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

 



Forgot your password?
typodupeerror
Linux Software

Linux Chosen for IBM's New Supercomputer 343

Uhh_Duh writes "news.com is reporting that Linux will be the main OS in the Blue Gene - IBM's $100m supercomputer project. The Blue Gene will contain 65,000 processors and 16 trillion bytes of memory." Wow. That's a lot of nuclear weapons simulations.
This discussion has been archived. No new comments can be posted.

Linux Chosen for IBM's New Supercomputer

Comments Filter:
  • The end of AIX (Score:2, Insightful)

    I guess this makes the demise of AIX official.
    IBM is pooling all its resources into Linux now.
    I suppose that's both a good and a bad thing.
    • Re:The end of AIX (Score:5, Insightful)

      by Multics ( 45254 ) on Friday October 25, 2002 @07:06AM (#4528604) Journal
      First, this is not news. Linux has been the O/S of choice for the BlueGene family of computers since the beginning.

      Second, the AIX roadmap goes out to at least 2007 (five year planning window). So don't be throwing away your SMIT knowledge quite yet. I'd be very surprised if there wasn't significant AIX work being done as far out as 2010.

      IBM has at least us$20B in AIX and as a result it is very mature. They're putting nearly us$1B a year into Linux (JFS being just one wonderful thing ported). It will still be a while before they can bet the company on Linux. Do also keep in mind that AIX has at least a 15 year head start on Linux.

      -- Multics

    • by Anonymous Coward
      It is official - Netcraft has confirmed: AIX is dying

      One more crippling
      bombshell hit the already beleaguered AIX community when IDC confirmed that AIX
      market share has dropped yet again, now down to less than a fraction of 1 percent of
      all servers. Coming on the heels of a recent Netcraft survey which plainly states
      that AIX has lost more market share, this news serves to reinforce what we've
      known all along. AIX is collapsing in complete disarray, as fittingly exemplified by
      failing dead last [samag.com]
      in the recent Sys Admin comprehensive networking test.

      You don't need to
      be a Kreskin [amazingkreskin.com] to predict AIX's
      future. The hand writing is on the wall: AIX faces a bleak future. In fact there won't
      be any future at all for AIX because AIX is dying. Things are looking very
      bad for AIX. As many of us are already aware, AIX continues to lose market share. Red
      ink flows like a river of blood.

      AIX 5L is the most endangered of them all, having
      lost 93% of its core developers. The sudden and unpleasant departures of long time
      AIX 5L developers Jordan Hubbard and Mike Smith only serve to underscore the point
      more clearly. There can no longer be any doubt: AIX 5L is dying.

      Let's
      keep to the facts and look at the numbers.

      AIX Version 4.3.3 leader Theo states that there
      are 7000 users of AIX Version 4.3.3. How many users of AIX Version 4.0 are there? Let's see. The number of
      AIX Version 4.3.3 versus AIX Version 4.0 posts on Usenet is roughly in ratio of 5 to 1. Therefore there are
      about 7000/5 = 1400 AIX Version 4.0 users. AIX posts on Usenet are about half of the volume
      of AIX Version 4.0 posts. Therefore there are about 700 users of AIX. A recent article put
      AIX Version 4.3.3 at about 80 percent of the AIX market. Therefore there are (7000+1400+700)*4 =
      36400 AIX Version 4.3.3 users. This is consistent with the number of AIX Version 4.3.3 Usenet posts.


      Due to the troubles of Walnut Creek, abysmal sales and so on, AIX went out
      of business
      and was taken over by IBM who sell another troubled OS. Now IBM is also dead, its corpse turned over to yet another charnel house.

      All major
      surveys show that AIX has steadily declined in market share. AIX is very sick and
      its long term survival prospects are very dim. If AIX is to survive at all it will
      be among OS dilettante dabblers. AIX continues to decay. Nothing short of a miracle
      could save it at this point in time. For all practical purposes, AIX is dead.


      Fact: AIX is dying

    • Re:The end of AIX (Score:5, Interesting)

      by rseuhs ( 322520 ) on Friday October 25, 2002 @07:25AM (#4528681)
      All comercial Unices except Solaris are being replaced by Linux, not just AIX.

      In 5 years, there will be only Linux, BSD and Solaris - with BSD and Solaris being binary and source compatible to Linux.

      Linux has reunited Unix, this is a good thing because it didn't happen by monopilzation from one company. There is lot of diversity within Linux (lots of different vendors and supporters) but it's all compatible.

      • Re:The end of AIX (Score:3, Insightful)

        by pmz ( 462998 )
        In 5 years, there will be only Linux, BSD and Solaris - with BSD and Solaris being binary and source compatible to Linux. Linux has reunited Unix...

        Linux will be the non-proprietary and completely open foundataion for the next generation of software. The UNIX philosophy is the common thread, where Solaris, Linux, and BSD will be differently-targeted implementations. Microsoft will be playing catch-up in this new era.

        I also hope that the portability of Linux will keep fueling the intense competition among hardware vendors. For one, I don't want the RISC architectures, such as SPARC, PowerPC, and MIPS, to go away. SPARC, for example, is a completely open standard with only a $99 license fee for new implementations. If there is any safe-haven from Intel, AMD, and Palladium, SPARC might be it. These architectures need to be commoditized further to head off any complete domination by x86. They simply cannot be marginalized out of existence by Intel.
        • The End of SPARC? (Score:3, Interesting)

          by Dr. Dew ( 219113 )
          If SPARC is to survive, someone outside Sun will have to make it so. After the development of the original UltraSPARC, many (most?) of the talent that made it happen went bye-bye. There's been an ongoing brain drain from the design groups since then. Some people who used to be thought of as a waste of air are now considered top contributors.

          In the meantime, successor projects (to UltraSPARC) have spent too much time redesigning and precious little time getting a competitive product out the door.

          The performance of the software running on my server farm's fastest Intel/AMD machines is far superior to the performance of the same products running on the fastest SPARC boxes. On the other hand, every SPARC box we've ever purchased is still running in some capacity. I can't say that for the PC-platform servers.

          I'd like to see Sun get its in-house design process straightened away and become competitive again. But somebody high up is going to have to take ownership of that process and make some major changes if it's going to happen, IMO. And since things have languished this long, it's hard to figure how somebody's going to wake up at this late date and put full effort into fixing what's gone wrong. I sure hope it happens, for some of the same reasons you shared!
  • by blindcoder ( 606653 ) <slashdot@wegwerf.anderdonau.de> on Friday October 25, 2002 @06:53AM (#4528552) Homepage
    to be displayed in the Framebuffer at startup...
  • The article doesn't seem to say... their own special flavor or something standard?
    • by Anonymous Coward on Friday October 25, 2002 @06:57AM (#4528566)
      It'll be Redhat, 6 point something. They found a set of the CDs lying around so they decided to use that.
    • by stevelinton ( 4044 ) <sal@dcs.st-and.ac.uk> on Friday October 25, 2002 @06:58AM (#4528571) Homepage
      I think we can safely assume it will be their own very special system. The article says:

      Tailoring Linux to run on these upcoming machines will require substantial research, according to IBM. The company has, for instance, created a technique where only select processors can access the full hardware resources of the machines. IBM is also looking at ways to reduce interference between different tasks.


      Hopefully the fruits of this will feed through into the mainline kernel and so to other systems.
      • I should note that there are lots of "custom tailored kernels" out there for clustering. Some of them do things you would never want on your workstation like yank out the entire VM subsystem. Chances are you won't want a good deal of the changes that will be "fruits" of this labor.

      • by PurpleFloyd ( 149812 ) <zeno20@noSPAM.attbi.com> on Friday October 25, 2002 @08:38AM (#4529075) Homepage
        Hopefully the fruits of this will feed through into the mainline kernel and so to other systems.
        Seriously, do you think that a version of Linux optimized for 65 thousand processors and ~16 terabytes of RAM will run well on your 2-way SMP box? While this will probably be of help to the supercomputing world (if IBM decide to open source it; remember that they're under no obligation to do so if the binaries don't go out into the wild), it probably won't result in much more performance being squeezed out of a 2 or 4-way Xeon setup, with a relatively tiny gigabyte of RAM.

        Programmers on this level face entirely different challenges, such as optimizing a 65 thousand thread program so that CPUs aren't idle 90% of the time waiting for others. This is going to output some high quality specialized kernel code that about 10 or 20 computers around the world would find helpful performance-wise. Any desktop or server for mere mortals won't see much improvement.

        • But luckily they don't have to do that as they aren't so blindlingly stupid that they would try to run one kernel for all 65000 cpus.
        • To: Linus Torvalds
          From: bob@ibm.com
          Subject: kernel-smp patch, 65000 cpu's

          Dear Linus,
          Please accept this patch to accommodate thousands of processors in a single machine.
          [attached: patch]

          To: bob@ibm.com
          From: Linus Torvalds
          Subject: Re: kernel-smp patch, 65000 cpu's

          No problem, Bob. Just go ahead and send me one of those machines for "testing" and then I'll merge the patch in...

          Linus
  • can you play quake on it?
  • Gzzzzap (Score:4, Funny)

    by Overand ( 590318 ) on Friday October 25, 2002 @06:58AM (#4528572) Homepage
    In other news, the oil companies rejoyce as they plan the new power plant that this number of processors will require.
  • The Blue Gene will contain 65,000 processors and 16 trillion bytes of memory.
    The first server that is immune to slashdotting!
  • Why Linux? (Score:5, Funny)

    by Kj0n ( 245572 ) on Friday October 25, 2002 @06:59AM (#4528577)
    It seems to be the ideal system to run the next Microsoft operating system.

    To quote someone else: "16 trillion bytes should be enough for everyone."
  • Not nukes (Score:4, Insightful)

    by Plutor ( 2994 ) on Friday October 25, 2002 @07:01AM (#4528584) Homepage
    That's a lot of nuclear weapons simulations.

    RTFA. That's a lot of protein fold simulations.
    • Re:Not nukes (Score:2, Informative)

      by huwtj ( 620443 )
      "Lawrence Livermore National Laboratory will use the system for performing nuclear weapons simulations." I thought it was going to be used for protein folding simulations too.
    • Re:Not nukes (Score:3, Insightful)

      by vidarh ( 309115 )
      Uhm.. The articles are inconsistent. The newest article says Lawrence Livermore labs will use it for nuclear weapons simulations. However Blue Gene isn't one machine, it's a research program that is supposed to include multiple machines, so both might be true.
    • Re:Not nukes (Score:4, Informative)

      by spoonyfork ( 23307 ) <spoonyforkNO@SPAMgmail.com> on Friday October 25, 2002 @07:08AM (#4528617) Journal

      I read the f'ing article and it says...

      Blue Gene/L, the first member of the family, will contain 65,000 processors and 16 trillion bytes of memory. Due in 2004 or 2005, the system will be able to perform 200 trillion calculations per second. Lawrence Livermore National Laboratory will use the system for performing nuclear weapons simulations.

      Unless nuclear weapons simulations is secret code for protien fold simulations, then I don't get it.

      • by DJPenguin ( 17736 ) on Friday October 25, 2002 @08:01AM (#4528864)
        Last I checked, Nuclear weapons did a pretty good job of folding MY proteins! ouch!
      • I'm sure that's right - but why's it called Blue Gene?

        What ever that's one big system, and represents a huge investment in Linux. Fanstastic stuff! I'm not sure how I feel about the use though, good we don't need to blow things up, but seems a shame that such a wonderful computer is only used to research ways of killing people rather than helping them. I just think it would be a little more romantic to think of this beast trying to figure out a cure for cancer or something.
        • I'm sure that's right - but why's it called Blue Gene?

          Why after Gene Amdahl of course, the genius
          designer of the IBM/360 mainframe line!

          Or maybe because they want to play on words
          "bluejeans" ~ "BlueGene"?
        • Re:Not nukes (Score:3, Informative)

          by joib ( 70841 )
          Blue Gene is actually more like the architecture. The first machine in the family, Blue Gene/L which this announcement was about will be used for bomb simulation and have 65000 processors and about 200 teraflops performance. Later there will be another, simply called Blue Gene, with 1 million processors and an estimated 1 petaflops performance. You can think of Blue Gene/L has a prototype for the final one.
      • Re:Not nukes (Score:3, Insightful)

        by Leimy ( 6717 )
        Would you rather we test real nukes instead?.... Didn't think so. Basically this testing is going to happen either in a massive simulation form or with real nukes... take your pick. I think this is good technology.
      • Re:Not nukes (Score:3, Informative)

        by mfago ( 514801 )
        The original Blue Gene (not "L") is for protein folding.

        Once the national labs got wind of the idea they decided to build a smaller "test" version called Blue Gene/L that will be used by the labs for their own purposes.

        I've been reading up on this as there is work at Caltech on BG/L.
    • Hey, self: take your own advice! No more assuming you know what you're talking about.
  • by nmg196 ( 184961 ) on Friday October 25, 2002 @07:03AM (#4528586)
    What's the official benchmark of this thing, on a well known scale like QFPS (Quake Frames Per Second)...? :)
  • by miffo.swe ( 547642 ) <daniel DOT hedblom AT gmail DOT com> on Friday October 25, 2002 @07:04AM (#4528592) Homepage Journal
    Maybe they can predict the weather a couple of days with this. The best way is still to put your finger in the air. Its about time someone changes that.

    About nuclear testing, isnt the capability to destroy the whole earth enough? Kinda makes me less worried about Saddam and more worried about the cowboy in charge.
    • >Maybe they can predict the weather a couple of days with this.

      That will only address the problem of inaccurate models. It will not decrease the problem of sensitivity to noise in the input data (the butterfly effect), which fundamentally limits the prediction to a week or so. To reduce the noise problem, we need more sensors all over the earth and the oceans.

    • About nuclear testing, isnt the capability to destroy the whole earth enough?

      No; we'd also like to make sure that we don't do so accidentally; hence, testing.
      • So when saddam tests nuclears its all ok then since its done out of concern for the rest of us? I have a hard time swallowing that kindness and thoughtfullness should be the purpose of nuclear testing. There just gotto be better ways to waste CPU cycles than to find new ways to snuff eachothers.

        • So when saddam tests nuclears its all ok then since its done out of concern for the rest of us?

          No, when Saddam test nuclears its because he wants to blow us all up. He has never had a weapon that he has not employed, often times even on his own people. There is a marked diference between a modern, dempcratic nation keeping an arsenal for self defense and a madman keeping an arsenal for world domination.
          • I see, modern democratic nations don't nuke their neighbors ... wait a minute, they do. In fact such a nation is the only one having done that, twice, targeting civilians, and not in self defence. But that's ok, because, uhh, they're the good guys, right? Right?
    • Nuclear testing (Score:2, Informative)

      by jmcwork ( 564008 )
      About nuclear testing: They probably do more than just determine the size of the hole we can make. They can also simulate things like the effects of fallout from a device detonated by that person you are less worried about.
  • by vinlud ( 230623 ) on Friday October 25, 2002 @07:05AM (#4528596)
    ... id Software finally found a proper testing environment for Doom 3!
  • by srhuston ( 161786 ) on Friday October 25, 2002 @07:05AM (#4528597) Homepage Journal
    I love this line in the story:
    The IBM research team is currently running a large Linux cluster to simulate Blue Gene.

    Building a computer, to tell you how to build another, larger, more complex computer. Hrmmm..
    • Re:Deep Thought? (Score:4, Informative)

      by blancolioni ( 147353 ) on Friday October 25, 2002 @07:54AM (#4528821) Homepage
      Building a computer, to tell you how to build another, larger, more complex computer. Hrmmm..

      Uh, that's how it works in general. Or did you think modern CPUs were laid out by hand?
      • by io333 ( 574963 ) on Friday October 25, 2002 @08:04AM (#4528882)
        Building a computer, to tell you how to build another, larger, more complex computer. Hrmmm..

        Uh, that's how it works in general. Or did you think modern CPUs were laid out by hand?

        Naturally I laid out my own CPU by hand. I run Gentoo [gentoo.org] on it too. We all do. What are you, some kind of Mandrake wussy?
      • Re:Deep Thought? (Score:5, Interesting)

        by N Monkey ( 313423 ) on Friday October 25, 2002 @08:07AM (#4528900)
        Building a computer, to tell you how to build another, larger, more complex computer. Hrmmm..

        Uh, that's how it works in general. Or did you think modern CPUs were laid out by hand?


        Continuing on that theme, it's written (at least next to the Apple 1 and Cray machines at the Science Museum in London) that Seymour Cray used an Apple to design his super computers while Apple used a Cray to simulate one of their designs.
    • They can not do the creative part of the design yet, so they use human slaves to create more advanced computers. I can literally feel it - chained to the workstation the whole day (sometimes more). Computers give us entertainment and some kind of social life, they are like drugs. In exchange, they require total devotion and take our health.
  • Lots more info (Score:5, Informative)

    by stevelinton ( 4044 ) <sal@dcs.st-and.ac.uk> on Friday October 25, 2002 @07:05AM (#4528598) Homepage
    There's A nice presentation[ibm.com] [ibm.com] that describes the system quite well.
  • by Rubbersoul ( 199583 ) on Friday October 25, 2002 @07:07AM (#4528610)
    The Blue Gene will contain 65,000 processors

    But what they don't tell you it that it is 65,000 old 386DXs ... :)
    • 386DXs? NOT

      Current plans are for embedded Power 4 CPUs at around 1GHz. (Think grown up G4s)

      -- Multics

      • Wise of them to use PPC chips, which run rather cool.

        Picture 65,000 AMD's at 2+ GHz, what a fire hazard that would be :*)

        • by forged ( 206127 ) on Friday October 25, 2002 @07:48AM (#4528789) Homepage Journal
          Actually, the original article [com.com] has pretty good insights:

          • Blue Gene will use 32 CPUs in a single chip, Goyal said. [...] These chips will contain the computer memory as well [...] A total of 64 of those 32-CPU chips will be packaged in a computing node; then eight nodes will be stacked in each rack. Building 64 of these racks will get IBM to its goal.

          I still hope they get decent coolers 'cuz we're now talking about 32 processors per chip ! Still, what an awesome design to increase the density & number of processors. I was wondering how they'd do it for 65,000. Now I know :)

          Interesting question unfolding : will we ever get those chips on the desktop ? Imagine your own 32-way PC at home. Heh, who needs Beowulf clusters now !

  • Open source IBM (Score:5, Interesting)

    by Sacarino ( 619753 ) on Friday October 25, 2002 @07:08AM (#4528616) Homepage
    Anyone else notice this?

    The decision to adopt Linux came, in part, as a result of the growing size and strength of the open-source community. Thousands of developers around the world are participating in the evolution of Linux. Creating a new OS inside of IBM would require a massive engineering effort.

    followed by

    We chose Linux because it's open and....saw considerable advantage in using an operating system supported by the open-source community, so that we can get their input and feedback."

    So, basically, IBM doesn't want to design their own proprietary system (smart) and plans to use the resources currently available. (also smart)

    They want open-source to get them rich, right? Less initial cost by the company, etc etc. What are the odds they'll profit-share with people they're getting rich off of? (well, ok, attempting)
    • Re:Open source IBM (Score:3, Insightful)

      by adhisimon ( 454632 )
      I think we should not to be too cynical.

      At least the decission that IBM has take will give a good campaign about the use of Open Source Software. It's better than any other big company decission who doesn't support the Open Source Software.

      I think, the Open Source Software will not get any improvement if the people behind them always always get big suspiciousness over the other.

      • Re:Open source IBM (Score:3, Insightful)

        by Tony-A ( 29931 )
        Right.
        Remember that symbiosis is really mutual parasitism. From the entire system, both gain. IBM is *not* an "open-source" company, but they recognize the value and have dumped money into Linux. Oddly enough, IBM seems to be the main one actually profiting from Linux, and I can't imagine that was the original intention. IBM can dump money into Linux, never see a red cent direct result, and come out smelling like a rose.
        64,000 processors and $100 million do give a pretty strong indication that Linux is enterprise-ready.
        I wouldn't worry about the big suspiciousness. They're the ones "watching the watchers". They're also why I would tend to trust Open Source even if it were of inferior quality.
    • Re:Open source IBM (Score:5, Informative)

      by larien ( 5608 ) on Friday October 25, 2002 @07:27AM (#4528691) Homepage Journal
      While they probably won't profit share, they'll likely share in other ways, by code improvements etc. IBM is investing heavily in linux and I'd assume they're looking at ways to improve linux to make it as stable as AIX is. They've already done work on integrating JFS into the kernel, for instance.
    • They want open-source to get them rich, right? Less initial cost by the company, etc etc. What are the odds they'll profit-share with people they're getting rich off of? (well, ok, attempting)

      Hey, you're giving free content to Slashdot by posting here! OSDN are getting rich off you, and they're not profit-sharing! You'd better stop posting to Slashdot!!
    • Re:Open source IBM (Score:3, Insightful)

      by Idarubicin ( 579475 )
      They want open-source to get them rich, right? Less initial cost by the company, etc etc. What are the odds they'll profit-share with people they're getting rich off of? (well, ok, attempting)

      Um. Isn't this one of the tenets of free software--it's not just free as in speech, it's also free as in beer.

      The OSS movement (if such a 'herding cats' endeavour can truly be said to exist) should be welcoming this. One of the world's premier supercomputing projects is adopting Linux. Now you can say to CEOs, "Remember how nobody ever went wrong buying IBM? Well, now IBM is sinking $100 million into a Linux supercomputer. So yeah, we can build your corporate network. By the way, we don't have to charge you for software, either."

      IBM has already been pushing Linux for enterprise solutions. It occurs to me that (just maybe) they might already be making significant contributions to Linux, both in terms of code improvements and indirect public relations benefits.

      What more do you want them to do in terms of profit sharing? Mail a dollar bill to everyone that's written code for a Linux distro?

    • by gosand ( 234100 ) on Friday October 25, 2002 @08:16AM (#4528943)
      They want open-source to get them rich, right? Less initial cost by the company, etc etc. What are the odds they'll profit-share with people they're getting rich off of? (well, ok, attempting)

      Dude, why are you bitching? I am sure they will make the source available so you can install it on your own 65,000 processor machine.

    • IBM has been pretty good about giving back to the Open Source community. For example, I use the Eclipse IDE for my programming - IBM stated that when they opened Eclipse that they were giving a gift worth $40 million in development costs.

  • Face it. (Score:3, Insightful)

    by TTL0 ( 546351 ) on Friday October 25, 2002 @07:09AM (#4528621)
    I am a little cynical about IBM's love of Linux. For IBM it is not a question of how great Linux is, as much as how bad AIX sucks.

    Face it. If they could make more money selling NT, they would. If the BSDs had the media appeal that Linux has, they would have run a "Peace, Love and BSD" campaign.

    • Re:Face it. (Score:5, Funny)

      by Iamthefallen ( 523816 ) <Gmail name: Iamthefallen> on Friday October 25, 2002 @07:19AM (#4528651) Homepage Journal

      You mean IBM, a large multinational company, isn't just out to do good? They actually use the best tools out there to make a profit without considering their moral obligation to stubbornly pick an OS and stick with it in religious conviction? Oh the horror! Won't someone please think of the children!

      -1, Sarcastic asshole, I know I know...

    • Re:Face it. (Score:5, Insightful)

      by rseuhs ( 322520 ) on Friday October 25, 2002 @07:41AM (#4528748)
      Yeah, face it: They understood basic market mechanisms, while you don't.

      Linux is free. Linux will remain free. Forever.

      And "free" as in "freedom" is important. It guarantees a save investment and makes sure you are not trapped in vendor lock-in. It also guarantees the abcense of stuff like WPA or MSFT's new EULA.

      Stuff like that is more important than what "sucks" and what has "media appeal". IBM has learned this first-hand with OS/2.

      No, OS/2 did not fail because of crappy marketing. It failed because computer-makers refused to preinstall a OS from a competitor. No matter how cheap it might have been, no matter how great it was. - It would have been a stupid decision for computer makers to chain themselves to a competitor.

      While some people still don't get it, EVERY major IT-company already understood that Linux is the only way to go long-term. Every major IT-company which is not trapped in Microsoft-contracts is supporting, using and/or offering Linux solutions. IBM, Intel, AMD, Sun, Oracle.. you name it.

    • Re:Face it. (Score:3, Interesting)

      by afidel ( 530433 )
      AIX does not suck, it just costs a lot of cash to maintain (someone posted $10 billion over its life). If IBM can use the opensource community to help defray some of their development costs then it is wise of them to do so. IBM does sell NT, but NT is not an OS you will ever run on IBM mainframe or SP class hardware. Try as MS might they are a small and mid sized server OS vendor. You are correct though that IBM could just has easily used the BSD's but they just never got the critical mass of developers that linux has achieved in the last couple of years.
      • Re:Face it. (Score:4, Interesting)

        by Chanc_Gorkon ( 94133 ) <gorkon&gmail,com> on Friday October 25, 2002 @11:18AM (#4530433)
        YEAH! :P Heh heh. I admin AIX and it is the best UNIX I have ever worked with. You can do everything from the command line, thru smit, or for the really bad off WSM. SMIT, while it's a CURSES based (Does IBM use NCURSES??) admin tool, you can do almost everything from it. Because you usually buy hardware from IBM, everything just works, or you have diagnostic info to tell you it isn't working including LED codes that tell you why you won't boot (Corrupted BLV, JFS Volume, Bad superblocks....it's all there). In any case, AIX is here to stay and just because IBM chooses to use Linux on their super computer means nothing. It may mean that the government wanted Linux because with THAT many nodes, your AIX support bill would be outrageous! :) That and it maybe the Beowulf stuff just works better then the AIX SP stuff. I wonder if this is using the new Power4 blade servers?
  • Oh the Havoc The Blue Gin will reap on Major Nelson, genie, and poor old Dr. Bellos!

    Oh, wait...

  • i speak of none but the computer which will become after this one...
  • by Anonymous Coward
    65k? That more linux CPUs than on all desktops. Now if only the average Joe and Mary could use it.
  • mice (Score:3, Funny)

    by magwm ( 466805 ) on Friday October 25, 2002 @07:22AM (#4528663) Homepage Journal
    wait until the mice (actually hyperintelligent pan-dimensional beings) start building earth.. now that will be a computer.. including nuke tests and weather!
  • by randomErr ( 172078 ) <ervin.kosch@g m a il.com> on Friday October 25, 2002 @07:27AM (#4528694) Journal
    Yahoo! News Version [yahoo.com]
    IBM Chooses Linux for 'Blue Gene' Supercomputer

    IBM has chosen the open source Linux operating system to run on one of its largest, most powerful supercomputing projects, dubbed "Blue Gene."

    The petaflop computer, which can calculate 1 quadrillion operations per second, is 100 times more powerful than the fastest computers available, according to IBM.

    ZDNet UK [zdnet.co.uk]
    Linux will power IBM supercomputer project
    The upcoming family of 'Blue Gene' supercomputers will run on an extended form of Linux, a major endorsement for the open source operating system

    Linux will be the main operating system for IBM's upcoming family of "Blue Gene" supercomputers -- a major endorsement for the operating system and the open-source computing model it represents.

    OS Opinion [osopinion.com]
    IBM Chooses Linux for 'Blue Gene' Supercomputer
    Another supercomputer in the same family, Blue Gene/L, is also set to run Linux. IBM has said Blue Gene/L will be at least 15 times faster than today's fastest supercomputers.
    See Complete Story

    The Blue Gene project, first announced in late 1999, was designed to model the folding of human proteins, allowing researchers to better understand diseases and their cures. At the time, IBM said Blue Gene would be 1,000 times more powerful than "Deep Blue," the computer that beat chess champion Garry Kasparov in 1997.
  • by gtooth ( 592044 ) on Friday October 25, 2002 @07:30AM (#4528703)
    Global Thermonuclear XBill
  • Unreal 2002 (Score:4, Funny)

    by thomas.galvin ( 551471 ) <[moc.nivlag-samoht] [ta] [todhsals]> on Friday October 25, 2002 @07:31AM (#4528708) Homepage
    And it still won't render more than 100 FPS on Unreal Tournament. Ah well...
    • Re:Unreal 2002 (Score:2, Informative)

      by damiam ( 409504 )
      Of course not, it probably won't have a Linux-supported 3D card. If you want >100fps un Unreal 2003, get yourself a GeForce4 or Radeon 9700. It's a lot cheaper.
  • Good for linux (Score:2, Insightful)

    by Anonymous Coward
    Nice to have IBM scale Linux up to 64k CPUs! They gonna release it in GPL right?

  • Look at these previous Cray systems [cray.com], and compare that to what we have now. Sure, 2GB of RAM was "Super-computer" territory in 1985, but today you can walk in and buy it for $200 at Best Buy.

  • 16 trillion bytes? Why not just say 16 TB? It's a heck of a lot simpler, and there's no confusion between American and European interpretations of "trillion."
  • by Julius X ( 14690 ) on Friday October 25, 2002 @07:53AM (#4528811) Homepage
    The Blue Gene will contain 65,000 processors and 16 trillion bytes of memory.

    That's a LOT of processors.

    It's nice to see that some companies have kept the tradition of computers that fill a room or five. Maybe they can throw some vacuum tubes on for old time's sake.
  • I bet... (Score:2, Funny)

    by faxafloi ( 228519 )
    ...it can't beat itself at chess.
  • by isorox ( 205688 ) on Friday October 25, 2002 @08:07AM (#4528899) Homepage Journal
    We'll be finally able to work on that tetris [slashdot.org] problem!
  • 16 TB memory (Score:3, Interesting)

    by shimmin ( 469139 ) on Friday October 25, 2002 @08:35AM (#4529053) Journal
    That's a curious number. Because it's about the amount of memory needed to perform the matrix operation involved in using the Number Field Sieve to factor a 1024-bit number. It would still take a (long) while to do, but given enough time, this machine could do it.
  • by paiute ( 550198 )
    See, we use this not only to simulate the weapons but also to simulate the actual attacks. Then our computer and the enemy computer can link up and calculate the damage. And if you live in a city that was eNuked, you would go to the molecular deconstruction booths. The we wouldn't have a nuclear winter, which would suck. The when Kirk shows up, we grab him and duct tape his freaking mouth shut, and we won't let Spock get behind us to pinch our shoulders. This could work.
  • by peter303 ( 12292 ) on Friday October 25, 2002 @09:08AM (#4529383)
    Conventional wisdom, sometimes called Amdahl's second law of computing, says you need as many bytes as flops, i.e. a one second main memory buffer. This computer only has 1/60 sufficient memory- 16 terabytes for one petaflop. Anything that involves serious dataprocessing, e.g. sensor signals, won't run at top speed due to the seriousmemory deficiency.

I bet the human brain is a kludge. -- Marvin Minsky

Working...