Slashdot Mirror


How to get 1.5 TeraFlops from Linux

Oak Ridge National Lab has purchased from SGI an Altix 3000 (flash movie). This article claims that: SGI Altix 3000 is recognized as the first Linux cluster that scales up to 64 processors within each node and the first cluster ever to allow global shared memory access across nodes. There is more here, here, and here.

280 comments

  1. Imagine. by inertia187 · · Score: 0, Insightful

    Imagine a beowulf cluster of those!

    --
    A programmer is a machine for converting coffee into code.
    1. Re:Imagine. by notque · · Score: 0, Offtopic

      Imagine. (Score:-1, Insightful)

      Never seen that before.

      --
      http://use.perl.org
    2. Re:Imagine. by Anonymous Coward · · Score: 0

      Known bug. It's because by definition, you cannot have:

      Imagine. (Score: -1, Overrated)

  2. Look Out! by TWX · · Score: 4, Funny

    "SGI Altix 3000 is recognized as the first Linux cluster that scales up to 64 processors"

    SCO will be all over your ass now!

    --
    Do not look into laser with remaining eye.
    1. Re:Look Out! by Anonymous Coward · · Score: 0, Offtopic

      Talk about beating a dead horse, SCO blasts are sooooo lame now. Please troll elsewhere.

    2. Re:Look Out! by monkey_jam · · Score: 4, Funny

      ..maybe you should learn to wipe better...

  3. Beowulf cluster jokes... by Kiriwas · · Score: 5, Interesting

    After all the beowulf cluster jokes, I am still incredibly curious about them. My goal is to build a small 5-6 node cluster by the end of the summer. The thing is, I still know very little about them. Every jokes about them, but no one puts any useful information. Are there specific langauges one must program in to tak advantage of the multiple processors? Or does the OS take care of that? How much speed can you actually get out of them? Is it pure processing power? Or is there more? I'm very curious and want to know.

    1. Re:Beowulf cluster jokes... by SkArcher · · Score: 2, Informative

      here and here are probably good places to look.

      --

      An infinite number of monkeys will eventually come up with the complete works of /.
    2. Re:Beowulf cluster jokes... by donutz · · Score: 1

      The thing is, I still know very little about them. Every jokes about them, but no one puts any useful information.

      Well maybe you're looking in the wrong place.

      Good luck.

    3. Re:Beowulf cluster jokes... by gladbach · · Score: 5, Informative

      just download clusterknoppix and knock yourself out. ; )

      http://bofh.be/clusterknoppix/

      --
      "Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd all be running around in darkened rooms,
    4. Re:Beowulf cluster jokes... by The_ForeignEye · · Score: 5, Informative

      Back in my days of parallel programming (read: 1998) on Beowulf clusters I used Fortran and C. The trick to make your program "parallel" is to use special programming libraries that will spawn instances of your program across the cluster and let them communicate between each other. The libraries I used were PVM and MPI.

      At that time they were working on a Java implementation, but I don't know what happened with that.

    5. Re:Beowulf cluster jokes... by 3141 · · Score: 2, Informative

      Another poster mentioned MOSIX, but openMosix is probably a better bet. It's released under the GPL, and is a combination of kernel-patch and user-space tools. Once you get these installed on each node, and connected via ethernet (all with networking set up of course... IP addresses etc) you should have yourself a cluster.

    6. Re:Beowulf cluster jokes... by Jeremy+Erwin · · Score: 1

      You might be able to find this book in the remainder bins. Alas, the publisher has recalled it, for some reason.

    7. Re:Beowulf cluster jokes... by RussianBeard · · Score: 2, Informative

      Take a look at OSCAR. We built a nine node cluster out of IBM e-servers using it. It was really quite straightforward.

      As far as languages go, you'll need an MPI library (like MPICH, or LAM/MPI (which is also a runtime environment), but the actual code used is usually C, C++, or Fortran. BTW, OSCAR comes with MPICH and LAM/MPI.

    8. Re:Beowulf cluster jokes... by fitten · · Score: 1

      Hmm... the Java interface should be there somewhere. I forgot where we put ours. But... depending on how coarse grained your app is, just using normal Java communication may be good enough. Only in the fine grained communication schemes would you need MPI.

    9. Re:Beowulf cluster jokes... by Anonymous Coward · · Score: 0

      I use google. What is this asshat you refer to? www.asshat.com appears to be some sort of "indie" music shite.

    10. Re:Beowulf cluster jokes... by wavedeform · · Score: 1
      Try using google and reading asshat.

      OK, I took your suggestion and went to Google to look up asshat, but I didn't find any information on Beowulf clusters. I did find out more that I ever knew possible about the word asshat, though.

    11. Re:Beowulf cluster jokes... by Helmut+Kool · · Score: 2, Funny

      You don't seem to remember the joke well enough. You're supposed to IMAGINE the beowulf cluster, not actually build it.

    12. Re:Beowulf cluster jokes... by oudzeeman · · Score: 2, Informative

      This SGI isn't a beowulf cluster. Traditionally beowulf clusters refer to clusters that use COTS hardware, don't have global shared memory, etc. Lots of people in the cluster community won't even call clusters of workstations beowulf clusters if they have some high speed network like Myrinet. We just call ours a Linux cluster, a cluster, a distributed memory supercomputer... You can program your beowulf cluster in C or Fortran using a free MPI(message passing interface) implementation called MPICH. I have even seen a scaled down version of MPI for Python, (which requires MPICH to use). So start learning MPI. MPI-1 has 129 functions, but you can write most programs using a small subset of these calls. If you don't want to pay much money I suggest using C, because g77 sucks and there are no free Fortran 90 compilers. We use the Portland Group Fortran and C compilers as well as the Intel Fortran Compiler. I think we are going to switch completely to Intel Fortran and C. Why do you want to use a beowulf cluster if you have no clue about them or parallel programming in general? Just because they are 'cool'? A beowulf cluster is very usefull for modeling or datamining, but unless you are running models that take days/weeks/months on your workstation you won't need the processing power of a cluster. Right now we have someone running a model on 76 processors that takes about 9 hours to finish a 1 year cycle in the model. They want to run the model for a total of 50 years. This is a model of the pacific ocean where they introduce carbon into the ocean, and then they see what effect that has on temperature change, etc. After they get their 50 year resluts for the Pacific they want to do a global simulation. This is the real use of beowulf clusters. They aren't for load ballancing web servers, playing quake, or any of the other things people post about every time there is an article about supercomputers/beowulf clusters. The speed up you will get really depends on your application. The more communication is necessary, the smaller the speed up will be. If you have a 5 node cluster, with 2 processors per node, the theoretical maximum speed-up is 10, but you will never achieve that because of parallel overhead(MPI calls, communication time, etc). If you want more information on parallel programming and cluster computing send me a private message telling me what you hope to do with your cluster.

    13. Re:Beowulf cluster jokes... by Anonymous Coward · · Score: 0

      Its called JPVM

      http://www.google.com/search?q=jpvm&sourceid=moz il la-search&start=0&start=0&ie=utf-8&oe=utf- 8

    14. Re:Beowulf cluster jokes... by sql*kitten · · Score: 0
      The trick to make your program "parallel" is to use special programming libraries that will spawn instances of your program across the cluster and let them communicate between each other.

      If you're compiler is smart, it can parallelize ordinary code too. Just as a trivial example, consider:
      for (i = 0; i < 1000000; i++) {
      doSomeCalculations(i);
      }
      A smart compiler would notice that you had 2 CPUs and would break that into two concurrent loops at compile time:
      for (i = 0; i < 500000; i++) {
      doSomeCalculations(i);
      }
      for (i = 500000; i < 1000000; i++) {
      doSomeCalculations(i);
      }
      And at runtime, the OS would run it on 2 processors. The really smart compilers like SGI's place wrapper code around it, and check at runtime to see how many processors are available, then break the loop into parallel threads on the fly.
    15. Re:Beowulf cluster jokes... by battjt · · Score: 2, Interesting

      Bull.

      int a;
      void doSomeCalculations(int i) {
      a = doSomethingElse(a + i);
      }

      Would fail (multiple threaded access to a). It is extremely difficult to detect sideffects in C. I've never seen a "smart" compiler as you put it, though there are systems where the programmers can explicitely parallelize a loop.

      --
      Joe Batt Solid Design
    16. Re:Beowulf cluster jokes... by sql*kitten · · Score: 0, Troll

      I've never seen a "smart" compiler as you put it, though there are systems where the programmers can explicitely parallelize a loop.

      Oh, because you've never seen it, it can't exist? Great logic there, I hate to think what your code is like. See here and here and here.

    17. Re:Beowulf cluster jokes... by appleLaserWriter · · Score: 1

      you will need monkeys.

      LOTS of monkeys.

    18. Re:Beowulf cluster jokes... by battjt · · Score: 1
      All three of your references cite the SGI optimizer, which has the limitation that I pointed out. From the MIPSpro(TM) Auto-Parallelizing Option Programmer's Guide:
      By default, the Auto-Parallelizing Option does not parallelize a loop that contains a function call because the function in one iteration of the loop may modify or depend on data in other iterations.
      Furthermore my mention that C is very hard to parallelize is also mentioned in the same chapter.
      The C and C++ languages have features that make them more difficult than Fortran to automatically parallelize. These features are often related to the use of pointers, such as implementing multidimensional arrays as pointer accesses and not as true arrays. The following practices involving pointers interfere with the Auto-Parallelizing Option's effectiveness:
      • arbitrary pointer dereferences
      • arrays of arrays
      • loops bounded by pointer comparisons
      • aliased parameter information
      --
      Joe Batt Solid Design
    19. Re:Beowulf cluster jokes... by brsmith4 · · Score: 1

      PGI Compiler does this. Its also good for optimizing floating point calculations. I run a beowulf cluster for my university and we make all of our users compile with the PGI compiler just cuz its better.

    20. Re:Beowulf cluster jokes... by Anonymous Coward · · Score: 0

      Once again you demonstrate how much of a fucktard you are. I'd personally love to see some software you write, if it's constructed with the care and intelligence your comments are, it should give me something to laugh at during my lunch break.

    21. Re:Beowulf cluster jokes... by goodchef · · Score: 1
      The OS and associated libraries will take care of the actual communication between the nodes. To this end, you can use any platform/OS and language that has support for it. However, it's still up to you, the application developer, to figure out how to parallelize your code. Some tasks lend themselves easily to this, some do not, and some can't be parallelized at all.

      At my work, we develop for a smattering of platforms, ranging from Linux, MacOS X and Cygwin to Solaris, IRIX, and a plethora of custom-built supercomputers. It's all done using C++ and MPI. MPI is a standard that specifies how nodes communicate with each other, and what methods/functions are called to do this. MPICH is an implementation put out by Arlington National Labs. LAM is another implementation, put out by Indiana University. LAM is a much nicer implementation IMO, but it's not available on quite as many platforms.

      I would recommend these two books: How to Build a Beowulf and High-Performance Computing.

      Bottom line: If you have computationally intensive calculations to do, beowulf clusters are a cheap alternative to pricey supercomputers. But if you've just heard they're cool but don't have anything to do with one, it's probably not worth it, as applications have to be specifically (re)written to take advantage of a cluster, they don't get automagically faster.

      --

      "Inflammable means flammable? What a strange country!" -Dr. Nick, The Simpsons

  4. Apple by zzzmarcus · · Score: 5, Funny

    Oh great... I can see Jobs wringing his hands already.

    "Now how am I going to make the G5's look faster than THIS?"

    1. Re:Apple by Trigun · · Score: 2, Funny

      He just has to tell the apple fanboys that a G5 is, and they'll believe it.

      For his next trick, Jobs is going to walk on water.

    2. Re:Apple by geekmetal · · Score: 1
      "Now how am I going to make the G5's look faster than THIS?"

      Am I missing something or are you comparing a G5 to a 256 processor system here?

      --
      There are two kinds of egotists: 1) Those who admit it 2) The rest of us
    3. Re:Apple by Umrick · · Score: 2, Interesting

      Rendevous will be used in 10.3 with Xcode to discover resources and distribute software builds across available 10.3 machines. If there's a perceived benefit to Apple, do you honestly think there's anything preventing the next version 10.4 from having distributed capabilities?

      You can already compile programs with LAM-MPI support, so in reality there is nada stopping you from building a Beowulf cluster of XServes. There may even be a compelling reason to use XServes over x86 boxes after XServers are updated to G5s.

      Rumor was the original XServes were built to spec for a distributed cluster for a Blast! genome search engine.

      People get hung up on Beowulf = Linux, and that isn't necessarily the case if you take Beowulf to mean a cluster of inexpensive machines.

      OS 10.2+ with Rendevouz autodiscovery using LAM-MPI for communicating could just be a killer configuration. Lord knows cluster management/monitoring would be outstanding, though perhaps the setup would not be as simple initially.

    4. Re:Apple by platypus · · Score: 1

      The point is, from the hardware side apple is lightyears away from machines like this. And then the OS must be capable of taking advantage/adapt to things like NUMA (which is not necessarily implement in this machines, but in many of this size), which I think OS X (or its BSD personality) is not able to.

    5. Re:Apple by Anonymous Coward · · Score: 0

      You're missing something. Here's your Kool-aid.

      Drink up.

    6. Re:Apple by Umrick · · Score: 1

      I went off at an angle to be sure. The machine in question has very little to do with a Beowulf cluster. It's more like a SGI Origin that happens to run Linux as the OS.

      No particular design makes sense in all situations. Hrm. Come to think about it, my knee jerk reaction would be fine comparing Linux Beowulf clusters to an Apple Beowulf cluster, but doesn't have much to do with the SGI box.

      Definately need more caffiene.

    7. Re:Apple by sql*kitten · · Score: 1

      Rendevous will be used in 10.3 with Xcode to discover resources and distribute software builds across available 10.3 machines. If there's a perceived benefit to Apple, do you honestly think there's anything preventing the next version 10.4 from having distributed capabilities?

      Back in the day, on the old black NeXT hardware running NeXTStep 3.3, there was an app called Zilla that could be used to distribute compute intensive jobs around a network of NeXT machines. They said that 100 NeXT Turbos was an even match for a Cray 1. It's funny when I look at OSX today... it's got a lot of bells and whistles, but a lot of the stuff that was in NeXTStep has actually been taken out!

    8. Re:Apple by Anonymous Coward · · Score: 0

      This Altrix things is a great workstation, much faster than that little G5. I just can't wait until SGI makes a server version. Then it would smoke!

    9. Re:Apple by Anonymous Coward · · Score: 0

      Actually, if you study the designs, the G5 is essentially identical to the SN-IA in most ways, and markedly superior in a few.

      SGI is, alas, light-years behind the curve in pure technical terms.

    10. Re:Apple by Anonymous Coward · · Score: 0

      They said that 100 NeXT Turbos was an even match for a Cray 1.

      Funny, considering they only ever made about 60 NeXTstation Turbos.

    11. Re:Apple by Anonymous Coward · · Score: 0

      Well obviously he's going to update the G5 by adding r-type stickers to all the cases, that should easily make up the speed difference and then some.

  5. yup by JimFromJersey · · Score: 0, Troll

    linux on the desktop ... it's a flop.

    --
    between the greater and lesser infinities sleep the dreams undreamt
    1. Re:yup by Sxooter · · Score: 1

      linux on the desktop ... it's a flop. You mispelled teraflop... :-)

      --

      --- It is not the things we do which we regret the most, but the things which we don't do.
    2. Re:yup by Anonymous Coward · · Score: 0

      linux on the desktop ... it's a flop.


      But with this technology, we can multiply that a trillion fold!

    3. Re:yup by Anonymous Coward · · Score: 0

      its even worse than that... its a flop and a half...

  6. kernel sources? by gladbach · · Score: 5, Interesting

    they going to release their kernel that allows them to globally share memory? or is it more of a hardware thing, than software?

    --
    "Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd all be running around in darkened rooms,
    1. Re:kernel sources? by Anonymous Coward · · Score: 0

      Its more a hardware thing.

      My suspicion is that they're using the generic NUMA stuff IBM has been working on. The sources are all available.

    2. Re:kernel sources? by Anonymous Coward · · Score: 0

      I'm sure it is a combination of hardware and software. Each vendor tends to have their own highly-specific implementation of NUMA, "industry standards" for NUMA hardware implementation have been tried, but everybody ends up doing their own proprietary version anyway. The one exception is add-on type NUMA where you use some fancy cards in PCI/PCI-X slots like myrinet and that other on DEC that I can't remember off-hand. But performance of those implementations don't compare to "native" interfaces like all the big boys (IBM, SGI, HP, Sun) use.

      So, certainly some of the generic NUMA stuff that both SGI and IBM have been working on is going to be mandatory to get reasonable performance, there is also certain to be a good chunk of memory-management code that is highly-specific to the SGI implementation.

    3. Re:kernel sources? by TheSurfer · · Score: 1

      Perhaps they use the open source Migshm patch for openMosix.

    4. Re:kernel sources? by Anonymous Coward · · Score: 2, Informative

      I'm 100% sure it's very much a hardware thing. SGI has a long history of building very large hardware shared memory machines (e.g., the Origin line - 02000 and 03000) based on proprietary MIPS processors. They still make those machines, but market pressures forced them to also develop and sell Intel-based shared memory machines. I'll be curious to see how much of SGI's extensive work on IRIX to let it scale to 1000's of processors efficiently will bubble out to their Linux systems.

    5. Re:kernel sources? by Anonymous Coward · · Score: 0

      "that other on DEC that I can't remember off-hand."

      memory channel???

    6. Re:kernel sources? by Anonymous Coward · · Score: 0

      It's called NUMA, and it's already there.

    7. Re:kernel sources? by SlowCoder · · Score: 1

      It's most definitly a hardware thing.

      It's a development of regular NUMA technology, called NUMAflex. See SGIs page about NUMAflex for more in-detail information.

  7. Better than Beowulf for normal use... by TWX · · Score: 5, Informative

    You're better off using mosix. It'll allow for more normal (ie, not beowulf specific) applications to thread across computers. I'd imagine that an open-mosix setup (like the ones using the knoppix boot CDs tailored to it) could probably make for a fairly powerful computing cluster very easily.

    --
    Do not look into laser with remaining eye.
    1. Re:Better than Beowulf for normal use... by yuvtob · · Score: 3, Informative

      while you are probably right that for most cases mosix will do just fine (I used it for a ~50 PC cluster at nights for DSP calcs), these machines are for super-computer calculations that require a lot of memory. If you even could run a 2GB process on mosix, it would be slowed down by the network, and these beasts can run 100GB processes at a 2GB/s interconnect !

    2. Re:Better than Beowulf for normal use... by ERJ · · Score: 5, Informative

      Mosix is nice, because it treats the cluster like a single, large, multi-cpu box by simply allocating threads to different boxes. The nice thing about this is that any multi-threaded program can take advantage (as stated in the parent post).

      However, this also can cause problems. Most threaded programs are written assuming that all the threads have high speed (i.e. system bus / cpu cache) access to shared information. When we introduce the latency incurred by a network, this can cause programs to run alot slower then they would if they simply had all the threads on a single box. Obviously, it all depends on how the program was written, and what it does.

      If you are writting a program specifically for a cluster, I would suggest instead looking at something like LAM-MPI. This allows for a much more controlling approach to be taken. It is more work (you have to decide how the work will be split) but it allows for much better control of where and what is being done and how to optimize it.

    3. Re:Better than Beowulf for normal use... by battjt · · Score: 2, Informative

      Threads can't be migrated. Only processes can be migrated.

      http://howto.ipng.be/openMosixWiki/index.php/App li cations%20using%20pthreads

      You have to write your application as a bunch of processes to take advantage of a mosix cluster.

      Joe

      --
      Joe Batt Solid Design
    4. Re:Better than Beowulf for normal use... by Anonymous Coward · · Score: 0

      Your response is out of place. The post you are responding to is answering a question about Beowulf clusters, not about the SGI machine.

    5. Re:Better than Beowulf for normal use... by Anonymous Coward · · Score: 0

      For this machine (as opposed to clusters in general), you'll benefit from using a threading package that assumes low inter-thread latency. The latency between nodes in the SGI machine is probably an order of magnitude lower (if not considerably more) than it would be on a conventional cluster with software-supported shared memory.

    6. Re:Better than Beowulf for normal use... by Nick+T.+Hammer · · Score: 1

      Yes, but Altix has something more: Global Shared Memory with NUMAFlex...
      The memory is globally shared between each nodes and make the cluster looks like an SMP machine.
      Very usefull if all your threads have to share a large amount of memory (database)...

    7. Re:Better than Beowulf for normal use... by haraldm · · Score: 1

      That depends a LOT on the application. If you want to run things like StarCD, for example, you need the MPI libraries and a low latency network, like Myrinet or SCI.

      --
      open (SIG, "</dev/zero"); $sig = <SIG>; close SIG;
    8. Re:Better than Beowulf for normal use... by Anonymous Coward · · Score: 0

      I thought this was how the old PTHREADS implementation worked on linux anyway (I should state it used to--hence the *old* PTHREADS implementation. I am aware that the kernel supports real threading now). Is the implementation on mosix different? (RTFA, right?)

  8. Rocket Haid... by poptones · · Score: 4, Funny
    Now those obsessed geniuses have even more reason to forget to change the oil in their cars...

    (Inside joke for my ol' friends at ORNL...)

    1. Re:Rocket Haid... by Raleel · · Score: 1

      that would be an inside joke for anyone who supports scientist

      (from an employee of PNNL)

      --
      -- Who is the bigger fool? The fool or the fool who follows him? --
    2. Re:Rocket Haid... by heir2chaos · · Score: 1

      Actually, it would be more of a reason for them not to crawl out of the various holes at ORNL to eat lunch and talk to themselves. Oh my, I need to get out of Oak Ridge :)

    3. Re:Rocket Haid... by confused+one · · Score: 1

      Sounds like people I know (JLAB)

    4. Re:Rocket Haid... by Obiwan+Kenobi · · Score: 1

      Hey! I too haven't gotten around to changing my oil...and suffered for it.

      And I'm a network admin at their credit union...

    5. Re:Rocket Haid... by poptones · · Score: 1
      I once took a valve cover off an Escort engine and found... black jell-o. Couldn't even see the rocker arms. When told her engine was ruined and not under warranty because it had obviously been abused through its 60,000 miles of life, she defiantly pointed out she "had too changed the oil! I had it changed twice!"

      ...they can put a man on the moon, but...

  9. SCO and Microsoft reactions? by mao+che+minh · · Score: 5, Interesting
    I wonder what kind of FUD Microsft and SCO will cook up to try to thwart this new display of raw power. McNealy seems intent on not only winning the Asshat award, but outright retiring it in his honor.

    It's funny that Microsoft always tries to downplay Linux's enterprise capabilities, when Linux has been scaled to far more power then Microsoft's best offering for years now. Windows 2003 is a clumsy, bloated, closed source chunk of green crap.

    1. Re:SCO and Microsoft reactions? by Anonymous Coward · · Score: 1, Funny

      The fun part is that this same thing is 100% impossible with Windows.

      it looks like that Microsoft's products are NOT enterprise class and Linux certianly is.

      It just is another nail I get to torture the MSCE's here with....

      "Hey, can your 2003 advanced server do this... oh wait... sorry, I was comparing powerful OS capabilities important to an enterprise to a low end operating system...."

      sorry guys.....

    2. Re:SCO and Microsoft reactions? by ichimunki · · Score: 1

      McNealy? I think you mean McBride.

      --
      I do not have a signature
    3. Re:SCO and Microsoft reactions? by Anonymous Coward · · Score: 0, Interesting

      And why would Microsoft even care to make a comment? This isn't exactly an area they have been even attempting to compete in. IBM is probably the most direct competition for this kind of power.

      As for your comment on Windows 2003, what qualifies you to make any comment on a system that hasn't been released yet?

      I won't even comment on your incorrect Mc usage (although Scott McNealy of Sun is an Asshat also).

      Oh I forgot. You are just a troll and I am wasting my breath.

    4. Re:SCO and Microsoft reactions? by Anonymous Coward · · Score: 0

      You fucking idiot. Don't attack people when you're just as clueless. Windows 2003 has been released a long while ago. In fact, I'm using it right now.

    5. Re:SCO and Microsoft reactions? by cgb8176 · · Score: 2, Interesting

      It's funny that Microsoft always tries to downplay Linux's enterprise capabilities, when Linux has been scaled to far more power then Microsoft's best offering for years now.

      RTFA. They are using this machine for research in the "sciences, clean energy management and production, environmental protection, and homeland security."

      It's not a web server, and it isn't demonstrating "enterprise capabilities." Windows has never been intended for, or used for, scientific computing on a large scale.

    6. Re:SCO and Microsoft reactions? by mjmalone · · Score: 1

      Pretty sure we have a win2k3 active directory server running in the NOC here... Or maybe im just insane?

    7. Re:SCO and Microsoft reactions? by Anonymous Coward · · Score: 0

      Windows 2003 has TOO been released... or did you mean working?

    8. Re:SCO and Microsoft reactions? by Valar · · Score: 1

      You are absolutly correct. As far as I can tell, Windows was designed to turn my computer into an overpriced Atari 2600 replacement. Or maybe it's some kind of value-added DOS distribution.

    9. Re:SCO and Microsoft reactions? by Anonymous Coward · · Score: 0

      You have got to be kidding.

      Microsoft pounds out their OS's "excellence in the enterprise" almost every day. Travel around the world. Look at the business channels on TV. Then, tell me again "Windows has never been intended for (...) scientific computing on a large scale."

      I know a few CEO's that would explode in your face at such a statement, to say the least.

      OTOH, Who cares? I dumped the suckers long ago. Somebody Elses Problem.

  10. Hey, at least it's not running IRIX by der_saeufer · · Score: 1

    The best part about it is that you can actually run something on it, unlike SGi's older IRIX-based crap.... unless you like re-writing code to blow your nose, be glad this one is running Linux.

    1. Re:Hey, at least it's not running IRIX by Chicane-UK · · Score: 4, Informative

      Um..

      I always liked Irix, and everyone I ever talked to who used Irix liked it. The GUI is about 500x more usable than the horrors of OpenWindows or CDE on Solaris.. bleugh.

      --
      "Hey! Unless this is a nude love-in, get the hell off my property!!"
    2. Re:Hey, at least it's not running IRIX by Anonymous Coward · · Score: 0

      Let me guess, you never used IRIX, or you tried to compile some brain-dead (as in x86 only, Little Endian only, 32bit only, and worst of all GCC only) Linux programs to it (in that case you will probably still have problems due to people thinking if it works on Linux it works on all other UNIXES)...

    3. Re:Hey, at least it's not running IRIX by Anonymous Coward · · Score: 0

      IRIX has some huge advantages over Linux for machines of this scale. SGI has spent enormous efforts to make Irix truly scalable to 1000+ tightly-coupled processors. To call this box a "cluster" is a major insult (but probably a marketing necessary).

    4. Re:Hey, at least it's not running IRIX by bigqueso · · Score: 1

      Yeah, irix works well if you already have code ported to it. Ever try porting code to an irix? You would not believe the hoops I had to jump though in order to get xmms to compile.

      On a desktop, irix is dead. However we have a few Origins where I work, and they kick some serious butt when it comes to writing parallel code. The problem with Origins is they cost way too much money for the amount of processing power they provide. The Altix is a step in the right direction where you can have industry backed CPU and OS support, but provide the flexibility of a SMP machine.

    5. Re:Hey, at least it's not running IRIX by Nexus+Seven · · Score: 1

      Why not just download it?

      SGI are kind enough to jump through all the hoops on your behalf, for most all freeware apps.

    6. Re:Hey, at least it's not running IRIX by caluml · · Score: 1
      The GUI is about 500x more usable than the horrors of OpenWindows or CDE on Solaris.. bleugh.

      I've never used Irix, but rating the GUI to Solaris' nastiness doesn't impress me at all ;) KDE baby...

    7. Re:Hey, at least it's not running IRIX by platypus · · Score: 3, Funny

      I hated it, if it helps.

    8. Re:Hey, at least it's not running IRIX by the+gnat · · Score: 2, Interesting

      I always liked Irix, and everyone I ever talked to who used Irix liked it. The GUI is about 500x more usable than the horrors of OpenWindows or CDE on Solaris.. bleugh.

      I vastly prefer 4DWM to GNOME or KDE as well. I'm helping a coworker set up a Dell inspiron 7500 (P3-700) with Linux, and he immediately complained that KDE was far too slow. I switched to WindowMaker, and he immediately noticed the difference. This is a three-year-old machine, with tons of memory and a reasonable processor, and it crawls with KDE3. Pathetic.

      Meanwhile, you can run the latest version of Irix on a seven-year-old SGI box (and even older) and it'll still be smooth. My Indy at home feels just as responsive as any PC I've ever used. I wouldn't call it *fast* by any stretch of the imagination, but the OS alone does not cripple the computer. I'm a huge Linux fan, but there are tons of examples like this where it just hasn't caught up to the more polished offerings.

    9. Re:Hey, at least it's not running IRIX by Anonymous Coward · · Score: 0

      IRIX has much better SMP support that can scale to thousands of processors. As a user of both IRIX and Linux I must say that when it comes to number crunching, IRIX can do more with less when compared to Linux.

    10. Re:Hey, at least it's not running IRIX by lga · · Score: 1

      IRIX is not dead on the desktop. It's dying, but could be revived yet. Perhaps the new Tezro workstations might do the trick, although most people won't see this kind of machine unless their employer buys one.

      IRIX is the most stable unix I have used. It hardly ever crashes. OS updates don't break the system. The desktop is quite nice although not as nice as gnome. But that's not a problem because IRIX can also run gnome!

      I know with a 64 processor machine I would prefer to use IRIX if I could. Linux is a good second choice.

    11. Re:Hey, at least it's not running IRIX by Anonymous Coward · · Score: 0

      Yawn. Chimera/Tezro/Whatever is nothing more or less than an Origin 350 with Odyssey graphics. There's nothing new there whatsoever.

      IRIX is possibly the finest operating system I've ever used. The SGI Origin is still, seven years later, miles ahead of anything anybody else has to offer except maybe Cray. But SGI workstations are crap now. Alas.

    12. Re:Hey, at least it's not running IRIX by Anonymous Coward · · Score: 0

      So you're comparing a desktop environment to a fucking window manager? You should have just stuck with Windows or MacOS.

    13. Re:Hey, at least it's not running IRIX by dwater · · Score: 1

      > The desktop is quite nice although not as nice as gnome.
      > But that's not a problem because IRIX can also run
      > gnome!

      And if you're otherwise inclined, you can also get the IRIX desktop for Linux (no, I haven't tried it) :

      http://www.5dwm.org/

      Personally, I prefer 4Dwm to all other desktops I've tried, *especially* this crappy thing I have to use on my Mac. How anyone can think the Apple GUI is good is beyond me...!

      (special 'hi' to b-a)

      --
      Max.
    14. Re:Hey, at least it's not running IRIX by PotatoHead · · Score: 1

      I agree with this. Anyone that says X is slow, should sit in front of an older SGI sometime.

      4Dwm is snappy --even on a 150Mhz machine.

      IRIX is a nice place to work. Now that I am getting more into Linux, the little differences annoy me. SGI does do its part though. freeware.sgi.com is a very nice resource. Get any old SGI, point it there and download most of the good OSS built and ready for your machine.

    15. Re:Hey, at least it's not running IRIX by Xolotl · · Score: 1

      Actually, the IRIX GUI _is_ a desktop environment, with a filemanager, trashcan, desktop icons and the rest. And all this more than ten years ago.

  11. Best of both worlds... by goats_in_boats · · Score: 2, Funny

    ...now you get obscene frame rates on quake III while searching for those pesky pockets of natural gas!

  12. Just wait. by Anonymous Coward · · Score: 0, Insightful

    1) Buy 100 PS3's.
    2) Cluster them together mit linux
    3) ???
    4) -1, redundant!

  13. Re:First one... by flyneye · · Score: 1

    damn,ya beat me to submit button
    -flyneye from 2 threads down-

    --
    *Repent!Quit Your Job!Slack Off!The World Ends Tomorrow and You May Die!
  14. Yanking from my journal entry of 6/30/03 by anzha · · Score: 4, Informative

    HPC Wire had an article that I referenced in my journal on 6/30.

    It's an interesting machine. I'd love to get one to play with. I'm sure our benchmarkers will have some even more interesting comments once they're done. Expect teething problems, folks. Systems of this size and complexity take time to break in.

    --
    Do you know why the road less traveled by is littered with the bones of the unwary?
    1. Re:Yanking from my journal entry of 6/30/03 by Anonymous Coward · · Score: 0
  15. lites by NetMagi · · Score: 3, Funny

    makes me just wanna turn off the lights and look at all those LED's blinkin!

    1. Re:lites by CoolVibe · · Score: 5, Interesting
      I've experienced it the other wat around once. At some previous $workplace, we had this humongous SGI Origin 3800 cluster. Due to a city-wide brown out, and due to the fact that we were just installing the diesel-powered generators, the thing had to survive for a couple of hours on the nobreak. Sure, all the lights in the building were out, but the behemoth was still churning. We (the venerable sysadmins) were trying to decouple a partition so we could hook up a console to ot to bring down the thing gracefully. Of course, that wasn't that easy.

      Suddenly the nobreak was all out, and the billion dollar machine went *poof* - down. Damage? A couple of SCSI disks, but of course everything was mirrored and had parity so even with the damaged disks, there was no data loss.

      Then (after a few hours) the powerfaillure ended, the lights went back on in the building, but the lights on the big cluster were still off. The other way round than you'd like to see. Although, when the building power was out, and the nobreak for the machine was active, it sure was a pretty sight. Although, with the impending doom, I didn't really have time to appreciate it.

    2. Re:lites by Leebert · · Score: 4, Informative

      the billion dollar machine

      What the hell kind of Origin 3800 do YOU have? ISTR ours (512-proc) was roughly $10M.

    3. Re:lites by CoolVibe · · Score: 1
      I was of course overreacting. The machine is expensive. The machine has 1024 procs, big fat arrays of scsi disks (think multiple multiple terabyte storage, raid5 and mirrored (can't say how much, it's too long ago, but it was heaps. Several room filling cabinets full of just SCSI disks), and really shitloads of RAM (think terabytes in total). It's probably not a billion dollars, but it's certainly quarter way if you count support contracts and upkeep.

      I was actually surprised that the thing managed to run for about an half an hour on just the nobreak, together with a Cray and a 256 node Beowulf cluster which we did manage to gracefully shut down in time. If we had the power generators ready (they were due to be operational in the next few days), they could've ran for almost forever without a power faillure.

      Ain't Finnagle a bitch.

      Of course I don't work there anymore. And no, I wasn't fired over this. Nobody was. The contracter that didn't make with the backup power (and the power company for causing a 5 hour power blackout) did get a hefty bill though.

    4. Re:lites by green+pizza · · Score: 3, Informative

      SGI Origin 3800 cluster

      Just to nitpick... most Origins are not clusters but rather one large single machine. It is possible to partition the machine in firmware and have each partition talk to others over the existing (and now unused) numalink interconnects... but it's much faster (even for plain MPI code) to just run the beast as one large single machine.

    5. Re:lites by Anonymous Coward · · Score: 3, Informative

      The machine has 1024 procs

      There are two 1,024-processor Origin 3000's in the world. One is in Eagan, Minnesota. The other is at NASA. The NASA machine is called chapman. It has 256 GB of RAM. Not terabytes.

      How do I know this? Because I'm sitting here looking at lomax right now.

      You're a... whaddya call it. Liar.

    6. Re:lites by Leebert · · Score: 1

      The NASA machine is called chapman. How do I know this? Because I'm sitting here looking at lomax right now.

      I have an account on turing. ;) In fact, I was through your datacenter back in February.

    7. Re:lites by Anonymous Coward · · Score: 0

      Somebody mod this fellow down, please. He is evidently making stuff up to try to impress the noobs.

      Thanks.

    8. Re:lites by CoolVibe · · Score: 1

      Nope, the Origin 3800 at this place has 1024 procs. That's where I worked. You're behind the times, so to speak. Europe has big fast number crunching behemoths too.

    9. Re:lites by Anonymous Coward · · Score: 0

      Dude, you're making shit up. The SARA machine was originally a 256p, later upgraded to a 512.

    10. Re:lites by CoolVibe · · Score: 1

      Nope I'm not. You are right that it originally was 256, then 512, but now it's 1024.

    11. Re:lites by CoolVibe · · Score: 2, Informative

      Oh, I found a little page on the sara website where it is clarified (can't get onto the intranet anymore, else I'd have mirrored some better specs). Anyway, more about TERAS here.

    12. Re:lites by Anonymous Coward · · Score: 0

      Caught you. The machine at SARA was originally sold as a 512p. It was never a 256p.

      You suck.

    13. Re:lites by Anonymous Coward · · Score: 0

      Hmm. Seems like the AC who called you on it is right. SARA's machine is not a 1024-processor at all. It's two 512-processors running in a cluster.

      Now, only one question remains. Are you a liar, like he said, or are you just really ignorant?

      (You fucked up the "terabytes of RAM" thing, too, but that doesn't answer the question. You could have been making stuff up, or you could have been just being stupid.)

    14. Re:lites by CoolVibe · · Score: 1

      I said 1024 proc "cluster", didn't I? And yes, a TB of ram is about right. The plural thing was just a brainfart.

    15. Re:lites by Anonymous Coward · · Score: 0

      If by "brainfart" you meant to say "vicious backpedal to keep from getting caught in a lie," then yes.

      Fucker.

  16. Obligatory: Mods on Crack! by panda · · Score: 5, Funny

    OK, so the moderators are on crack today. What's with all these obviously "funny" posts getting moderated as "insightful?"

    Guess it's time to meta-moderate!

    --
    Just be sure to wear the gold uniform when you beam down -- you know what happens when you wear the red one.
    1. Re:Obligatory: Mods on Crack! by darkov · · Score: 5, Funny

      OK, so the moderators are on crack today.

      Only today?

    2. Re:Obligatory: Mods on Crack! by duguk · · Score: 1

      Mod parent +1 insightful!

  17. O.R.N is not the only one by Anonymous Coward · · Score: 0

    The israeli biological research center also got one this week.

  18. Even better! by gladbach · · Score: 1

    Imagine running a Counter-Strike server on this!!!

    --
    "Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd all be running around in darkened rooms,
    1. Re:Even better! by Trigun · · Score: 1

      Wall-hacking at the speed of light. YES!

  19. Oops (RTFA) by Anonymous Coward · · Score: 5, Informative

    The machine has 256 processors for 1.5 teraflops, not 64.

    1. Re:Oops (RTFA) by Anonymous Coward · · Score: 1, Informative

      The 64 is how many CPUs it can use in a single node...RTFA =)
      It has 4 nodes...4 x 64 = 256

    2. Re:Oops (RTFA) by Repran · · Score: 1

      64 in each node. Times 4 is 256. RTFA yourself ;-)

      --

      -- Contradictions only exist in thought - not in reality.

  20. So... by Mipsalawishus · · Score: 3, Funny

    How hard would it be to /. one of these things??

    1. Re:So... by Anonymous Coward · · Score: 0

      its on 56k baud, nae bother...

    2. Re:So... by Drakonian · · Score: 1

      Processing power has very little to do with bandwidth.

      --
      Random is the New Order.
    3. Re:So... by Nick_dm · · Score: 1

      Yes, but some sites are taken down because of the strain on the databases generating dynamic pages (you sometimes notice people switching their website to a static index.html to display a message when they are slashdotted) so in these cases the server is the issue not the bandwidth.

    4. Re:So... by ocelotbob · · Score: 3, Insightful

      Depends on the bandwidth into the machine more than anything else. Most /.ings, unless the database explodes into a shower of sparks, are limited by the bandwidth of the machine more than anything else. It'll be quite easy to /. it if it's only got a T1 or so. If it's got a 10Gb connection or two, I'd imagine that the system load wouldn't even be noticed.

      --

      Marxism is the opiate of dumbasses

  21. Re:Why Linux? by Anonymous Coward · · Score: 0

    I think the funniest part is that this got modded UP. Wow, was retarded mods.

  22. Re:Just wait again by Anonymous Coward · · Score: 0
    1) Buy 100 PS3's.
    2) Cluster them together mit linux
    3) ???
    4) -1, redundant!
    5)implement global memory
    6)???
    7)1, interesting ?!

    Or did i miss something? aint like i rtfa...

  23. Re:Why Linux? by Anonymous Coward · · Score: 0

    They enjoyed Jizz cola

    Uh...I'm speechless.

  24. RETARDED MODERATORS by mhesseltine · · Score: 0, Troll

    WHO THE FUCK MODERATED THIS AS "INTERESTING"? DID YOU READ IT? -1 TROLL, -1 OFFTOPIC, -1 FLAMEBAIT

    Posting logged in, with bonus to point out the egregious nature of this posting

    --
    Overrated / Underrated : Moderation :: Anonymous Coward : Posting
    1. Re:RETARDED MODERATORS by Carbonite · · Score: 0, Offtopic

      I really should have known something was amiss after "Jizz Cola", but I kept reading. I even clicked the link to display the entire comment. It was horrible, possibly written by a geek descendent of Lovecraft. Excuse me now while I wash my brain out with soap...

      --
      ich muß mehr Kuhglocke haben
    2. Re:RETARDED MODERATORS by mhesseltine · · Score: 0, Flamebait

      What's funnier is that I get a -1 Flamebait for pointing out the fucked up moderation.

      Oh well. I have karma to burn. Bring it on bitches

      --
      Overrated / Underrated : Moderation :: Anonymous Coward : Posting
    3. Re:RETARDED MODERATORS by Anonymous Coward · · Score: 0

      I should have some mod points in the next day or so. You made a very valid point and got modded "Flamebait" while the shit you were replying to got modded UP ... that's fucked up. But don't worry, when I get my mod points you'll get your karma back.

    4. Re:RETARDED MODERATORS by mhesseltine · · Score: 1

      Thanks for the concern, but don't worry about my karma. I've said it before, I have plenty to burn. I just wish moderators would actually read the comment before going "Look, it's really long, therefore it must be at least interesting"

      --
      Overrated / Underrated : Moderation :: Anonymous Coward : Posting
    5. Re:RETARDED MODERATORS by Q+Who · · Score: 1

      Say, are you stupid or something?

      The trolls themselves mod it up.

    6. Re:RETARDED MODERATORS by mhesseltine · · Score: 1

      <voice char="Forrest Gump">Stupid is as stupid does</voice>

      Seriously, I understand how the trolls have multiple accounts, get points in one, post crap in another, and use the first to mod up the second. However, the editors have unlimited points, along with the rest of the "legit" moderators, and comments like that shouldn't stay at +2 informative for that long.

      --
      Overrated / Underrated : Moderation :: Anonymous Coward : Posting
    7. Re:RETARDED MODERATORS by BigBadBri · · Score: 1
      More Burroughs than Lovecraft, surely?

      I thought Gibson was the descendant of Lovecraft...

      --
      oh brave new world, that has such people in it!
  25. Re:Why Linux? by Anonymous Coward · · Score: 0

    LTOR !yag si xuniL

  26. Great.. by nother_nix_hacker · · Score: 1

    Great.. so which option in the .config do I enable?

    1. Re:Great.. by Creepy+Crawler · · Score: 1

      >>Great.. so which option in the .config do I enable?

      All of them. And turn on Oompha-matic CPU eater.

      --
    2. Re:Great.. by Anonymous Coward · · Score: 0

      CONFIG_IA64_SN2

      though YMMV, because you'd be pretty crazy to run a stock kernel on these machines. SGI's patches (which come with the machine, are under GPL, and will no doubt be accepted in the main kernel one day (they've already been going in for months now, bit by bit) are what you want.

  27. Conversion scale? by jackDuhRipper · · Score: 5, Funny

    What's that in bogomips?

    1. Re:Conversion scale? by Alien+Being · · Score: 1

      Oh, about 1 megabogomip.

    2. Re:Conversion scale? by brakk · · Score: 1

      But, how many Volkswagen Beetles is that?

  28. Re:O.R.N is not the only one by Trigun · · Score: 0, Offtopic

    Ooh, those poor Palestinians...

    SARS is going to look like a field trip to the zoo :)

  29. now.. by ciupman · · Score: 1

    will it be bundled with neck plugs?

    --
    I fuse with Mercer every single day...
  30. Re:Why Linux? by Anonymous Coward · · Score: 0

    Alright, I know its hot on the East Coast in the US (and just about every where else in the US) and folks are going to ridiculous lengths to stay cool, but gaddamn! The moderators must have their heads in their asses and their toes on the keyboard to actually mod this shit up.

    HELLO! PAY ATTENTION! I'm being deliberately insulting here in the vain hope that this will never occur again.

    In between starting this comment, and ending it, I see the moderators got their act together....'bout damn time. Took your damn time too. Try not to slip up again: we're losing faith here.

  31. SGI: Unsung coorporate heros ? by Anonymous Coward · · Score: 5, Interesting

    Perhaps this will finally get SGI's Open Source Software efforts in the spotlight.So far every other major hardware vendor has jumped the bandwagon making a lot of noise, and trying to get free publicity. SGI however has always quietly contributed large amounts of knowledge but always in a modest or even shy way (sometimes even publicly denying involvement, but working in secret :) ).
    In the meantime their additions have contributed quite a bit to open en free thinking in software, take OpenGL and open Inventor, or even to the kernel directly as with the XFS filesystem.
    I always liked this approach more than the hyping others have done with linux, but unfortunately this has kept them unadorned within the community. With the Altix cluster (as with their GNU/Linux workstations,which unfortunately failed) I think they have shown that they put their money where their mouth isn't.

    I think it's only fair that when we are talking about the large coorporate players in the OSS field SGI at least deserves a footnote for their efforts instead of just hammering exclusively on IBM,Sun etc..as the great backers.

    I know, I know. It's a coorporation, so they inherently put money over freedom, it's just something I noticed because of the lack of their name in any high-profile discussions, which I think is unfair.

    1. Re:SGI: Unsung coorporate heros ? by jd · · Score: 2, Interesting
      One thing I've considered for some time is a "league table" of companies involved in "Open Source Software".


      The table would record the number of packages released, the number of patches, and the licenses used for each. Originally, I was going to make a four-way split - open-source packages or patches, and packages/patches for open-source OS'.


      From this, you could create some kind of scoring system, and thus compare the "open-sourceness" of companies. (From the above, it should be obvious that I consider promotion of an Open Source OS to be important, whether the company actually releases any Open Source code itself or not.)


      In every league table I've drawn up - I've just not had the time to complete or maintain such a table - IBM ranks first, and SGI is second. The gap is surprisingly close between these two. No other major corporation even comes close.


      If you want to understand why SGI ranks so highly, look at their oss.sgi.com site, under projects and also under propack. (Propack is the collection of a lot of their Linux-specific code, including XFS for Linux.)


      Between Propack, OpenGL, GLX, Open Inventor, Coin, their OB1 code dump, their Apache 1.3.x acceleration patches (which apparently resulted in a political war between the Apache group and SGI), Rhino, their patches for Mozilla, plus all of their Open Source code for IRIX, there can be no serious question that SGI has done a lot.


      (Patches SGI used to support, but dropped, include AIO - Asynchronous I/O, and Scheduled Transfer Protocol. Both were for Linux.)


      In comparison, Compaq has a patch and a package for clustering, HP has a plug-in scheduler system, and The Open Group now provides a version of Motif for Linux.


      Hmmm. Yeah. Even with Compaq's take-over of HP, SGI are still so far ahead of the game that it's not funny.

      --
      It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
    2. Re:SGI: Unsung coorporate heros ? by Anonymous Coward · · Score: 0

      You are on the right track. One of the best ways for us to show appreciation for Good Work (tm), is to acknowledge and support the companies we like.

      Yeah, yeah, "voting with your wallet, old cliché".

      Yep. But still. We would be really stupid not to.

  32. Re:Why Linux? by CoolVibe · · Score: 0, Troll
    That's an interesting piece of fiction. Guess I'm lucky I'm not a UNIX coder but more of a FreeBSD coder. :-)

    There's also a lot of stereotypical imagery in your story. Did you know that, yes, lots of people in the Open Source world are gay (Eric Allman is, Alan Turing is, to name but a few). Needless to say, without mr. Turing's efforts, you wouldn't be typing this story on this website. A little more respect for people's sexual preference please. I'm not homosexual, but many of my friends are. And sure, there might be some asshats around, but I can assure you they are in the minority.

    Are you sure this isn't just misdirected homophobia? Oh well, we will never find out, since you are an anonymous coward.

    (and no, Free/Net/OpenBSD and Linux are not UNIXen, and you desperately need a girl/boyfriend. Or a life. Maybe both)

  33. And just how did they accomplish this... by Kevinv · · Score: 4, Funny

    without SCO's help?

    1. Re:And just how did they accomplish this... by Mostly+a+lurker · · Score: 1
      And just how did they accomplish this without SCO's help?

      It's really 99% SCO Xenix inserted with IBM's help.

  34. uh huh by notque · · Score: 2, Insightful

    Mod parent +1 Funny!

    --
    http://use.perl.org
  35. 64 processors = 1.5 Cells by Jeff+Carr · · Score: 1

    What I find amazing is that the Cell is supposed to run up to a TeraFlop when it reaches production. That compared to a 64 processor Linux cluster.

    --
    The television will not be revolutionized.
    1. Re:64 processors = 1.5 Cells by Anonymous Coward · · Score: 0
      What I find amazing is that the Cell is supposed to run up to a TeraFlop when it reaches production. That compared to a 64 processor Linux cluster.

      Yeah, right. Minus the 64-bit word size, more than 3/4 of the per-processor cache, more than 90% of the memory, nearly all of the interprocessor bandwidth and I/O bandwidth.
    2. Re:64 processors = 1.5 Cells by AmishSlayer · · Score: 2, Informative

      What I find amazing is that the Cell is supposed to run up to a TeraFlop when it reaches production. That compared to a 64 processor Linux cluster.

      thats 64 processors per node

    3. Re:64 processors = 1.5 Cells by sydres · · Score: 1

      the geforce4 ti 4600 when fully loaded with matrix math is said to hit the teraflop mark but
      i guess that only helps for non programmeables
      wonder what the ati r300 and nv3x chips could hit
      in a programmed useage

    4. Re:64 processors = 1.5 Cells by Anonymous Coward · · Score: 0

      Only if you count every single 8-bit add and multiply as a "FLOP". Which you don't.

  36. You live in Redmond by chance? by Jerk+City+Troll · · Score: 1

    So this is what Microsoft developers sit around and dream up all day, right?

    1. Re:You live in Redmond by chance? by Anonymous Coward · · Score: 0

      Probably true...most of Microsoft's successes involve copying the truly great.

  37. 1.5 Teraflops!? by Phishcast · · Score: 1

    Great Scott!

  38. That really great by Anonymous Coward · · Score: 0

    What a beautiful advertisement slashdot is running for SGI. Granted, it's an important step for Linux. Still, it's a completely biased write-up. Why not compare it to what Windows is capable of. Who knows, Linux might have already surpassed the SMP ability of Windows.

  39. Intel and math by geekmetal · · Score: 1
    With 256 new Intel Itanium 2 processors, 2TB of global shared memory, and 1.5 TFLOPS of computational power at their disposal, ORNL researchers can simulate and analyze data sets of extraordinary size and complexity. The groundbreaking capabilities of the SGI Altix 3000 system will help ORNL drive new generations of scientific applications hungry for increased computing power and capacity. The applications include those used in computational biology and genetic research, as well as climate modeling, in which researchers project the potential long-term impact of such environmental threats as pollution and ozone depletion.

    Looks like the number crunching reliability of intel processors has taken a good boost here.

    --
    There are two kinds of egotists: 1) Those who admit it 2) The rest of us
  40. Re:Just wait again by Anonymous Coward · · Score: 0

    1) Buy 100 PS3's.
    2) Cluster them together mit linux
    3) ???
    4) -1, redundant!
    5) implement global memory
    6) ???
    7) 1, interesting ?!
    8) ???
    9) Profit!

  41. Re:Why Linux? by NetMagi · · Score: 1

    I seriously thought it was funny. . . definetely unexpected

  42. You're my hero by Anonymous Coward · · Score: 0

    MY OWN SPECIAL HERO!

  43. ALMOST as good as my "Screwed Out of a Job" troll by Anonymous Coward · · Score: 0

    ...except yours was a little too long to fit on the page. Good imagery, though, and way to get those suckers to read your whole post! A good Troll Tuesday special!

  44. Cluster at Oakridge by ratfynk · · Score: 1

    A cluster at Oak Ridge?
    Better check the rads and the terraflops. Might even mutate into a giant deranged Penguin that eats Seattle.

    --
    OH THE SHAME I fell off the wagon and use sigs again!
    1. Re:Cluster at Oakridge by pdbogen · · Score: 1

      > giant deranged Penguin that eats Seattle.

      I don't see the downside?

  45. 1.5 TFLOPS, but which flops bencmark and weighting by Animixer · · Score: 1


    I wonder what FLOPS benchmark they are using, and what the weightings are for floating point divides, multiplication, etc (didn't see it in the article.)

    If you want to really cut a machine down to size, see how many FLOPS it gets with a heavy floating point divide weighting. Using a flops benchmark with light divides can make your machine look awesome. :)

    --
    man tunefs | grep fish
  46. Re:Beowulf cluster! by Anonymous Coward · · Score: 0

    Imagine life in infinite debt.

  47. ...scales up to 64 processors by 1s44c · · Score: 1

    scales up to 64 processors

    Imagine a Beowulf cluster of those.

    1. Re:...scales up to 64 processors by lpret · · Score: 1

      You know, the most amazing thing about your post was that it wasn't modded down to oblivion or modded up as hilariously funny. It just sits at 1. That's when you know it's bad, when no one takes the time to even mod you.

      --
      This is my digital signature. 10011011001
  48. Re:First one... by JudgeFurious · · Score: 0, Offtopic

    See this is all wrong. At this time it's modded "Troll" which is a travesty. It's +1 Funny damnit!

    No mod points, can't offset rectal moderator, sorry.

    --
    Appended to the end of comments you post. 120 chars.
  49. Setting one up now by jimshep · · Score: 5, Informative

    We just got ours installed yesterday. I'm still installing software and am starting benchmarks. It's only the deskside version (12 cpus, 24GB RAM, 1TB disk), but still more powerful than the 4-cpu SGI Origins that we have been using.

    It is the first one that the regional SGI reps had actually installed, but since it is almost exactly the same as the MIPS-based origin 3000 servers (with the exception of the obviously different Itanium 2 cpus and supporting chipsets), they ran into almost no problems getting it online. I have also been suprised as to how many commercial codes have already been ported to the platform.

    The main reasons we purchased this machine is for the ease in parallelizing code and the floating point performance of the Itaniam 2 cpus. We're computational materials engineers and the less time we have to spend optimizing codes so that the nodes of a cluster are always kept busy and minimizing I/O bottlenecks gives us more time to concentrate on the theoretical issues.

    It runs RedHat 7.2 with some tweaks by SGI called SGI ProPack. The Propack modifications come on separate CDs, with the proprietary software on separate CDs from the open source software. So far, from the command line, everything works just like my PC. It's kind of strange running Linux on a >$100K machine, but it sure beats dealing with the annoying differences between IRIX and Linux. Now to see if it performs as well as we expect...

    1. Re:Setting one up now by platypus · · Score: 1

      Maybe you'd like to post some kernel compile times? ;)

    2. Re:Setting one up now by Jaeger- · · Score: 2, Funny

      Don't you know you can't follow "only" with the words "12cpu, 24gb ram, 1tb disk"...

      --
      E V E R Y T H I N G I W R I T E I S F A L S E
    3. Re:Setting one up now by Mentally_Overclocked · · Score: 1

      If you have the time, what do you plan on using to benchmark the system. I have looked around, but don't really know what to use. Not quite the same system or setup, but any suggestions are welcome.

      Thanks in advance

      --

      Mathematician, n.:
      Someone who believes imaginary things appear right before your i's.
    4. Re:Setting one up now by Anonymous Coward · · Score: 0

      hpl is the benchmark used to compare systems for the top500 list. This is from the readme:

      High Performance Computing Linpack Benchmark (HPL)
      HPL 1.0 - September 27, 2000

      HPL is a software package that solves a (random) dense linear
      system in double precision (64 bits) arithmetic on
      distributed-memory computers. It can thus be regarded as a
      portable as well as freely available implementation of the
      High Performance Computing Linpack Benchmark.

      The HPL software package requires the availibility on your
      system of an implementation of the Message Passing Interface
      MPI (1.1 compliant). An implementation of either the Basic
      Linear Algebra Subprograms BLAS or the Vector Signal Image
      Processing Library VSIPL is also needed. Machine-specific as
      well as generic implementations of MPI, the BLAS and VSIPL
      are available for a large variety of systems.

  50. Re:To the tune of "Bad boys" by Anonymous Coward · · Score: 0

    rhyme and metre. learn it.

  51. Re:Why Linux? by Anonymous Coward · · Score: 0

    How the fuck is Alan Turing in the 'Open Source World'?

  52. Or, Try Quantix, which comes with some apps by coyote1 · · Score: 2, Informative

    or, try Quantix, which is derived from cluster knoppix. A self-booting ISO with data analysis software, based on Knoppix. This is geared more for scientific apps; it doesn't come with open office, etc, which cluster knoppix does.

    --
    Eat Lamb, 1 million coyotes can't be wrong
  53. Re:Why Linux? by Anonymous Coward · · Score: 0

    I just metamoded someone as "Unfair" who modded this story up, if anyone cares.

  54. Re:hot grits! by karlandtanya · · Score: 0, Offtopic

    a-la Al Green?

    What'd Natalie Portman ever do to you?

    --
    "Reality is that which, when you stop believing in it, it doesn't go away." - Philip K. Dick
  55. Re:Why Linux? by Anonymous Coward · · Score: 0

    Shut up you M$-loving lamer!!!@@!$!$!@$!@$!@$

  56. Benchmarks I can Understand? by Greyfox · · Score: 2, Funny
    How about the Quake 3 framerate?
    Kernel compile time with make -j?

    Hmm, what are some other good ones?

    --

    I'm trying to teach myself to set people on fire with my mind... Is it hot in here?

    1. Re:Benchmarks I can Understand? by Anonymous Coward · · Score: 0

      As for Quake, the Altix does not have any graphics options. I think you hook a dumb terminal to it for console access. Everything else is done through the network.

  57. Macs are better by Anonymous Coward · · Score: 0

    They have a pretty case, and are MUCH faster at Photoshop.

  58. Mosix... by wowbagger · · Score: 2, Informative

    The thing about Mosix is the costs of process migration.

    First, you have to understand process migration. In a mosix cluster, a running process can be moved, lock stock and barrel, from one CPU to another. All that is left behind is a "stub" process that forwards all file I/O across the network to the new location. So, if the program was a 3D raytracer that had the source description file and the output file open, after migration all file accesses to those files would be forwarded over the network to the stub (since you cannot guarantee that the remote machine can access those files in the same way.)

    Now, this is great for programs that do little file I/O but lots of computing (for example the ray tracer I just described.)

    However, the process must be set up on the local node first, then migrated. If the process has a 3 G core image (is taking up 3G of memory), then 3G of stuff has to be shoved across the wire, while the program is frozen. Thus, migrating a process is expensive.

    Now, if you have a bunch of long-running compute bound processes this is a net win (for example, rendering a movie might benefit). But something like building the Linux kernel won't benefit, since what you have is a bunch of short running, high I/O jobs.

    We have a Mosix cluster at work. I tried using it as a compile farm, and the results were disappointing. Not surprising - I was NOT using it for what it was designed for.

    However, if we can ever get the FPGA synthesis tools running natively under Linux, the hardware types are going to be quite happy....

    1. Re:Mosix... by Ensign+Nemo · · Score: 1

      At my workplace we setup a openmosix cluster. We use synplify_pro, for which there is a native linux port, for synthesis and it works BEAUTIFULLY on the cluster in batch mode.

  59. Yes. I do that now. by Anonymous Coward · · Score: 0

    And StarOffice only takes 2 hours to load now!

    1. Re:Yes. I do that now. by Anonymous Coward · · Score: 0

      On the other hand, it only takes a few minutes to crash, so it's well worth the wait.

  60. Cost?$ by Anonymous Coward · · Score: 0

    What do you reckon one of these here suckers will set yall back for?

  61. Re:Why Linux? by Anonymous Coward · · Score: 0

    Want a cookie?

  62. Re:Just wait again by Anonymous Coward · · Score: 0

    1) Buy 100 PS3's.
    2) Cluster them together mit linux
    3) ???
    4) -1, redundant!
    5) implement global memory
    6) ???
    7) 1, interesting ?!
    8) ???
    9) Profit!
    10) Invest money.
    11)???
    12)All your base are belong to us!!!

  63. Re:ALMOST as good as my "Screwed Out of a Job" tro by Anonymous Coward · · Score: 0

    Please post this, I am not familiar with this material.

    Thanks,
    Trollaxor

  64. so uh... by Anonymous Coward · · Score: 0

    How long before SkyNet^H^H^H^H^H^H ORNL becomes self-aware?

  65. One Word: by mbourgon · · Score: 1

    Zilla. For OS X. Make a cluster of G5s. C'mon Steve, time to port Zilla to OS X.

    (more info: http://slashdot.org/comments.pl?sid=45647&cid=4722 113)

    --
    "Sometimes a woman is a kind of religion, she can save your soul & set you free from all your sins" - Bad Examples
    1. Re:One Word: by cyberassasin · · Score: 1

      Check this out for info on Zilla.

      http://www.apple.com/scitech/research/papers/acg /

      and

      http://developer.apple.com/hardware/ve/acgresear ch .html

      I tend to belive that there will be some developments on this front in the next release or two.... especially with the newly released distributed compile tech... I tend to think that is a testbed for some more distributed computing apps coming our way....

      --
      Who is the master of foxhounds, and who says the hunt has begun? -Pink Floyd
  66. They must be running their web server on it. by twoslice · · Score: 2, Funny

    'Cause it is surviving a /.ing with a Flash intro even!

    --

    From excellent karma to terible karma with a single +5 funny post...
  67. So how fast by earthforce_1 · · Score: 1

    Can it recompile its own kernel?

    --
    My rights don't need management.
  68. Re:Why Linux? by Anonymous Coward · · Score: 0

    Yes please.

  69. Re:Why Linux? by Anonymous Coward · · Score: 0



    You've got 29 hours and 24 minutes. Long enough?

  70. IMAGINE A BEOWOLF CLUSTER OF THESE by Anonymous Coward · · Score: 0

    ANd playing quake 1 on it

  71. FLOPS? by AvengerXP · · Score: 1

    Is that a measure of how many 1.44MB blocks it can write in a second? Or the number of bad movies/records it can create?

    --
    Trolls dont like to be Flamebait, because they burn so well. Protect our Troll heritage!
  72. And how can this be? by Anonymous Coward · · Score: 0

    For SGI is the Kwisatz Haderack!

  73. The sad thing is, by Anonymous Coward · · Score: 0

    ... most of that processing power is used for linear algebra. I.e. big honking matrix multiplication. Grunt work, and relatively uninteresting from an implementation perspective no less.

    Then again, I suppose it's a good thing that such an useful abstraction exists in the first place. Matrix manipulation is, after all, rather simple and painless if you're using the right toolkits, compared to writing the message passing or shared memory control stuff yourself...

  74. What IS a cluster, anyway? by Jhan · · Score: 1

    My gut reaction is that this isn't a cluster. A cluster is a network of independent computers collaborating on delivering some service.

    This is a parallell (super) computer. Key difference: all the processors share a single memory space with each other. Programs will run exactly as if this was a single (multitasking) computer.

    Most clusters I've worked on are just a bunch of computers with a fast network, using various protocols to synchronize their behavior ("Hey, node 19 isn't pinging, he must have died. Wasn't he running Muckatron? Who takes that job? "Me!" Ok, Joe, start up Muckatron, it's your baby now...)

    Also, does anyone know if anyone except SGI makes massively parallell unified memory computers anymore? Those Origins are some kick-ass machines...

    Does anyone have a good link to some "official" definitions of cluster and parallell computer?

    PS.

    Kids today would probably call this a "massively" parallell computer. Hah! Back in my day, we used Connection Machines with 65536 1-bit processors! All the processors had to run the same program (synchronized cycle by cycle), though the data was different at each node. We had to program it in *Lisp, the sickest programming environment then invented, but we were happy with it! We wept for joy, every morning and sometimes at lunch!

    --

    I choose to remain celibate, like my father and his father before him.

    1. Re:What IS a cluster, anyway? by TheOrquithVagrant · · Score: 1

      The Altix itself is a big 64-CPU NUMA machine, but as far as I can tell, this is a cluster of four Altixes...

    2. Re:What IS a cluster, anyway? by Jhan · · Score: 1

      this is a cluster of four Altixes...

      Yes, granted, but they still share a single memory space (and IMHO are one parallell computer):

      SGI Altix 3000 is recognized as the first Linux cluster that scales up to 64 processors within each node and the first cluster ever to allow global shared memory access across nodes.

      In other words, 64 processors in each node (read box), but globally shared memory between all the nodes.

      Sure, memory access won't be super fast across the borders (like maybe fibre channel network instead of direct optical connections between each pair of nodes), but it's the same sort of thing that happens inside each Altix. After all that's what the N in NUMA stands for! (NON uniform memory access)

      --

      I choose to remain celibate, like my father and his father before him.

    3. Re:What IS a cluster, anyway? by cr0sh · · Score: 1
      Back in my day, we used Connection Machines with 65536 1-bit processors!

      Do you (or anyone else here) know how a 1 bit processor works? I have read a bit about the CM, and always come across the "1-bit processor" thing - but never have found enough information on how a 1-bit processor really works?

      I have a feeling it is something like a single-instruction cpu (subtract then branch or something like that), but even more esoteric in practice. Furthermore, I tend to wonder how you write programs for such a machine (nevermind reducing a problem, if possible, to a parallel processing structure) - I mean, how would the real (assembler, or if need be, machine) code look like coming out of the compiler.

      Does anyone know of a site detailing this kind of information? Kinda like "parallel processing for dummies" or something (I understand the high-level concepts of PP - I just don't understand all the background)?

      --
      Reason is the Path to God - Anon
    4. Re:What IS a cluster, anyway? by Jhan · · Score: 1

      No, no, it's the data bus that's one-bit. Basically, you stream data to the processor over a serial line, either from memory or from another processor. There are a bunch of instructions. All the normal boolean operations, operations for streaming data aka NEWS, sleep-if, wake... No arithmetic operations, obviously.

      No links handy, sorry.

      --

      I choose to remain celibate, like my father and his father before him.

  75. Re:Asshat by brakk · · Score: 1

    Hi

    I just like the word "asshat"

  76. but does it fix flash typos? by Anonymous Coward · · Score: 0

    In the flash movie link, the tech specs say that "Linux is a copyright of Linux Torvalds" (emphasis mine). Is his life imitating his art?

  77. Linux == Terrorism weapon of mass destruction? by Anonymous Coward · · Score: 0

    Next on Slashdot:

    At Redmond, WA, George W. declares Linus Torvalds as a top 10 most wanted criminal because he creates an weapon of mass destruction capable of delivering 1.5 Flops of Terror. Meanwhile, Microsoft Chairman Bill Gates offers 1 million reward for the first Linux distro company switching from Linux to Windows.

  78. Re:beowulf cluster problem! by FuzzyDaddy · · Score: 1
    20 minutes. At home, on my Pentium Pro 200 running NT 4, which by all standards should be a lot slower than this Mac, the same operation would take about 2 minutes. If that.

    It knows you're charging by the hour!

    --
    It's not wasting time, I'm educating myself.
  79. Not crack by Anonymous Coward · · Score: 0

    You guys have obviously never smoked crack... I can tell you from experience that this kind of behavior is more typically associated with meth.

  80. So what? by quigonn · · Score: 1

    The Johannes Kepler University of Linz/Austria purchased an Altix 3700 (the first one in production use) in April. Here is a link (sorry, German only): http://www.news.jku.at/ARCHIVE/archivnewsroom/2003 maerz-april/newsroom/supercomputer.htm

    --
    A monkey is doing the real work for me.
  81. Re:Just wait again by Anonymous Coward · · Score: 0

    1) Buy 100 PS3's.
    2) Cluster them together mit linux
    3) ???
    4) -1, redundant!
    5) implement global memory
    6) ???
    7) 1, interesting ?!
    8) ???
    9) Profit!
    10) Invest money.
    11) ???
    12) All your base are belong to us!!!
    13) ???
    14) In Soviet Russia, Make Money Fast list makes YOU!

  82. I'm offering a Bounty to all posters by hellfire · · Score: 3, Funny

    $500 for the scalp of anyone who says the words "Beowulf" and "cluster" in the same post in response to this article.

    --

    "All great wisdom is contained in .signature files"

    1. Re:I'm offering a Bounty to all posters by lpret · · Score: 1

      Here's the scalp you want.

      --
      This is my digital signature. 10011011001
    2. Re:I'm offering a Bounty to all posters by bestguruever · · Score: 1

      can I have my $500 now or do I have to pull it out of your wallet after I scalp ya?

      --
      if you think this is bad, you should have seen my last sig
    3. Re:I'm offering a Bounty to all posters by Dr_Cornholio · · Score: 1

      I noticed that your post had the offending pair of words. I want my $500! Gimme gimme gimme!

      --
      In Soviet Russia, the monkey spanks you!
  83. Ms Might move to Hanford by ratfynk · · Score: 1

    Microsoft might decide to move to Hanford so they can compete, with giant mutated Doggy like reborn versions of Clippy.

    --
    OH THE SHAME I fell off the wagon and use sigs again!
  84. How to get 1.5 TeraFlops from Linux by Anonymous Coward · · Score: 0

    Use it to generate the special effects in Ishtar II.

  85. Re:ALMOST as good as my "Screwed Out of a Job" tro by Anonymous Coward · · Score: 0
  86. distributed shared memory by 73939133 · · Score: 1

    It seems unlikely that the SGI is the first Linux cluster with global shared memory. There are plenty of distributed shared memory systems in software, some of them open source. You can find a list here. For most computations and most hardware, you are probably still better off with MPI or PVM rather than shared memory.

    Note also that there are several high speed interconnects for Linux clusters available from many different vendors, including InfiniBand, Gigabit Ethernet, FireWire, and Myrinet.

    1. Re:distributed shared memory by green+pizza · · Score: 2, Informative

      You can find a list here. For most computations and most hardware, you are probably still better off with MPI or PVM rather than shared memory.

      Note also that there are several high speed interconnects for Linux clusters available from many different vendors, including InfiniBand, Gigabit Ethernet, FireWire, and Myrinet.


      SGI systems (Origin and Altix) have massive interconnects that hold together the single-system architecture. They're fast for shmem-type shared memory apps, but also for MPI. In fact, SGI keeps tweaking their MPI implementation with every release of IRIX and the Linux ProPack, even though MPI is not the "best" way to run apps on their systems.

      The interconnects in most Origins and Altix systems are 3.2 gigaBYTE per second with extremely low latency. I don't know about Infiniband, but I do know that GigE is only 125 MB/sec with really high latency... FireWire 800 is 100 MB/sec with better latency.... and I think the bst version of Myrinet is 500 MB/sec (4 gigabit) with about 5x the latency of SGI's 'numalink'.

      The smaller Altix systems (and supposedly, future Altix and Origin systems this fall) can be double cabled or can run at a higher speed... for 6.4 gbyte/sec per interconnect.

      Also, the Altix can handle up to 64 processors per single machine / single node (or 128 with a very beta set of patches). The cluster in the article is actually four Altix systems, each with 64 processors. The Origin 3800/3900 can handle 512 processors per node (or 1024 with a special "XXL" IRIX kernel).

      Great stuff for I/O intensive tasks, but massive overkill for 3d rendering or calculating pi.

    2. Re:distributed shared memory by 73939133 · · Score: 1

      The interconnects in most Origins and Altix systems are 3.2 gigaBYTE per second with extremely low latency.

      I think SGI is playing with the numbers there: while that is the limit for how much data you can push through a link, it seems unlikely that you get that kind of performance for arbitrary processor-to-processor communications or as aggregate bandwidth.

      In any case, very fast interconnects are not cost-efficient when they are not needed. 3.2GB is the memory bandwidth of a 400MHz bus motherboard. Most numerical applications do not spend all their time moving stuff around between different processors, and it just isn't worth paying a premium in order to get that kind of bandwidth if you don't need it.

      Also, the Altix can handle up to 64 processors per single machine / single node (or 128 with a very beta set of patches). The cluster in the article is actually four Altix systems, each with 64 processors. The Origin 3800/3900 can handle 512 processors per node (or 1024 with a special "XXL" IRIX kernel).

      Again, you pay a premium for a product that just doesn't make a lot of sense (at least to me): an abstraction of parallel computation in which 64 or 1024 processors all read and write a gigantic shared memory just isn't a good way of thinking about parallel computations, even if you could build it cheaply. And because it's NUMA, it only looks like a shared memory, but it really isn't anyway.

      I think the SGIs have a niche: you can take dusty FORTRAN and multi-threaded C applications and scale them up to many processors with a minimum of fuss. That's very valuable, even if you pay a premium for the convenience. But I sure hope that the future of parallel programming won't look like that.

  87. Yes by Anonymous Coward · · Score: 1, Funny

    You're missing something.

  88. Great but does it run by sydres · · Score: 1

    Doom III?

    1. Re:Great but does it run by Mr.+Sketch · · Score: 1

      Actually a better question would be if it gets at least 30fps in DoomIII, after all, my 1GHz will probably _run_ it, but it would look like a slide show.

  89. problem solved by Anonymous Coward · · Score: 0

    I have one teraflop in my pants

  90. Java implementation by Anonymous Coward · · Score: 0

    At that time they were working on a Java implementation, but I don't know what happened with that.

    It's still loading.

    1. Re:Java implementation by BigBadBri · · Score: 1
      Nah - it'll only take one VM per node, won't it?

      Just so long as you only need to run one Java process, that is...

      --
      oh brave new world, that has such people in it!
  91. Re:Asshat by Alan+Partridge · · Score: 1

    Can you tell me what it means? Some kind of a hat for your Ass, presumably. My Donkey seems to get by without a hat, why would an Ass need one?

    --
    That was classic intercourse!
  92. Cluster??? by Anonymous Coward · · Score: 1, Insightful

    first cluster ever to allow global shared memory access across nodes...

    In other words, its not a cluster but a big shared memory system running a single OS image, just like their Origin 3000s. They just call it a cluster because that's the big buzzword in the Linux world.

  93. Re:First one... by JudgeFurious · · Score: 1

    Figures. Mod this one Offtopic too while you're at it.

    --
    Appended to the end of comments you post. 120 chars.
  94. Re:Asshat by Anonymous Coward · · Score: 0

    There's no telling where some words come from. Like, for instance, who was the genius that one day said "I'm going to invent the word 'fuck' just so people can scream it at each other while driving down the highway."? Well, for the most part, I'm guessing we will never know how some words came into origin. But I have come up with a clearly outlined history of the greatest word ever, along with definitions and a useful synonyms list.

    Part One: The Origins and Spread to the United States

    From what I can trace back from archived IRC chats, the term 'asshat' was first used in the large European country of Sweden as an alternative for the word 'asshole.' From Sweden, by way of both train and ship, the word found it's way to the shores of western France, where it was used by a software coder named Louis (LOO-ee) in a telephone conversation with his second cousin in Newfoundland. This great man was responsible for sending this fashionable European word to North America for all children to enjoy!

    Once on the North American continent, there was no stopping it's spread to the states. In a June 2000 memo to US-Canadian Customs agents, they were warned to not let this word slip across the border. At this point the word had such a Canadian backing that there was no way to stop it's glorious spread. And so, 'asshat' was first used in the US by an Inuit man named Pukkeenegak who slipped across the border in his kayak. God bless the Inuit!

    Americans were, at first, wary to this new word. They had grown up using 'asshole,' 'assclown,' (premier in Office Space) and even 'asshead.' What was to force them to add yet another word to their vocabulary? The Internet, of course! Using radiocarbon dating, we have determined that some of the first widespread usage of the word 'asshat' was in several Kevin Smith fan newsgroups, all located in California. Luckily, there are also Kevin Smith fans in the northeast United States, so the word quickly bounded an entire continent and made it to New York City.

    Everyone knows that New York City is a melting pot of culture, drugs, and comedians. The Bamboni family of Brooklyn, NY was the first to use the word 'asshat' in common day talk. Donny Bamboni was quoted in July 2001 as telling a cashier at 7-11 to, "put the money in the bag, asshat!" Several scared patrons of this convenience store quickly jumped on planes and flew back home, carrying the word with them. And Donny took the word to Rikers Island, where all the prisoners began using the word in a different context than an insult.

    It is feasible to guess that by November, 2001, the word asshat was now in full usage all around the United States. Secretary of State Collin Powell was quoth in early 2002 as saying the following at a State Dinner: (in regards to Usama bin Laden) "We have not yet found that asshat [bin Laden], but we're sure as hell trying."

    In terms of an age breakdown, it seems that colleges students between the ages of 18-25 most commonly use the word asshat, but children as young as six have been heard using the word. Current demography predicts that by the end of 2003, the word asshat will be a commonly used word among high school students and among the 35-35 year old working class. It also appears that, through misinformation, the word is loosing it's set definition. The next section will discuss its proper usages.

    Part Two: Using Asshat Correctly

    Asshat is by no means just a replacement for the word 'asshole,' but at the current state of the economy there is no better word to compare it to.

    asshat
    n. slang
    1. A thoroughly contemptible, detestable person.
    2. Tight fitting underwear
    [sny: asshole, asshead, assclown, asswad, asshot, assface, prick, faggot, whitey tighties]

    Part Three: Why Use Asshat?

    Quite frankly, there is no better insult in the world than the word asshat. Think about it.

    Scenerio 1: You are once again late for work, and the boss decides to chew you out about it. So he fin

  95. Makes sense. by MattGWU · · Score: 1

    Why the hell not? You have desert maps (some of my favorite), and the oil rig. Might as well find where the coders hid the sweet, sweet oil. You can use the seismic waves put out by the HE grenades, or the terrorist bombs.

    --
    "These people look deep within my soul and assign me a number based on the order in which I joined" --Homer re:
  96. Yep. Here. by Anonymous Coward · · Score: 1, Informative
  97. Mod parent up by Anonymous Coward · · Score: 0
    gotta be a +5:Funny

    Unless there's a parallel compiler somewhere out there...

    1. Re:Mod parent up by Anonymous Coward · · Score: 0

      Unless there's a parallel compiler somewhere out there...

      Well, apart from the obvious fact that there IS, dipshit, have you ever heard of this little thing called MAKE? See, it's neat. It runs ONE TASK on EACH PROCESSOR in your computer.

      Fuckin dipwit.

    2. Re:Mod parent up by arkane1234 · · Score: 1

      make -j13 bzImage :)

      --
      -- This space for lease, low setup fee, inquire within!
  98. Re:Asshat by WTFmonkey · · Score: 1

    I asked about this yesterday or the day before. See here

  99. must be cheap by danZenie · · Score: 1

    i think i'm gonna buy me one next week. does it come with XP also?

    --
    You need people like me so you can point your fuckin fingers and say, "That's the bad guy." So what that make you? Good?
  100. Re:ALMOST as good as my "Screwed Out of a Job" tro by Anonymous Coward · · Score: 0

    pfff. rather weak.

  101. Re:MMMM....Quake server by flyneye · · Score: 1

    what moron would call the desire to run this computer as a quake server offtopic?
    sadly the mentally inferior and unimaginative walk among us because we havent learned to feed on them yet.

    --
    *Repent!Quit Your Job!Slack Off!The World Ends Tomorrow and You May Die!
  102. Obsolesence at last. by skvngrx · · Score: 1

    Great... so can I finally get rid of my Indigo2?

  103. Even this guy? by corkhead0 · · Score: 0

    http://science.slashdot.org/comments.pl?sid=70313& threshold=2&commentsort=0&tid=106&mode=thread&pid= 6392800#6393985 ;)

  104. How to get 2+ TeraFlops from Linux by tobiashm · · Score: 2, Informative

    This does not seem to have been mentioned before:
    Niflheim at Danish University of Technology

  105. Hammer Time! by Anonymous Coward · · Score: 0

    Would a similar configured, AMD Hammer based computer give better specs?