, IDEs, libraries, algorithms, etc. Most of the stuff on it right now is from SourceForge, but they are asking for user contributions to add to it." (Terra Soft's also been providing a Cell-compatible Linux distro for a while now, and according to Wikipedia the kernel's supported it since version 2.6.16.)
This discussion has been archived. No new comments can be posted.

Toshiba Launches First Cell-based Laptop

Comments Filter:
  • And who is going to start writing custom built apps or patches to utilize the hardware on this one laptop?
    • by electrosoccertux (874415) on Wednesday July 16, 2008 @02:59PM (#24217431)

      Probably the people who buy it.

      Linux will support it of course. Offload video decompression (XVID/MPEG4/whatever) and audio decompression to the PPEs.

      I'm sure Audacity and any other audio processing tools will support it. The GIMP could make use of it as well. Matlab of course.

      Those are just off the top of my head.

      • by negRo_slim (636783) <mils_oRgen@hotmail.com> on Wednesday July 16, 2008 @03:04PM (#24217539)
        When is the last time a company come along with a proprietary technology like this and received market acceptance? I seem to recall companies like Aegia and Rambus... Hell RDRAM didn't even require any change to software and provided higher performance, but it was one company and of course the price remained high.... Nah I don't expect to see this go much further than it already has, a few people will buy it and it will ship with some in house programs on it and just like every other system from a major manufacturer support and software will slowly fade into obscurity. Sure a few websites will be started in dedication of this thing and they will light up their message boards with how superior and awesome it was/is/could of been...
        • by bucky0 (229117)

          A good example would be PhysX processors or ultra-high end soundcards, NOT rambus. Rambus was prohibitively expensive and high latency to boot. Dual channel DDR + clock scaling came across and ate its lunch.

        • Availability (Score:5, Interesting)

          by DrYak (748999) on Wednesday July 16, 2008 @09:23PM (#24222441) Homepage

          I seem to recall companies like Aegia

          The big difference is the availability.
          Although both the Cell and the PhysX share some architecture design, it's about the only thing they have in common.

          PhysX was only available on 1 single type of board. No tools available at all to develop code for the chip, only a physics library which only provided 1 single API.
          The only thing you could do as a user is buy it, stick it into the computer and hope that game developper will release patches supporting it.
          The only thing you could do as a developer is write some physics simulation into the game you're developing.

          Cell has lot of tools to develop code to run of it. Including open source compilers (gcc for example), and including frameworks dedicated at doing stream computing (RapidMind can produce SPE code). Thanks to the fact that its main CPU part is a plain simple PowerPC, there is even a lot of prior knowledge that can be recycled.

          And the Cell is available on lots of devide ranging on small device on which the would-be developer can test some code like PS3 (compatible with Linux out-of-the-box) and this laptop (x86-based with Vista, but offers a cell as a coprocessor) all the way up to big servers with several cell boardlet inside, ready to do some crazy super computing for scientist.

          Anyone can develop for Cell and run pretty much everything they want on it, and even have access to a significative range of platform to test the code.

          The cell is much more likely to experience some success that the PhysX did.

        • What's up with Op Ivy in your sig?

        • What's this proprietary thing you speak of? IBM has free SDKs for the CBE (Cell Broadband Engine), the PS3 itself can have Linux loaded on it (mine runs Ubuntu for the record) and the documentation on how to program it is very public.

          Even the game development platform for the PS3 is highly based on open source and free software toolkits. Stop with the FUD.

      • Chances are big that the laptop would remain an expensive toy for rich geeks.

        Linux main driving force are people who try to make out of literally junk computers something usable (and at large they succeed). Majority of people are those who can't afford computers nor proprietary OSs - and they have natural interest in such stuff.

        I'd say that if the laptop costs below $600 mark - then it might have chance. But something tells me that at $1600 it would find few followers.

        P.S. But you might expect rab

      • What about ps3 emulation?
      • by Ilgaz (86384)

        Toshiba/IBM guys should get some commit access to ffmpeg/x264/VLC and start working on enhancing the tools for Cell CPU. ffmpeg or x264 aren't "nerd things" anymore, they are used by every kind of user embedded in easily used products.

        Firefox became huge and everyone optimises for it or fix its code right? If I remember correct, the Download record was 10 million+... Look at VLC without any kind of record campaign, you will be amazed: http://www.videolan.org/vlc/ [videolan.org]

        It is 92 million right now and counting.

    • by LWATCDR (28044)

      I don't think that will be too much of an issue. My guess it will be used as a device like a GPU or DSP on a sound card.

      • by negRo_slim (636783) <mils_oRgen@hotmail.com> on Wednesday July 16, 2008 @03:10PM (#24217633)

        I don't think that will be too much of an issue. My guess it will be used as a device like a GPU or DSP on a sound card.

        And we all know how easy it is to add hardware acceleration for products from major vendors like Creative, Nvidia, AMD... The architectures these companies have produced have far greater market penetration then Toshiba could dream to see and yet there isn't across the board support for such common devices. Not many are going to be coding for a piece of hardware only one manufacture is producing. Unless this thing is seen in rigs across the board and/or it demonstrates a highly tangible benefit it simply won't be supported by the fast majority of software.

    • I've been thinking about porting a CFD code to PS3, but this would be interesting. There's been at least one company making Cell add-in cards but they have been prohibitively expensive.

      It's kind of like using your GPU for fast math. No wait, the exact same thing :)
  • by ALecs (118703) on Wednesday July 16, 2008 @02:53PM (#24217289) Homepage
    According to the article (and hinted by the summary), the thing has an ordinary Intel Core2Duo CPU. I'm assuming the cell is the "Toshiba quad code HD Processor" mentioned in the article. So it's a co-processor, then. My best guess it it's a 4-SPU cell processor without the PowerPC core. Weird...
    • by crabbz (986605) on Wednesday July 16, 2008 @02:55PM (#24217347)
      Yes, it isn't a Cell, it is Toshiba's Spurs Engine [wikipedia.org] with 4 SPEs and no PPE.
      • Re: (Score:3, Interesting)

        by ALecs (118703)
        Wow - this could be a real win for scientific computing. Traditional GPU-based computation is hindered by the poor I/O (I guess PCI-e is fixing this but I don't know). And PS3-based cell computation is hindered by lack of RAM _and_ poor I/O. But this thing looks like it's hooked up to some nice RAM and a PCI-e bus and could really crunch some FFTs.

        We have an experimental PS3-based cell data reduction system here but it's just too slow. I can almost hear my phone ringing as one of our scientists starts ask

        • Re: (Score:3, Informative)

          by stephentyrone (664894)

          Single precision only, non IEEE-754 arithmetic isn't a "real win for scientific computing". It's a win for getting the wrong answers really, really quickly.

          Yes, I know that there are problems for which the limitations of the SPEs don't kill the accuracy of the solution, but people (even scientists) rarely do a complete analysis of whether or not their problem is one of those before they set off to use the new faster hotness.

          • Re: (Score:3, Insightful)

            by MatanZ (4571)

            Most scientific calculations use integer opertation of the CPU, impelementing their own Floating/Fixed point if needed. The type of calculations for which single precision gets the wrong answer really quickly, but double precision is perfect is very rare.

            • This was true once (before Wilkinson/Kahan/others showed that you could do meaningful error analysis for floating-point all those years ago). It's not really true anymore. The vast bulk of scientific computing is now done in floating-point, except for specialized problems that require more than double precision, or problems that are fundamentally discrete in nature and are best performed in integer.

              The performance benefit from using hardware floating-point vs. soft float is simply too huge to be left on t

            • That's not true. For quad / multi-precision it makes sense to do the work on FP words in the FP pipes rather than integer. It's faster and the free normalisation helps things out. There are well-known tricks like Dekker splits to ensure bit-accuracy across multi-precision values.

              Even for integer work (like crypto) people are moving towards using FP because it can speed up calculations. Not that FP arithmetic is intrinsically faster, it's just that manufacturers throw area at FP units and so the throughput i

        • by PopeRatzo (965947) *

          Wow - this could be a real win for scientific computing.

          I'd like to see what a company like Metric Halo could do with using this Cell processor as one or more DSPs as an onboard digital audio adapter.

          I've often wondered why there aren't portable audio-centric systems for DAW users like me. It would be great to have a portable box that I could use for live performance or field recording that wouldn't require some outboard Firewire audio adapter. Whenever I have to carry a Powerbook or Thinkpad with a fir

      • by Sockatume (732728) on Wednesday July 16, 2008 @03:23PM (#24217817)
        So combined with the comment above that Ben Heck already did a PS3-based laptop, we find that the only accurate words in the headline are "toshiba", "launches", and "laptop".
    • Re: (Score:3, Informative)

      by negRo_slim (636783)
      This appears to be some earlier info on the kit they are using in this laptop, here [reghardware.co.uk], with pictures!
  • http://home.comcast.net/~SupportCD/XPMyths.html [comcast.net]

    Modern hardware except one particular Pentium M stepping (which was popular for a while) handles PAE. 64G RAM on 32-bit

    • You are a rube.
    • Re:PAE mode? (Score:4, Insightful)

      by Anders (395) on Wednesday July 16, 2008 @03:08PM (#24217595)

      Modern hardware except one particular Pentium M stepping (which was popular for a while) handles PAE. 64G RAM on 32-bit

      But Windows does not [microsoft.com].

    • Re:PAE mode? (Score:4, Insightful)

      by QuantumRiff (120817) on Wednesday July 16, 2008 @03:18PM (#24217757)

      Because of the way that windows "pages" memory (and I'm assuming your running a server version of the OS, cause XP and vista don't work with PAE) you still can't have a single process with much more than 3GB of ram on a 32bit system. You can have multiple processes running with 3GB of ram, but then you get some slowdowns from paging in and out the memory.

      If Memory serves, this is part of the reason that Exchange 2007 requires 64-bit OS's and processors. (except for the demo and SMB versions)

      • No, that's because of how protected mode works. Also, you have 78 processes running, and you want one with more than 3GB of RAM? Browser uses 100M, outlook uses 250M, music player uses 80M....

  • Having a laptop that does Gesture based and Face navigation...So what happens if you give it the finger? Does it shut down? or does rolling your eyes bring up email?
    • Re: (Score:3, Funny)

      by gdog05 (975196)
      I think giving it the finger will boot Windows, and rolling your eyes will start Firefox and head to YouTube. Sneezing of course, will run an anti-virus scan.
  • Big, big let-down (Score:3, Interesting)

    by Jah-Wren Ryel (80510) on Wednesday July 16, 2008 @03:04PM (#24217523)

    I thought this was going to be the first fuel-cell based laptop.
    Especially after reading how a fuel-cell the size of a regular battery can operate a cell phone for 2,700 hours of talk-time. [slashgear.com]

    • That's a bit unclear. They didn't claim that you can run a phone for 2700 hours. You still have to recharge the fuel cell every few hours like a battery. They are saying that's how long the same fuel-cell can be used without severe charge deterioration. In other words, you wouldn't have to replace the fuel cell as often as you have to replace rechargeable batteries that have lost too much of their capacity for recharging. It's a an improvement, but not as large as it seems at first glance.
    • by Aetuneo (1130295)
      But you can't take it on airplanes. In fact, if you even bring it to the airport, you're going to be taken aside for additional questioning, and then put on a flight to ... well, that destination is classified.
  • by ivan256 (17499) on Wednesday July 16, 2008 @03:05PM (#24217549)

    For almost 10-years now, Slashdot has pipmped Terra Soft and Yellow Dog. There's Debian, Fedora, Ubuntu, Gentoo, and SuSE (you know, distros people actually use out there) available for the Cell processor and PS3, and Slashdot shills for Terra Soft. This was true back when PPC linux was mildly popular too... Debian, Slackware, SuSE... They all supported it, but Slashdot pimps Yellow Dog. What gives?

    • by Jason Earl (1894)

      Maybe Malda really liked the movie "Old Yeller".

    • by Otter (3800)
      I've submitted a number of those stories, and from my side I've been surprised at how little interest there is in them compared to yet-another-video-card and vaporware stories, let alone throwing-raw-meat-to-the-mob stuff about SCO and the RIAA.

      Anyway, there's a huge difference between a distro specifically designed for a platform and some half-assed port. I don't have a PS3, but can tell you that there was no comparison between Yellow Dog and things like SuSE PPC that maybe could install a bunch of recomp

      • Re: (Score:3, Informative)

        by ivan256 (17499)

        As somebody who bought a PowerBook G3 when they came out specifically to play with Yellow Dog Linux on it, my experience was the opposite. Perhaps you just don't like SuSE?

        My experience was that Yellow Dog was a half-assed port of RedHat to PPC, and Debian for PPC was Debian. With Yellow Dog you felt like you almost had a working RedHat system, but things were out of date, and many of the things you were used to were unavailable. Debian had none of those problems.

        Admittedly, I've not gone back and tried Yel

      • by PopeRatzo (965947) *

        I've submitted a number of those stories, and from my side I've been surprised at how little interest there is in them compared to yet-another-video-card and vaporware stories, let alone throwing-raw-meat-to-the-mob stuff about SCO and the RIAA.

        Otter, maybe there's something about the stories you submitted besides the content or point of view that's keeping the interest-level down.

        • by Otter (3800)
          Perhaps, but I always confuse "it's" and "its", link perfectly obvious words to Wikipedia and conclude with an idiotic "question" like "Could this be the end of x86 on the desktop?" Is there anything I'm missing?

          Anyway, my point was less to complain about my submissions than to respond to the OP's conspiracy theory. TerraSoft makes genuinely interesting stuff, and I certainly don't think they're getting too much attention here!

    • Re: (Score:3, Funny)

      by mhall119 (1035984)

      Yeah, Slashdot has been giving Ubuntu nearly enough coverage!

    • [...] and Slashdot shills for Terra Soft.

      TerrSoft was one of the initial (and major) developers of Cell support on Linux. As well they were more or less official supplier of Linux for PS3.

      They as well remain one of the major PowerPC/POWER supporters and developers of Linux on PPC*/POWER*. (Largest PPC users now are in embedded market - not in desktop/workstation market where TerraSoft is working.)

      Rest of distros, in large part, merely have used the GPLed work done by TerraSoft.

      That might sound like shilling, but the guys deserve credit f

    • by nurb432 (527695)

      FreeBSD and NetBSD also support PowerPC remember.

      NetBSD has for a long time now.

    • by Ilgaz (86384)

      Do you actually have a PowerPC based machine? Except Debian and Gentoo there is a huge difference between "supported" and "designed for it".

      I actually tried YDL on my Quad G5, it was no different then installing/running OS X Leopard. It was that smooth with everything working by default.

      Ubuntu? Tell them to revert back to officially support PowerPC, the POWER family in fact and apologise us, PowerPC owners first. Someone should have reminded them IBM, Cell, POWER, Toshiba, Sony, the Power.org foundation. Yo

  • Please, someone! The new PowerPC's look to be combined with Cell like this in the U of Illinois supercomputer (to dwarf Roadrunner). I don't mind an asymmetric combination as long as x86 is out. The Linux version should be good, but Linus has maintained that x86 will be the most important target for awhile. I can't see it handling the helper cores very well--it's just barely getting NUMA working well.
  • shameless (Score:1, Troll)

    by Gewalt (1200451)
    another shameless slashvertisement. Just because someone does something stupid with the wrong type of hardware, that doesn't mean slashdot needs to feed hothardware more ad revenue.
    • Actually to me it looked moke like Toshi, while the giants - nVidia/ATI/Intel - are wrestling on GPU-CPU split, tries to stab them in a back.

      If they had ever tried to deliver on promise of cheap Cell, they might have already won the ongoing CPU acceleration war.

  • Good price too (Score:3, Informative)

    by Scotteh (885130) on Wednesday July 16, 2008 @03:07PM (#24217581)
    It's only $1549.99 which is the average price of Sony VAIOs [sonystyle.ca]
  • by Anonymous Coward

    Windows XP?

  • PS3 + linux = shit (Score:2, Informative)

    by Anonymous Coward

    The RSX is still locked away and there is no decent video driver. It's like using an old Pentium machine.

    • by Tweenk (1274968)

      Because somebody could figure out how to make Linux PS3 games or buy PS3s for the hardware alone while completely ignoring the games, and this could theoretically leave Sony's asshole in ruins, because they sell the consoles at a loss and make it up with high margin game sales...

  • by JSBiff (87824) on Wednesday July 16, 2008 @03:16PM (#24217723) Journal

    I've been wondering, if Cell technology were integrated into general purpose PCs, what kind of tasks it would help with. Could it be used to accellerate. . .

    * Crypto functions (like whole-disk encryption, or encrypted volumes (like TrueCrypt)?

    * High resolution video decoding, so the processor doesn't have to chug so much on it? From the article, it sounds like this might be one use of the cell?

    * Grid computing - things like World Community Grid, distributed.net, SETI@home etc? I imagine this probably depends, at least in part, on the specific types of computations being done for the project you participate in, but would you commonly be able to do more computation, faster, for those types of projects if you had cell processors?

    Can a GPU like one from Nvidia or ATI potentially work together *with* the cell processor to increase the GPU's capabilities? (I'd guess that would probably depend on the drivers having support for the Cell, and I'm guessing that current generation drivers probably wouldn't take any advantage of the Cell?)

    • by schwaang (667808)

      Can a GPU like one from Nvidia or ATI potentially work together *with* the cell processor to increase the GPU's capabilities? (I'd guess that would probably depend on the drivers having support for the Cell, and I'm guessing that current generation drivers probably wouldn't take any advantage of the Cell?)

      Not likely, since they are competing with each other for the same markets (acceleration of graphics and computation, including all the examples you gave like encryption, etc.). They are two approaches to

      • What's the price of a Cell vs a GPU? Nowadays, I think you can get a pretty decent GPU for less than $100? Still, I wonder if, for low-end systems, would it make sense, financially, to use a low-end CPU (like an Atom or Intel's 'economy' Pentium mobile CPUs), which wouldn't maybe be able to handle high-def video, and add on a Cell? Although, at that point, it's probably cheaper to get a Core2 CPU, and not worry about GPU or Cell.

    • by mhall119 (1035984)

      At that point you'd probably be better off with a couple of FPGAs.

    • My guess would be - Yes, No, Yes.

      For second - HD video decoding - to be any efficient, it has to be done very close to video output. Otherwise, HD video might take up some huge chunk of internal buses bandwidth. After all you first have to transfer data from RAM to Cell, then from Cell to video adapter. (*) That can get really messy without proper integration of Cell into computer architecture. And as long as Cell isn't main CPU, the problem would remain. PS3 doesn't have this problem because Cell is us

      • (*) Fact: CPUs for quite some time are capable of decoding HD video in real time - it's the transfer (1) to/from RAM and (2) video hardware which takes most of the time.

        Interesting fact, and I most certainly believe it. In fact, the marginal system memory bandwidth of most PCs is the very reason we have video cards (and now even sound cards) with local cache. We make the system memory slow because we want lots of it. We can afford to make local memories faster because there is less of it.

        I have to agree

    • by nxtw (866177)

      High resolution video decoding, so the processor doesn't have to chug so much on it? From the article, it sounds like this might be one use of the cell?

      It might be possible, but within a year or two, any system sold will probably be able to playback HD video.

      Most CPUs sold today can decode 1920x1080 MPEG2 and many can handle VC1 (including the 2.0 GHz Core 2 in this laptop), and just about any GPU sold today can deinterlace 1080i video (although some cards are better than others.) High-end/overclocked Core

  • by WilliamBaughman (1312511) on Wednesday July 16, 2008 @03:17PM (#24217739)

    Yes, think PS3 technology, developed jointly by Toshiba, Sony, and IBM.

    Saying that the Cell BEA was developed for the PlayStation 3 is like saying the wheel was developed for razor scooters. The PlayStation 3 uses the Cell, but the Cell was not made solely for the PlayStation. The Cell was developed to be a floating point and vector arithmetic monster that would be at home in a supercomputer, which it is.

    I have nothing against the PlayStation 3, but I get upset when a myth like this is perpetuated. Saying that one of the most powerful processors available today was 'made to play video games' detracts from it and gives readers an incorrect impression (in my humble opinion).

    • by eebra82 (907996)

      Yes, think PS3 technology, developed jointly by Toshiba, Sony, and IBM.

      Saying that the Cell BEA was developed for the PlayStation 3 is like saying the wheel was developed for razor scooters. The PlayStation 3 uses the Cell, but the Cell was not made solely for the PlayStation. The Cell was developed to be a floating point and vector arithmetic monster that would be at home in a supercomputer, which it is.

      I think you're reading his/her words incorrectly. The way I see it, the author wrote that you get technology used in the PS3 in your computer. After all, the Cell is PS3 technology since it consists of one.

    • by dbIII (701233)
      It looks that way because the way the cell is used. From looking around for a cell solution that compares to a dual quad core server I came to the conclusion that you need some kind of black ops military budget that you never have to justify to anybody to buy anything in the mid range with a cell in it instead of eight similar x86 machines. The PS3 is a very different story but has very little memory. Hopefully things like this Toshiba will help make it the commodity it should be instead of being conside
    • You can say that all you like, but being a hobbyist, try to get a cell for yourself. Go on, write up IBM! I'll sit here and wait while the same "sorry" letter gets mailed to you, informing you that it's not possible for you as an individual outside of sony to get the cell, and that they've got linux running now on the ps3.

      I should have scanned mine
    • Do you think the Cell would have been made without the guaranteed millions of sales?

  • Fail for Windows (Score:1, Interesting)

    by Anonymous Coward

    Do Toshiba sell these things without an OS?

  • by MDMurphy (208495) on Wednesday July 16, 2008 @03:29PM (#24217907)

    There's got to be some upper limit to be called a laptop. I looked at screen resolution first, it's 1680x945. It's an odd size, not as many pixels as some other laptops. Then I noticed the size in inches: 18.4! Base weight: 10lbs.

    I don't have a problem with large computers you carry from room to room with a built-in UPS. But at some point it's a desktop all-in-one or something else.

    • by doconnor (134648)

      The technical term is "Luggable".

    • by pimpimpim (811140)
      I don't have a problem with large computers you carry from room to room with a built-in UPS.

      With a laptop of that weight you need indeed someone from the United Parcel Service to carry it around for you :)

      Disclaimer: proud EEE owner. Actually that was what I was hoping for, couldn't they make a netbook with a cell cpu. I'd buy it! Housewifes would, too! As a PS3 can already play movies, games, and do internet stuff, it is actually commodity hard and software.

    • by Fred_A (10934)

      Then I noticed the size in inches: 18.4! Base weight: 10lbs.

      I don't have a problem with large computers you carry from room to room with a built-in UPS. But at some point it's a desktop all-in-one or something else.

      Actually it can hover on the exhaust of its built-in fan when you need to move it.
      (requires large battery sold separately, may reduce computer run time, depending on applications used)

  • by Tarlus (1000874)

    Cell CPU is not about gaming, but about the multimedia experience.

    Not sure about anyone else, but by my definition, gaming is a multimedia experience

  • Did anyone else parse Qosmio as Quasimodo?
  • This is not a Cell based laptop. It's a PC laptop with a Cell processor inside.

    The Cell is a cool add-on but it does not make this a very interesting laptop by itself.

    I would love to see a pure Cell-based laptop, mostly because it would be a decent performer and an outstanding number-cruncher. The fact that it would be completely Windows-proof would be a nice bonus.

  • Anyone else curious about what kind of battery life this thing gets? I know that the ps3 sucks enough power to black out a small country, so I'm interested (read: worried) in how long this thing will last on battery power.
  • Qosmio? (Score:3, Insightful)

    by Squiffy (242681) on Wednesday July 16, 2008 @03:49PM (#24218329) Homepage

    Qosmio? What a stupid name. It totally drips with unnecessary marketing affectation. It's like they were trying to cram as much cheesy bullshit into one name as they possibly could.

    Cosmo? Not quite.
    Cosmio? Hm, needs a little more bullshit.
    Qosmio! Yes, good job. That 'q' really ratchets up the puke factor. Well done.

  • by Doc Ruby (173196)

    I hope the appearance of the Cell in actual PCs, not just the RAM-hardwired and GPU-lockedout (and no PCI) PS3 will reignite official support of Cell Ubuntu. Until last year, Ubuntu was officially supporting the PPC-based Cell version of their distro. Now it's just a community effort that needs your help [psubuntu.com]. Ubuntu is working, with some bugs (right now mainly the installer, and beta bugs in the Cell SPE video driver). If there were more diverse Cell PC HW, and a larger, more diverse developer community coming

    • Not a Cell CPU (Score:3, Informative)

      by Doc Ruby (173196)

      On closer examination of the specs, this laptop isn't a Cell CPU at all. It's Toshiba's "Spurs" coprocessor, which is like a Cell but with the central PPC core stripped out and only half the Cell's 4 SPE DSPs, hooked up to a Pentium Core 2 Duo instead. That might be an interesting platform for experimenting with Linux and DSP, but it's not a Cell, and has practically no relation to any Cell/Linux project, nor Ubuntu in particular.

      Both the Slashdot story and the actual article lie about the CPU being a "Cell

  • I'm just surprised that I haven't seen anyone mention that this machine is actually quite capable and very reasonably priced.

    You get a decent Intel Core2 Duo processor, a fantastic graphics card and, oh, by the way, a 4 SPE unit Cell processor. You also get two 250 GB drives that you could probably run in RAID to increase speed, an 18.4 (!!) inch LCD screen. All for around $1550.

    Seems like an ideal machine for someone who might want to start developing for the Cell or for entire research, business or
  • Interestingly (and necessary, with 4 GB of RAM), the system comes with 64-bit Vista installed by default, but 32-bit Vista ships as an option as well.
    .

    Vista 32 with 4 GB installed will reserve about 1 GB of RAM for the OS, GPU. etc.

    I would expect Vista 64 to do the same.

    I would also expect an NVIDIA 9600M to be reasonably competent as a media player. What am I missing here?

    • by jmorris42 (1458) *

      > Vista 32 with 4 GB installed will reserve about 1 GB of RAM for the OS, GPU. etc.

      > I would expect Vista 64 to do the same.

      You might expect it, but you would be wrong. 32bit Microsoft OSes (desktop, not server) are strictly 32bit flat memory model, thus can only address 4GB, from which the area reserved for the PCI bus and other I/O must be deducted. Depending on chipset this can vary a little, but is normally at least 512MB.

      Much the same way DOS could address 1024KB but because the top of the addr

  • Quosmio (Score:4, Funny)

    by bugs2squash (1132591) on Wednesday July 16, 2008 @09:52PM (#24222717)
    Esmirelda; the cells, the cells...
  • A desktop PC I could understand,but a laptop? From what I've read the cell is simply not built with power conservation in mind. I can only imagine that in a laptop it will kill the battery in no time flat. And how much heat does a cell chip generate,anyway? IMHO while the cell might be great built into a desktop machine,it just sucks too much juice for a laptop to be practical. And I wonder why they aren't selling an add-on card for desktops? I can imagine there are a lot of folks that would happily buy a c
  • Since the Cell processors are a couple of years old now and GPUs are being used more and more for offloading some computation, wouldn't the latter be the way to go? Or is the Cell architecture so fundamentally different that it is much better suited for some tasks than GPUs?
  • I was looking for a new computer (could be a laptop or a desktop), and I'd be sold on this one, BUT:

    - I don't know if the hard drives can be used in a RAID0 setup (all indications are that they can't). Nothing pisses me off more than a slow storage device. It doesn't help that this model isn't available with 7200RPM drives, while the X305 is. SSDs aren't an option either, which isn't a plus.

    - IT'S NOT AVAILABLE WITHOUT SOFTWARE. I don't want to pay the Microsoft tax, and I sure as hell don't want Vista!

When all else fails, read the instructions.

Working...