Slashdot Mirror


Cell-based "Roadrunner" Tops Elusive Petaflop Mark

prunedude writes "The NY times is reporting that an American military supercomputer, assembled from components originally designed for video game machines, is more than twice as fast as the previous fastest supercomputer, the I.B.M. BlueGene/L. To put the performance of the machine in perspective, Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day."

269 comments

  1. Re:Ummm by francium+de+neobie · · Score: 2, Funny

    They're trying to pull 1000 times your lab's results.

  2. Summary should have a shout out by HolyCoitus · · Score: 2, Informative

    1350 IBM Linux cluster team. xCAT for pwning.

    --
    That's scary.
    1. Re:Summary should have a shout out by Thalin · · Score: 1

      Yay for 1350. Go us. :)

      --
      What? You want a sig?
    2. Re:Summary should have a shout out by b4upoo · · Score: 1

      It has been obvious that the Cell processor can be really potent when combined with other Cell processors. What bothers me is that we haven't been seeing PCs made with a few Cell processors at their core. Expense isn't much of an issue with the Cell units.
              There is also a wicked fast computer built with a few graphics cards that recently made news. Apparently graphics cards can process data faster then conventional CPUs.
                So why aren't such things easily available?

    3. Re:Summary should have a shout out by cnettel · · Score: 1
      Graphics cards are easily available. The PCI Express interface is fast enough in both directions.

      Code is not available for any task, though.

    4. Re:Summary should have a shout out by DrEldarion · · Score: 4, Interesting

      From what I've heard (I'm no expert), these processors are good at certain types of calculations, but horrible at others. Ask a Cell to run Folding@Home, and it'll be blazing. Ask it to do general-purpose computing, though, and you'll quickly see the strength of other processors.

    5. Re:Summary should have a shout out by jacquesm · · Score: 4, Interesting

      The biggest problem is not to acquire the hardware (the graphics boards can be bought easily enough and some manufacturers make motherboards that will host 4 of them), the problem is in the software.

      As you probably know your current desktop computer executes software 'in parallel' on multiple cores. Unless that software is written to use multiple cores then the parallelization will probably take place at the task level, in other words, unless special care was taken during the development each core will be executing a different process (or parts of a process, slicing itself up between multiple processes as long as the number of tasks is greater than the number of cpus).

      The cell processor and the graphics cards you are talking about (I assume you refer to the nvidia 'cuda' architecture) are not your ordinary processor (though the cell comes a lot closer). They need special software to get the maximum out of their parallel features, which means that you can only exploit that speed when you are trying to solve a particular class of problems.

      The issue is that not every problem is easily adapted to a parallel solution, and for some classes of problems it's simply impossible. For general purpose computing a general purpose cpu (with or without multiple cores) is good enough.

      If you have a problem that you can solve in a parallel way (say computing chess moves, analyzing genetic strings, ray tracing an image or fluid dynamics, add your own examples here) then it makes sense to invest the extra time to code the problem up in such a way that if you have lots of cpus that they can each work on a small part of the problem.

      The graphics cards are closer to a traditional vector processor than the cell, and thus even less suitable to be used as general purpose cpus.

    6. Re:Summary should have a shout out by chrysrobyn · · Score: 4, Informative

      From what I've heard (I'm no expert), these processors are good at certain types of calculations, but horrible at others. Ask a Cell to run Folding@Home, and it'll be blazing. Ask it to do general-purpose computing, though, and you'll quickly see the strength of other processors.

      You're precisely correct. Cell's strength is in very predictable workloads (ones it can perform without branch mispredict penalties), very parallelizable workloads (ones that can be distributed over 6-8 SPU's / SPC's) that fit within 256 KB of local storage per SPU (manually managed cache, mapped to main memory). The non-double precision floating point enhanced version's (the version in the PS3) strength is further limited to integer and single precision floating point workloads. Roadrunner's Cell-DP eliminates that last limitation. While video games, encryption, nuke simulations and anything else that involves matrix manipulation can really stretch their legs on such a beast, general purpose computing won't find a benefit.

    7. Re:Summary should have a shout out by Anonymous Coward · · Score: 0

      i HATE xcat! there are far better solutions out there than that cobbled together piece of admin annoyance.

      On an unrelated note, props to Don Grice for getting a pic in the NY Times!

    8. Re:Summary should have a shout out by adisakp · · Score: 2, Insightful

      The non-double precision floating point enhanced version's (the version in the PS3) strength is further limited to integer and single precision floating point workloads.

      BTW, the original Cell can do Double Precision in hardware. The big limitation it had was the DP was not PIPELINED so all DP instructions caused huge stalls in processing. You can use DP on the PS3 just fine and it's still fairly fast (especially compared to software DP) -- it's just not nearly as fast as SP.

      The Cell's double precision hardware attains a very respectable 25 Gigaflops per second (peak), but its single precision performance is a phenomenal 256 Gigaflops (also peak).

      The main new features of the PowerXCell 8i Processor are that DP is now fully pipelined and can attain over 100 GFLOPS (about a 5X improvement in DP execution due to stall removal) and that the memory interface now supports industry standard DDR2 memory so 16 GB of RAM per Cell can be used. The memory limitation with XDR was just as bad as the DP in limiting more common use of the Cell since XDR is expensive and hard to come by.

    9. Re:Summary should have a shout out by HolyCoitus · · Score: 1

      Check out 2.0 and let me know. If there is a problem, we can fix it.

      --
      That's scary.
  3. But can it run.... by CastrTroy · · Score: 3, Funny

    By can it run Crysis?

    --

    Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    1. Re:But can it run.... by Anonymous Coward · · Score: 2, Insightful

      Why would anyone want to?

      Crysis is a miserably boring game.

    2. Re:But can it run.... by setagllib · · Score: 3, Funny

      The marketing for that game was done all wrong. Insiders report that it was meant as an interactive preview of 3DMark2020.

      --
      Sam ty sig.
    3. Re:But can it run.... by Warll · · Score: 0, Offtopic

      Crysis is a miserably boring game. It is? Too bad no one told me that before I had plenty of fun playing through it.
    4. Re:But can it run.... by Anonymous Coward · · Score: 0

      If you had fun playing Crysis, you gotta try this game:

      http://en.wikipedia.org/wiki/Paddleball

      It should absolutely blow you away. And you won't have to waste hundreds to thousands of dollars on you 'rig' to do so...

      Rig, LOL, what a bunch of losers.

    5. Re:But can it run.... by beav007 · · Score: 3, Funny

      Can it even run Vista Ultimate?

    6. Re:But can it run.... by nawcom · · Score: 4, Funny

      Hah you think the military is dumb enough to even install Vista? Haven't you ever heard of military intelli... errm.. nevermind. Yeah, I'm sure they have it installed already.

    7. Re:But can it run.... by Warll · · Score: 3, Funny

      Mah I'm a PC gamer myself, haven't used a handheld in years.

    8. Re:But can it run.... by voidcoder · · Score: 1

      By can it run Crysis? yes, but in LOW settings only...
    9. Re:But can it run.... by SanityInAnarchy · · Score: 1

      Nah, really should've been sold as Doom 4. ...which would be about the same thing. Never mind.

      --
      Don't thank God, thank a doctor!
    10. Re:But can it run.... by Hojima · · Score: 4, Funny

      Windows: End Program - Untitled - Nuclear Defense. This program is not responding. To return to Windows and check the status of the program, click Cancel. If you choose to end the program immediately, you will loose any unprotected civilians. To end the program now, click End Now. Army personnel: Sir, I think we should send an error report.

    11. Re:But can it run.... by x2A · · Score: 1

      Know how we can tell that's made up and not a real error message?

      Microsoft knows how many o's there are in 'lose'.

      Completely ruined it's believability.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    12. Re:But can it run.... by pha7boy · · Score: 3, Funny

      military supercomputer, assembled from components originally designed for video game machines yes. yes it can. and, most likely, it will also blend.
      --
      -- All this knowledge is giving me a raging brainer.
    13. Re:But can it run.... by PhasmatisApparatus · · Score: 1

      With that kind of processing power, it could likely run many instances of Crysis without any sort of graphics adapter. A GeForce 8800 (which can easily run Crysis on medium settings, and can run high settings with intermittent slowdowns), for example, has somewhere around 500-600 gigaflops. Meanwhile a core 2 quad is somewhere around 30-40 gigaflops (but is obviously much faster than the typical 500mhz GPU). Using one of the aforementioned hand calculators, we can determine that the Roadrunner could probably run over a thousand copies of Crysis. We can also determine that I am bored, although that determination does not require a calculator.

    14. Re:But can it run.... by Anonymous Coward · · Score: 1, Informative

      The DoD Like all large corporate entities refuses to install Vista.

    15. Re:But can it run.... by asc99c · · Score: 2, Funny

      I was just assuming while the nuclear defenses were down, the unprotected civilians were being catapulted to safety. Or certain death for the ones not wearing parachutes.

    16. Re:But can it run.... by CastrTroy · · Score: 2, Insightful

      But would any of them be full speed. By your argument, based solely on flops, you could get a cluster of pentium II's to run Crysis. Given enough of them, you could get enough instructions to run many instances of it. However, could you ever accomplish the task of displaying it full screen, with a high frame rate? The game only runs as many threads as it is programmed to run. You can only parallelize it out to as many processors as you have threads. And then the distribution and gathering of the data streams in order to put the actual game on the screen would be so slow, that the game would be unplayable. FLOPS is not the only measure of speed that is important.

      --

      Anthropic principle: We see the universe the way it is because if it were different we would not be here to see it.
    17. Re:But can it run.... by Kgosi+Makwati · · Score: 1

      No. However, it can run through an infinite loop in less than a second!

    18. Re:But can it run.... by Anonymous Coward · · Score: 0

      Know how we can tell this was not a actual pedant's comment, but merely a made up one?

      It's.

    19. Re:But can it run.... by x2A · · Score: 1

      Oh yeah, well done... plus of course the absurdity of implying that the only way someone could tell that error message wasn't a genuine windows error message was because of the spelling of a word, but the whole nuclear defense bit, yeah that's totally believeable. You missed that bit.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
  4. Question by SeeSp0tRun · · Score: 0, Troll

    What exactly would the military use a supercomputer for? Being a pessimist, the only thing I can really think of is the air force doing the obvious shady things that it does. But there has to be some statistical purpose for such a beastly machine.

    --
    Something witty.
    1. Re:Question by avalys · · Score: 4, Informative

      It will be used for nuclear weapons simulations - primarily for investigating issues related to how warheads will perform as they age.

      --
      This space intentionally left blank.
    2. Re:Question by attemptedgoalie · · Score: 3, Insightful

      Lets say you have designed a nuclear weapon.

      Wouldn't it be really neat to run some tests before you build it?

      For instance, how cool would it be to have a simulation that could test a weapon being mishandled, or shot. At every single point from every possible angle at every possible velocity?

      It would be nice to know that there is a possibility of detonation if it were to drop off of a loading rack.

      --
      My mom says I'm cool.
    3. Re:Question by Brett+Buck · · Score: 1

      "shady things" ? What in the world are you talking about?

                Brett

    4. Re:Question by anaesthetica · · Score: 5, Funny

      What exactly would the military use a supercomputer for?

      The military will use this advanced technology to assist and perhaps automate the RTFA process, also known as Reading The Fucking Article, which would allow you to answer your query without posting.

    5. Re:Question by Anthony+Rosequist · · Score: 5, Informative

      Before it does weapons simulations, it will first work on some scientific problems, like model testing to predict climate change.

      After it's done with that (I wonder how they will determine what done is...), it will go classified and do nuke simulations.

    6. Re:Question by mikael · · Score: 1

      They can also use a system like that to simulate the response of the populations of other countries to particular events.

      DARPA Files Patent On Predictive Simulation

      Modeling Urban Panic

      --
      Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
    7. Re:Question by ScrewMaster · · Score: 3, Insightful

      the only thing I can really think of is the air force doing the obvious shady things that it does.

      Uh ... what exactly do you mean by "shady things"? If you have a problem with what our armed forces are doing, you'd be better off leveling your charges at Congress. Ultimately, they're the ones that fund any "shady" things the military does.

      --
      The higher the technology, the sharper that two-edged sword.
    8. Re:Question by snl2587 · · Score: 1

      Uh...codebreaking?

    9. Re:Question by DigiShaman · · Score: 0, Troll

      You know what they say. Garbage in garbage out.

      I've heard rumors about how the data is collected for global temperatures. For example, some thermometers have been found on the roof tops of building coated with black tar while others close by a heat exchanger. But naw, that couldn't skew the results could it?

      --
      Life is not for the lazy.
    10. Re:Question by TechyImmigrant · · Score: 1

      >You obviously know nothing about nuclear weapons.

      So nobody has to shoot him.

      --
      Evil people are out to get you.
    11. Re:Question by Anonymous Coward · · Score: 3, Funny

      And I've heard that republicans eat babies. As someone who has worked with climate models including data collection I can safely say you're full of shit. There are thousands of research stations collecting the data. For it to be generally corrupted, there'd have to be some vast global conspiracy whereby publically competing research stations and countries agree to privately skew their data.

      Now there IS something of a vast global conspiracy (PNAC, Republicans, Bilderberg, etc), but, er, it's not on the pro-environmental-sanity side.

      FWIW, if anything, the climate change stuff you usually see is an underestimate. 8-(

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

      If all 6 billion people on earth sat down and used hand calculators to do anything, we'd have world peace and no need for this computer. But it was paid for by the Bush Administration, so this is not a surprise.

    13. Re:Question by Anonymous Coward · · Score: 1, Insightful

      And I've heard that republicans eat babies. As someone who has worked with climate models including data collection I can safely say you're full of shit. There are thousands of research stations collecting the data. For it to be generally corrupted, there'd have to be some vast global conspiracy whereby publically competing research stations and countries agree to privately skew their data. Or, you could create a system for funding research that provides financial rewards to over-the-top alarmism.

      That'd work, too.
    14. Re:Question by Anonymous Coward · · Score: 1, Informative

      The military didn't build Roadrunner. The U.S. Department of Energy built it, one in a long line of supercomputers used for (in addition to many other things) simulations to evaluate the reliability of the nation's aging nuclear weapons stockpile.

    15. Re:Question by Brian+Gordon · · Score: 3, Funny

      Quick, make it play tic tac toe against itself.

    16. Re:Question by kylehase · · Score: 5, Funny

      Perhaps they should invest in a computer to track warhead parts.

      --
      You want fun, go home and buy a monkey!
    17. Re:Question by Kirkoff · · Score: 2, Interesting

      I understand that you really don't like certain politicians and certain stances and ideas. I have to wonder though, can you really say that a major political party be considered a conspiracy? This is a political party that holds open elections, has existed since the 1800's and has 55 million members (roughly 1/6 the population of the US) in all 50 states. It doesn't seem to really make great argument for conspiracy.

      --
      There are exactly 42,935,718 letter sized sheets in a square mile.
    18. Re:Question by bersl2 · · Score: 2, Insightful

      You'd need a very large number of abnormalities to make a significant difference, unless there is a disappointingly low number of measuring devices, which I find unlikely. But by all means, if you can find a systematic skew in the results that reinforces itself (rather than cancels itself out by erring in both directions randomly), please present this evidence; accuracy of evidence is a legitimate issue.

    19. Re:Question by bpkiwi · · Score: 3, Insightful

      Yes, but it would only work if you made that funding available to everyone in the world, or else the data out of wherever it wasn't available would conflict with where the funding was available.

      In fact you could analyse such systemic bias by comparing data sets from varying geopolitical areas (say comparing Russian data to US data).

    20. Re:Question by Anonymous Coward · · Score: 0, Informative

      Perhaps, but that's not what we have. In fact, funding bodies for the most part try to "encourage" skewing the data downward, particularly in the corporate reich of america.

      And don't forget it's pretty much the same data as goes towards the weather forecasts. Those things on TV? Sure, they're not always right. But they're pretty good nowadays. You can't have it both ways, like some creationist happily microwaving his dinner.

    21. Re:Question by ediron2 · · Score: 2, Funny

      What exactly would the military use a supercomputer for?

      They could tell you, but then they'd have to kill you.
    22. Re:Question by ediron2 · · Score: 1

      Sorry, didn't mean to reply directly to Anaesthetica, but to grandparent. Wasn't logged in when I got snarky, then the new (stupid!) ajax login framework fried my reply and dumped me back at the /. base page. Pissed me off and I was preoccupied with finding my old spot rather than making sure I was replying to the right level.

    23. Re:Question by Anonymous Coward · · Score: 1, Informative

      If a dozen nukes dropped off their loading racks and exploded before being launched, the world would become a much better place.

    24. Re:Question by Chibi+Merrow · · Score: 0

      Ahh those ignorant of history and all that... Considering modern electronic computers got their start as military projects (Colossus MkI/MkII and ENIAC) and that the Internet came out of a military research agency--just to give two examples--is it any surprised you were modded as a Troll?

      As others have pointed out, the US military makes heavy use of supercomputers to do nuclear weapons simulation so that they can determine how weapons perform as they age without rolling them out into the desert and blowing them up, which is the way they used to get that data. That's a pretty good use of resources in most people's opinions, as the alternative is unpleasant. There are plenty of other uses for these machines, of course, but the USAF is not in the business of doing "shady" things with them. The NSA, maybe, but you won't hear press releases about those... :P

      --
      Maxim: People cannot follow directions.
      Increases in truth directly with the length of time spent explaining them
    25. Re:Question by Anonymous Coward · · Score: 0

      For instance, how cool would it be to have a simulation that could test a weapon being mishandled, or shot. At every single point from every possible angle at every possible velocity?

      It would be nice to know that there is a possibility of detonation if it were to drop off of a loading rack. Anything with enough kinetic energy to pierce the shielding around a nuclear warhead is going to turn it into a large radioactive paperweight.
      A lot of effort went into designing the triggering mechanisms for nuclear weapons and they're all designed with a 'don't fire' failure mode.
    26. Re:Question by Anonymous Coward · · Score: 0
      Y'know, "shady" things. Like funding Jaffa rebellions and trying to defeat Anubis.

      Oh, wait, Congress^H^H^H^H^H^H^Hthe SciFi channel killed off that program. Never mind, it's not really that shady anymore.

    27. Re:Question by x2A · · Score: 0

      Hmm wonder if countries could use that method of throwing off the scent...

      "We are doing computer simulations to see how much butter it takes to set off a nuclear reaction"

      "Yeah, whatever you say Iran... (hehehe butter! They know nothing!)"

      "(Fools!)"

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    28. Re:Question by x2A · · Score: 1

      But then how would the people who did RTFA know what people who didn't RTFA are wondering?

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    29. Re:Question by x2A · · Score: 3, Funny

      The previous model used hundreds of dual core P4s, just running NOP's at full speed. The heat generated, being equivelant to that outputted by a nuke, meant they could run simulations without having to actually write any code.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    30. Re:Question by flayzernax · · Score: 1

      The military wants to ask it the answer to "the question".

      42

    31. Re:Question by attemptedgoalie · · Score: 1

      You're probably right.

      The supercomputers that were doing that kind of testing that I saw were really just eye candy.

      I can't believe I fell for it.

      I'm just glad somebody posted anonymously to prove me wrong.

      --
      My mom says I'm cool.
    32. Re:Question by GregPK · · Score: 1

      do realize that data like that will have an average temperature over time. Global warming is a periodic thing. So if the average temperature last year was 88 degrees and this year it is 90 degrees the data would show a moving average of 89 degrees over those two years. Or it would show a temp change of 2 degrees. Plus, I don't think they'd use data from a new setup unless it was very close to another device in someplace like death valley.

      Even then, its all about averages.

    33. Re:Question by aproposofwhat · · Score: 1
      Though such an analysis would inevitably be flawed if it did not take into account the location of the measurement stations - in the US, a lot of the stations are in urban areas, which skews the data upwards, while in Russia the stations are far more representative of the climate as a whole.

      I'd hazard a guess that over 50% of the reported rise from the US is due to increased energy utilisation in populated areas - there is some anthropogenic warming, but the US figures vastly overreport the effect.

      --
      One swallow does not a fellatrix make
    34. Re:Question by Anonymous Coward · · Score: 0

      previous AC is abrupt but correct, nuclear weapons do not detonate from being dropped.

    35. Re:Question by bestinshow · · Score: 1

      Lets say you have designed a nuclear weapon.

      Wouldn't it be really neat to run some tests before you build it? We're software engineers here ... testing is something we do when it screws up during actual deployment!
    36. Re:Question by Elbowgeek · · Score: 1

      Isn't that what we have Iranians for?

      --
      Who is this delectable creature with an insatiable love of the dead?
    37. Re:Question by neomunk · · Score: 1

      That would just lead some of the more reactionary citizens of the U.S. to call for an embargo on dairy to Iran. Some of the more extreme ones will even start claiming cows as proof of Iranian WMDs. Unfortunately for us, many of these people seem to be in positions capable of influencing policy, if the Iraqi adventure is any indication.

    38. Re:Question by k1773re7f · · Score: 1

      And don't forget it's pretty much the same data as goes towards the weather forecasts. Those things on TV? sure, they're not always right. But they're pretty good nowadays.
      Yeah! They can predict current conditions with 80% accuracy!
      --
      This sig. intentionally left blank.
    39. Re:Question by halcyon1234 · · Score: 1

      What exactly would the military use a supercomputer for?

      Military shit. Move along, citizen.

    40. Re:Question by Anonymous Coward · · Score: 0

      Nuke simulations aren't scientific problems?

    41. Re:Question by RockDoctor · · Score: 1

      What exactly would the military use a supercomputer for?

      Military shit. Move along, citizen.

      s/citizen/civilian/

      Hold it there, civilian! Just why are you asking? Put these handcuffs on and come along with us to this cellar in an anonymous building. Or we'll hit you with the next-generation Tazer. We designed it using this machine, for longer battery life and better interaction with your nervous system. You won't die of a heart attack, but you will feel like you're having one.
      --
      Birds are not dinosaur descendants;birds are dinosaurs, for all useful meanings of "birds", "are" and "dinosaurs"
    42. Re:Question by JustOK · · Score: 1

      The real danger, 'tho, would be from goat cheese (not going to link to the site...)

      --
      rewriting history since 2109
    43. Re:Question by Anonymous Coward · · Score: 0

      It would be nice to know that there is a possibility of detonation if it were to drop off of a loading rack. Hey wait...I just watched that episode of Battlestar Galactica!!
  5. exaflop, zettaflop, the yottaflop and the xeraflop by anaesthetica · · Score: 3, Insightful
    Military taking the lead on computing as usual. Why is the military so much more progressive (with practical results) than any other institution of government?

    It will be used principally to solve classified military problems to ensure that the nation's stockpile of nuclear weapons will continue to work correctly as they age. The Roadrunner will simulate the behavior of the weapons in the first fraction of a second during an explosion. Before it is placed in a classified environment, it will also be used to explore scientific problems like climate change.
    So, it also has Cell-based processors AND Opterons. I wonder what the functional division between the two chip types is?

    "If Chevy wins the Daytona 500, they try to convince you the Chevy Malibu you're driving will benefit from this," said Steve Wallach, a supercomputer designer who is chief scientist of Convey Computer, a start-up firm based in Richardson, Tex. Those who work with weapons might not have much to offer the video gamers of the world, he suggested.

    Who cares? It's awesome sui generis.

  6. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    > Before it is placed in a classified environment, it will also be used to explore scientific problems like climate change.

    Great. It will be the first classified nuclear simulation to be infected with the Storm worm.

  7. The future by Renraku · · Score: 1

    Who didn't see this coming from Cell?

    --
    Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
    1. Re:The future by the_humeister · · Score: 0, Troll

      I certainly didn't, especially when Sony's proclamation of the PS2 being a supercomputer resulted in no actual emotion engine based system other than the PS2 being produced.

    2. Re:The future by Thalin · · Score: 5, Informative

      This is actually based on Cell 2 or as IBM marketing likes to say it "Cell eXtreme"!

      Cell 1 (the Playstation chip) didn't have the double precision floating performance to achieve the petaflop mark; Cell 2 is far better on that front.

      --
      What? You want a sig?
  8. Yes, but... by parsnip11 · · Score: 1

    ...what if you had a beowulf cluster of these?

    1. Re:Yes, but... by thatskinnyguy · · Score: 2, Interesting

      That might be just the standard meme here, but this time it really got me to think. Having run a Beowulf cluster in the past, what would one use to connect all of these supercomputer nodes together to make a more massive computer? Cat5 just won't cut it in this situation. I have to wonder if fiber could even keep up.

      --
      The game.
    2. Re:Yes, but... by JasterBobaMereel · · Score: 1

      I take your Beowulf cluster meme and raise you a RTFA

      It is already a massively parallel computer cluster (as are the BlueGenes and the vast majority of the fastest computers)

      --
      Puteulanus fenestra mortis
  9. So what is it for? by sweet_petunias_full_ · · Score: 1

    "if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day."

    This contraption makes lots of people really, really, tired of punching on calculators?

    --
    You can't send a takedown notice to an already printed newspaper.
    1. Re:So what is it for? by kylehase · · Score: 1

      Wow, Roadrunner is going to put a lot of calculator operators out of work.

      --
      You want fun, go home and buy a monkey!
  10. Re:exaflop, zettaflop, the yottaflop and the xeraf by Renraku · · Score: 4, Insightful

    The military is more progressive because there's not a whole lot they can do to advance things.

    They can hope for random breakthroughs, mostly based on chance/luck/etc..

    Or they can follow the natural progression of things. If you want to make things explode you have to know the nature of the explosion. And to know the nature of explosions you have to know all about high-energy physics at a molecular level. And to know about high-energy physics you have to know about how molecules and atoms interact. Now, with all of these things you can either make them yourself and study the real explosion, or you can simulate it and confirm with real-world results..which is what they're doing.

    They have the resources AND the desire to do so, and therefore, they are doing so. Private industries will rarely do things like this on their own. They're much more likely to wait for someone else to do the research, or research with grants and then patent the results for their own profit. Its the same reason NASA has spurred many developments and improvements in the rest of the civilian world.

    This setup will make it easier to study weather, physics, etc, etc. On the other hand, it'll also make it easier to figure out how to make bigger sticks that are lighter and sharper.

    --
    Job? I don't have time to get a job! Who will sit around and bitch about being broke and unemployed then?
  11. Re:exaflop, zettaflop, the yottaflop and the xeraf by the_humeister · · Score: 1

    Military taking the lead on computing as usual. Why is the military so much more progressive (with practical results) than any other institution of government?
    I wouldn't say "as usual." The prior computer at the top for more than 2 years was at the Lawrence Livermore National Laboratories.
  12. Change in paradigm by karvind · · Score: 5, Informative

    If one looks at http://www.top500.org/ list and compare the CPU frequencies of the top supercomputers - all BlueGene CPUs were running at less than a GHz. And it seemed those low power cores were key to HPC (high performance computing). Cell and opteron - both run at multiple GHz and (presumably consume more power). IBM still has next generation of BlueGene/Q in works and is also for +Petaflop computation.

    1. Re:Change in paradigm by the_humeister · · Score: 1

      The Bluegene/L uses a PowerPC 440 chip. However these chips don't include an FPU in the core, unless IBM made special ones with an FPU or they're using a separate FPU chip on the processor boards. I think it's interesting their Linpack score/processor is on par with the #3 spot consisting of quad-core Xeons.

    2. Re:Change in paradigm by TubeSteak · · Score: 2, Informative

      all BlueGene CPUs were running at less than a GHz. And it seemed those low power cores were key to HPC (high performance computing). Supercomputing is on its way to a water cooled infrastrucure.
      IBM is already selling a product under the name bluefire
      http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=207100873
      http://www.eweek.com/c/a/Infrastructure/IBM-Ships-First-WaterCooled-Supercomputer/

      I hope we see more water & less air in the future
      --
      [Fuck Beta]
      o0t!
    3. Re:Change in paradigm by Anonymous Coward · · Score: 0

      The BlueGene/L system has 3328 cards with 106496 processors. For theoretical performance, this is 596TF at 1.81MW (330MF/W).

      In comparison, the Cell systems achieves 'better than a petaflop' at 'roughly 3 megawatts', so thats around 333MF/W.

  13. I feel bad for Whyle E. ... by CaptainPatent · · Score: 4, Funny

    ...There's no catapult in the world that will catch THAT roadrunner!

    --
    Well, back to rejecting software patent applications.
    1. Re:I feel bad for Whyle E. ... by Anonymous Coward · · Score: 0

      That's why ACME invented rocket skates.

  14. But my pocket calculator by Anonymous Coward · · Score: 0

    is a macbook pro!

    1. Re:But my pocket calculator by Anonymous Coward · · Score: 0

      So... much... trash talk... can't... resist...

      You must have clown pants if you can fit a macbook pro in your pocket.

      Is that a macbook pro in your pocket or are you just happy to see me.

      Better get some asbestos underwear; I hear those macbooks get pretty hot.

  15. Better comparison by Clarious · · Score: 1

    Can we have a better comparison? Everyone know that computer caculate much faster than a man with a caculator.

    1. Re:Better comparison by Anonymous Coward · · Score: 0

      How's about a monkey with a typewriter? "Roadrunner can write the entire works of Shakespeare faster than an infinite number of monkeys typing for an infinite amount of time." Yea, that's better.

    2. Re:Better comparison by zippthorne · · Score: 1

      If you have an infinite number of monkeys, it only takes you one keystroke in time. You only need the infinite amount of time if you don't have enough monkeys.

      --
      Can you be Even More Awesome?!
    3. Re:Better comparison by Lord+Lode · · Score: 1

      Maybe this is one of the American units of measurements? Yards, feet, number-of-men-with-calculator-years?

  16. so what else is new? by WheresMyDingo · · Score: 5, Funny
    the roadrunner always wins, so it no surprise it topped this "petaflop mark" guy (yeesh, what a name).

    and roadrunner's always been cel-based, at least in the modern era. i bought one of those cels from the warner bros. store before they went under, nice one too with his tongue sticking out

    1. Re:so what else is new? by Anonymous Coward · · Score: 1, Funny

      you suck
      your sense of humour is extremely weak
      fuck off

    2. Re:so what else is new? by Anonymous Coward · · Score: 1, Insightful

      your sense of humour is extremely weak I like how you spelt humour. But apart from that you have a total lack of it.
  17. Re:exaflop, zettaflop, the yottaflop and the xeraf by DogDude · · Score: 5, Informative

    Military taking the lead on computing as usual. Why is the military so much more progressive (with practical results) than any other institution of government?

    Are you kidding?

    --
    I don't respond to AC's.
  18. Re:exaflop, zettaflop, the yottaflop and the xeraf by students · · Score: 1

    Livermore uses their Blue Genie/L for mostly the same thing. They are responsible for the country's nuclear technology. The upcoming Blue Genie/P will also do weapons simulations, among other things.

  19. Perspective? by Bob54321 · · Score: 5, Funny

    To put the performance of the machine in perspective, Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day."
    That really put it in perspective for me. I normally judge a supercomputer by how many "all Earth people hand calculation years" it can do in a day...
    --
    :(){ :|:& };:
    1. Re:Perspective? by Digestromath · · Score: 2, Funny

      Normally I would compare computers by floating operations per second. However sicne I guess we are going back to the old style of comparing it to people doing calculations by hand. What about all the people on earth using abaci 24/7? Or by leagues per bushel over cubits squared?

    2. Re:Perspective? by Thalin · · Score: 2, Funny

      At IBM we like to measure things in "Libraries of Congress per second", or perhaps "747s of phonebooks per second". ;)

      --
      What? You want a sig?
    3. Re:Perspective? by Arathon · · Score: 1

      What I need to know, is how many candlepower is its HDD access light, and how many horsepower would it take to drive this thing from Tokyo to Timbuktu?

    4. Re:Perspective? by Miseph · · Score: 1

      Yes, but can you convert that to automobiles?

      --
      Try not to take me more seriously than I take myself.
    5. Re:Perspective? by x2A · · Score: 1

      How about metres per second of ink drawn on paper by the number of people required to work out as many long division sums per second as the computer can do?

      That's be cool, you could measure the speed of your computer as a fraction of the speed of light.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    6. Re:Perspective? by Anonymous Coward · · Score: 0

      To put the performance of the machine in perspective, Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day."


      That really put it in perspective for me. I normally judge a supercomputer by how many "all Earth people hand calculation years" it can do in a day... To really put thing into perspective, how many Libraries of Congresses is that?
    7. Re:Perspective? by MadMidnightBomber · · Score: 1

      ...especially since the number of people on Earth is known to be a constant.

      --
      "It doesn't cost enough, and it makes too much sense."
    8. Re:Perspective? by Anonymous Coward · · Score: 0

      Yes... but the real question is: was Mr. D'Agostino talking about African or European calculators?

    9. Re:Perspective? by CAIMLAS · · Score: 1

      What point font are the phone books printed in?

      --
      ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
  20. err by RockoTDF · · Score: 0

    Correct me if I'm wrong, but the Cell wasn't designed originally for the PS3, even though that is the primary role it has found itself in.

    --
    There is more to science than physics!

    www.iomalfunction.blogspot.com
    1. Re:err by Anonymous Coward · · Score: 3, Informative

      It was designed originally for the PS3. But not solely for it.

      Cell was the brainchild of Ken Kutagari of Sony and Peter Hofstee of IBM.

  21. ummm... by WheresMyDingo · · Score: 5, Insightful
    if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day

    probably because most of those people would either try to eat the calculator or sell it for food and medicine

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

      Wouldn't work. If everyone had a calculator, no body would want to buy yours. Better off eating it, must contain a LOT of fiber.

    2. Re:ummm... by x2A · · Score: 4, Funny

      Grr, we're already overpopulated without everyone on the planet multiplying t the same time.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    3. Re:ummm... by Anonymous Coward · · Score: 0

      1. Publicly claim to perform calculations for food.
      2. Eat calculator once received
      3. ???
      4. Profit

      ---

      Sorry, yes I know life is unfair to a lot of people, but no one is really going to give a calculator to every person on the planet and Mr. D'Agostino is not responsible for the state of it (yes he is guilty of a really really bad analogy)

  22. Enough With The Fanboyism by Anonymous Coward · · Score: 2, Interesting

    "I certainly didn't"

    He was obviously asking about people with a clue, not stupid little fanboys who love to spout obviously false quotes and claims supposedly made by console makers.

    Sony NEVER once claimed the PS2 was a supercomputer. Not ONCE.

    The EE WAS powerful, cheap, and power efficient enough that at the time of its arrival on the market it fell under government scrutiny for its potential military uses. The EE utterly SHIT over any other chip on the market or would be on the market for another two years after its release with regards to its floating point power and heat/power usage combination.

    Let me guess, you're another one of those pathetic little fanboys who go around repeating that tired old lie about Sony, the PS2, and Toy Story graphics...

    http://builder-news.com.com/2100-1040-250632.html

    "One of the basic premises of the Xbox is to put the power in the hands of the artist," Blackley said, which is why Xbox developers "are achieving a level of visual detail you really get in 'Toy Story.'"

  23. Back it my day! by Anonymous Coward · · Score: 3, Funny

    The kids these days are lazy, back in my day if we wanted to know if a nuke worked we'd take it out back test it!
    Whatever happened to nuked marsh mellows or sitting round with Geiger counters trying to make funny sounds?
    Kids are lazy these days!

    1. Re:Back it my day! by mgblst · · Score: 1

      Or do what the French and British used to do, and take it out to somebody elses backyard to test it. Much better idea.

  24. calculators by hansoloaf · · Score: 2, Funny

    I bet if everyone had the TI-57, it'll take the aforementioned 46 years.
    But the TI-68 will cut it down to 23 years.

    1. Re:calculators by MilesAttacca · · Score: 1

      Does it count if I do my figures on a TI-99?

      --
      98% of America's teens drink alcohol, smoke, and have sex. Put this in your sig if you like bagels.
    2. Re:calculators by Anonymous Coward · · Score: 0

      Throw some HP 32SII's in there and knock it down to a couple of hours.

    3. Re:calculators by Reverend528 · · Score: 1

      Only if you successfully find and kill the wumpus.

  25. Re:exaflop, zettaflop, the yottaflop and the xeraf by anaesthetica · · Score: 4, Insightful

    Are you kidding?

    Not really. The post you link to describes the defense budget as it dwarfs other spending, but doesn't really argue why or why not that spending is progressive/regressive.

    The military was one of the first racially integrated public institutions in the U.S., it researched and funded the Internet, it's pouring money into synthetic fuels right now, and it's pushing the limits of computing power as seen in this article. There are numerous other scientific and social areas in which the military advances society, with far more practical results than do-gooders in other government or public institutions.

  26. Not in perspective by justinlee37 · · Score: 3, Insightful

    To put the performance of the machine in perspective, Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day.

    That does not put the performance of the machine in perspective at all. Technical details would be much more accurate and effective.

    1. Re:Not in perspective by Anonymous Coward · · Score: 0

      In fact the author got it all wrong too. 46 years is 1.5 Billion second. So six billion people get 10^19 human seconds. Computer gets one day or about 10^20 flop. Does the author say that human can do 10 flops????

    2. Re:Not in perspective by pH7.0 · · Score: 1

      To put the performance of the machine in perspective, Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day.

      I also notice some media change "hand calculators" to "handheld computer(s)" (USA today
      (LA Daily News. Mirror.co.uk) and make it even more confusing. Today's "handheld computer" can be pretty fast.
  27. Mouse brain at half speed? by dgarbett · · Score: 0

    The previous model was able to simulate a mouse brain at quarter speed IIRC

  28. Ahh, analogies, so unhelpful by lawaetf1 · · Score: 0, Redundant

    Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day."
    And if you tried to re-produce the energy present in a single tank of gas it would take you a year of back-breaking labor. Probably more like five years.
    Which isn't to say that I don't think the machine is impressive. Were it only around a few years ago it might have calculated that the Iraq war wouldn't be a lil' "let freedom ring!" jaunt.
    --
    CommentBot 0.7a running with args "-module irritate,disagree -target random"
    1. Re:Ahh, analogies, so unhelpful by CrazyJim1 · · Score: 1

      Especially if the people on the calculators are just pressing +1 over and over again for 46 years.

    2. Re:Ahh, analogies, so unhelpful by x2A · · Score: 1

      That would certainly explain all those global warming estimates!

      --
      The revolution will not be televised... but it will have a page on Wikipedia
  29. A link from LANL itself -- some propaganda by gardyloo · · Score: 1

    I missed all of the general-audience talks which have been given at LANL so far about Roadrunner, but this makes me want to try a bit harder at making it to some of them:

    http://www.lanl.gov/news/index.php/fuseaction/1663.article/d/200805/id/13277

  30. Take that petaflop with a grain of salt by jimhill · · Score: 4, Interesting

    As a software developer who's worked on the Lab's previous ASC machines (Blue Mountain, Q, Lightning) I can say that once the calculation is run to get a machine atop Jack Dongarra's gee-golly list, it's partitioned, segmented, divided, and subjected to such crappy resource management that if I could trade the entire machine for a pair of coupled 8-core Mac Pros I'd do it in a heartbeat.

    The real PITA with these machines is that the powers that be are trying to kill two birds with one stone: they want an R&D platform for advanced computing, but they also want to certify an aging and untestable nuclear stockpile. That rather requires a fairly static platform, and so far our experience with ASC has been that when a machine hits that sweet state, they yank it and give us the next one.

    --
    Learn to spell: nickel, missile, lose, solely, amendment, speech, kernel, probably, ridiculous, deity, hierarchy, versus
  31. yes but... by strength_of_10_men · · Score: 1

    if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day.

    if those calculators were RPN, it would take only 32 years.
  32. The CELL processor is single precision by ancient_kings · · Score: 0

    which translates into wrong answers most of the time. Me guess is that for problems requireing double-precision numerics, you should divide CELL based supercomputer by 10 to 100 (software emulation of double precision is MASSIVELY SLOW), so this is really a teraflop machine. No big whoop...NEXT!!!

    1. Re:The CELL processor is single precision by Anonymous Coward · · Score: 0

      It really is amazing how dumb some readers are here.
      I mean, it's easy enough to read the article,
      or perhaps read the LANL article above. The
      machine in question is an enhanced Cell system,
      in fact capable of double precision ops as opposed
      to first generation Cell silicon,

      Really, RTFM is most applicable when considering
      answering an engineering related article. Putz.
      (sorry, this nonsense just pisses me off....why
      so few really deep thinkers on slashdot?????)

    2. Re:The CELL processor is single precision by DegreeOfFreedom · · Score: 1

      HPL, the code that is run for the Top500 number (or an equivalent implementation, optimized for the target architecture), is double precision. This is perhaps the most important rule of the benchmark: the calculation must be carried out in full double precision. The 1.026 Pflop/s number does represent a double-precision workload.

    3. Re:The CELL processor is single precision by genericpoweruser · · Score: 1

      That's easy to answer: time. The GP didn't have the patience and/or time to RTFA, but thought he knew enough about the Cell processor to make a remark. How about cutting him some slack instead of making a personal attack. Putz.

      --
      A fool and his lamb are worth two in the bush.
  33. Re:exaflop, zettaflop, the yottaflop and the xeraf by blahplusplus · · Score: 5, Insightful

    "There are numerous other scientific and social areas in which the military advances society, with far more practical results than do-gooders in other government or public institutions."

    It's because the military doesn't have the scrutiny and oversight other institutions do, lets face it. Do public institutions besides the miilitary get secret prison's and liscense to do whatever the want? The military is not held back by moral qualms. We've seen this with all sorts of classified documents coming out of the government. The military has budgets that are kept secret. For anyone to claim the 'military helps us' vs public institutions, we'd have to do an analysis. But that would be fairly difficult and politically sensitive, now wouldn't it?

  34. Computing the data pyramids by Eternal+Vigilance · · Score: 1, Funny

    Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week... I think this analogy may reveal a little more than intended about government's vision for humanity.

    "Let my people goto!"
    1. Re:Computing the data pyramids by anaesthetica · · Score: 1

      You're right. The Bush Administration has had plans from before 9/11 to pour vast sums of money into actuarial and accounting schools through a secret CIA slush fund operating in various former Eastern bloc countries.

    2. Re:Computing the data pyramids by Anonymous Coward · · Score: 0

      And he was a long time in his hole, human population has since grown to 6.7billion :> And thus it would only take us about 41years to do what said supercomputer can do in 1 day.

  35. The result of their research: by suck_burners_rice · · Score: 4, Funny

    The answer is 42. The question is left as an exercise for the reader.

    --
    McCain/Palin '08. Now THAT's hope and change!
    1. Re:The result of their research: by prestomation · · Score: 1

      1. 42
      2. ???
      3. Profit!

      Can't go wrong with that formula.

      Well, you can probably cut out step 2, it's really not needed.

  36. Re:exaflop, zettaflop, the yottaflop and the xeraf by anaesthetica · · Score: 2, Interesting

    ...secret prison's [sic]...not held back by moral qualms...

    Are you really arguing that the scientific and social advances from the military arise from secret prisons and lack of moral qualms?

  37. Military?? (Dept of Energy) by Anonymous Coward · · Score: 0

    Folks,
    It is the Department of Energy that runs Los Alamos National Laboratory. Even if it is on a military base.

  38. Yet... by Tastecicles · · Score: 1

    ...a source from the last week (I can't put my finger on it right now) says that some boffs have built a cluster with over 120,000 cores; this system was used during testing to simulate ten seconds of processing in half a mouse brain. This processing took three /days/ to complete.

    Does this mean that mice are smarter than humans?

    --
    Operation Guillotine is in effect.
  39. Re:Not in perspective - this is a media number by mykepredko · · Score: 5, Interesting
    This is a perfect example of a propellor head trying to come up with an analogy for a media/marketing type. I suspect that this was the only one that the powers that be felt non-techies could relate to. I've been asked to come up with these analogies a couple of times and it can be pretty frustrating on both sides.

    I suspect the first example of this happening was trying to estimate how many angels could fit on the head of a pin.

    Other meaningless analogies could be:
    • How long it would take Roadrunner to count all the atoms in the universe
    • What speed your car would run at if the speed difference between Roadrunner and your home computer was multiplied by 60 mph
    • If we could go this many times faster than the speed of light, how fast could we cross the universe
    • If in every instruction it could take in one byte of text, how long it would take to read the Library of Congress
    • How fast it could render "The Incredibles" compared to how long it took the original server farm (actually, this might be one that's understandable)
    • How fast it could break the 128 bit encryption used when you log onto your bank's web page to pay your bills (this might also be understandable and would probably be a bit scary)

    The simple fact is that a petaflop computer works faster than humans can conceive and any kind of analogy cannot be comprehended.

    myke
  40. WTF happened to plain old units of measure? by PseudoThink · · Score: 3, Funny

    if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day

    I'm glad to see the continuing trend of creatively "dumbing down" units of measure (in this case, flops) to the point where they are not only practically useless, but entirely divorced from reality. I would like to propose the following similar, hype-worthy measure for fuel economy:

    Old: Miles per gallon
    New: Number of miles from which one would smell the excrement from the number of cattle one could feed for a day with the amount of corn it would take to produce one gallon.

    1. Re:WTF happened to plain old units of measure? by mrbluze · · Score: 1

      I'm glad to see the continuing trend of creatively "dumbing down" units of measure (in this case, flops) Yeah, my dad would have flipped if he'd read the summary.
      --
      Do it yourself, because no one else will do it yourself. [beta blockade 10-17 Feb]
    2. Re:WTF happened to plain old units of measure? by ozbird · · Score: 1

      I'm glad to see the continuing trend of creatively "dumbing down" units of measure (in this case, flops) to the point where they are not only practically useless, but entirely divorced from reality.

      Yes - are they regular (infix) or Reverse Polish (postfix) notation calculators? Dumbed-down minds need to know!

  41. Re:exaflop, zettaflop, the yottaflop and the xeraf by religious+freak · · Score: 1

    Why is the military so much more progressive (with practical results) than any other institution of government? Military is the expression of the power of a nation as a whole... the means by which nations, using their national resources, keep ahead of the pack. The rest of the government is bloated bureaucracy designed to keep that powerful military in check and keep society functioning.
    --
    If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
  42. Yes, It Does Run Linux by Doc+Ruby · · Score: 2, Informative
    As I posted the last time this story was reported (in IBM Touts Supercomputers for Enterprise") in "Yes, It Does Run Linux":

    From IBM's detailed press release [ibm.com]:

    the QS22 boasts an open environment, utilizing the flexibility of Red Hat Enterprise Linux as the primary operating system and the open development environment of Eclipse.


    That means that a PS3 running Linux, even with its ridiculously low 512MB RAM, can be used as a $500 development platform for these CellBE BladeServers.

    And, in turn, some QS22 SW might be usable on the PS3, if it can be ported to use the tiny RAM. Or if someone hooks an i-RAM bank to the SATA port as swap/ramdisk, using perhaps iSCSI over its Gb-e for storage.


    Now get out there and supercompute!
    --

    --
    make install -not war

  43. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    The opterons are used to feed the cell processors. They are paired up, one opteron for each cell.

  44. Re:exaflop, zettaflop, the yottaflop and the xeraf by anaesthetica · · Score: 1

    You sound like a lot of the scary fin de siècle German political theorists that I have to read for my poli-sci Ph.D. studies. Calm down--we all know how that story ended.

  45. Old News by MBHkewl · · Score: 3, Informative

    This was covered last year, and the Los Alamos website had a few interviews with some people involved on what the uses of Roadrunner are. They had a time-line of what phases are to be done, and as far as memory serves me, they were going with Opterons for the first phase, then performance assessment, then add the Cell processors in the third phase.

    From these pictures, it clearly shows they're using IBM Blades (4 chassis in each rack), and IBM already offers BladeQ servers which use Cell processors for HPC applications. The IBM BladeQ servers pack double the CPUs of a PS3.

    If you take a look at the Folding@Home project statistics, you can see the performance of PS3 boxes, and almost relate...
    --
    Mod points are a dangerous tool. Abuse them wisely.
  46. Does not compute... by fahrbot-bot · · Score: 0, Redundant
    if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day.

    Can I have that figure in something more useful, like Library of Congresses / Fortnight or an automotive analogy?

    --
    It must have been something you assimilated. . . .
  47. Explore scientific problems like climate change? by hamster_nz · · Score: 2, Funny
    The article mentions that it will be used to explore climate change. At 3MW, perhaps it is likely to cause climate change!

  48. Re:Not in perspective - this is a media number by jareds · · Score: 4, Informative

    How fast it could break the 128 bit encryption used when you log onto your bank's web page to pay your bills (this might also be understandable and would probably be a bit scary)

    No, not at all scary. It's apparently twice is fast as the BlueGene/L, which apparently set a record of 478.2 teraFLOPS. Let's assume it takes 1 floating-point operation to test a single key, which is a gross underestimate. We'll thus assume the Roadrunner can test 10^15 keys per second. Testing 2^128 keys would then take about 10^16 years.

  49. Re:Not in perspective - this is a media number by jthill · · Score: 1

    The analogy I've liked best is distance: in this case, in the time it takes ars's God Box to get to the supermarket, this puppy has been around the world a few times.

    --
    As always, all IMO. Insert "I think" everywhere grammatically possible.
  50. Re:exaflop, zettaflop, the yottaflop and the xeraf by blahplusplus · · Score: 3, Insightful

    My point is they do not have the same barriers other institutions do: i.e. the gaps funding and scrutiny. My point about mentioning secret prisons was merely an example of the previous point.

  51. It will take new technologies. by Jane+Q.+Public · · Score: 1

    It is highly doubtful that the "zettaflop", a million times this petaflop, will be achieved by "conventional" circuitry. That will take optical or some other kind of computing (probably not quantum). The yottaflop will likely be quantum or molecular-state computing, or something unexpected.

    1. Re:It will take new technologies. by genericpoweruser · · Score: 1

      I remember reading about a hypothetical computer where the processor sends the completed operation's results back in time to when it was first asked for. Since the processor would then already have the data it would not actually have to process it. Thus the processor could be as fast as anything.

      --
      A fool and his lamb are worth two in the bush.
    2. Re:It will take new technologies. by Anonymous Coward · · Score: 0

      That would...huh?
      This is a tech site, not a philosophy forum.

  52. 1st PETAflop computer by karmer · · Score: 1

    Roadrunner = Audrey III ?

  53. Re:Not in perspective - this is a media number by mykepredko · · Score: 1

    Thanx - I should have done the back of envelope like you've done here before coming up with that particular analogy.

    myke

  54. wikipedia by Anonymous Coward · · Score: 2, Informative

    Each Compute or I/O node is a single ASIC with associated DRAM memory chips. The ASIC integrates two 700 MHz PowerPC 440 embedded processors, each with a double-pipeline-double-precision Floating Point Unit (FPU), a cache sub-system with built-in DRAM controller and the logic to support multiple communication sub-systems. The dual FPUs give each BlueGene/L node a theoretical peak performance of 5.6 GFLOPS (gigaFLOPS). Node CPUs are not cache coherent with one another.

  55. Yes by Konster · · Score: 0, Offtopic

    But will it blend?

  56. This CELL is not single precision by raftpeople · · Score: 4, Informative

    The CELL processor is single precision, which translates into wrong answers most of the time. Me guess is that for problems requireing double-precision numerics, you should divide CELL based supercomputer by 10 to 100 (software emulation of double precision is MASSIVELY SLOW), so this is really a teraflop machine. No big whoop...NEXT!!!

    Things move fast in technology Jethro, including this 2nd gen of the CELL proc, this is what you missed:

    Double Precision FP - 190TFLOPS (5 times faster than 1st CELL)
    Memory: Expanded to 32gb
    Memory: DDR2 instead of Rambus
    65nm (I know, I know, but it's better than 90nm)
    1. Re:This CELL is not single precision by afidel · · Score: 2, Interesting

      IBM's process at 65nm is pretty good since it includes SOI and strained silicon. They are going to make the original cell at 45nm for Sony so I'm sure cell-2 will move to that process eventually, the East Fishkill fab has to be pretty busy since AFAIK it's IBM's only 45nm plant and it just started volume production this quarter.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    2. Re:This CELL is not single precision by Anonymous Coward · · Score: 0

      Do the new PS3s have this double precision FP Cell2 processor?

    3. Re:This CELL is not single precision by faragon · · Score: 2, Informative

      The difference between the two Cells it is not "abysmal", let's do some comparisons:

      a) PS3 "Classic Cell" 1 PPC64 w/ 2 threads and 7 SPEs (8, but one disabled, defective or not):

      GFLOPS 32-bit (float): 3.2GHz * 8 FLOPS/Hz * 7 SPEs = 179.2 GFLOPS
      GFLOPS 64-bit (double): 3.2GHz * 1 FLOPS/Hz * 7 SPEs = 22.4 GFLOPS (huge penalty, because of simulation via unoptimized simple precission operation)
      GFLOPS 64-bit (double) a optimized 32-bit operation: 3.2GHz * 3.9 FLOPS/Hz * 7 SPEs = 87.36 GFLOPS

      b) Roadrunner "New Cell" 1 PPC64 w/ 2 threads and 8 SPEs:

      GFLOPS 32-bit (float): 4GHz * 8 FLOPS/Hz * 8 SPEs = 256 GFLOPS
      GFLOPS 64-bit (double): 4GHz * 4 FLOPS/Hz * 8 SPEs = 128 GFLOPS

      P.S. ad hoc rewrite, based on my own Journal at Barrapunto (spanish /.).

    4. Re:This CELL is not single precision by Anonymous Coward · · Score: 0

      You left out "Splinter Cell"

  57. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    From my understanding of the "So, it also has Cell-based processors AND Opterons. I wonder what the functional division between the two chip types is?" question is that the Cell is awful for doing any type of management, so a supercomputer requires the Opteron to handle the delegation of tasks to the Cell processors. I have to hand it to the Cell's for their fp calculations though.

  58. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 3, Informative

    So, it also has Cell-based processors AND Opterons. I wonder what the functional division between the two chip types is?

    Each node has two Opterons and 4 PowerXCell 8 processors (an upgrade to the PS3's Cell processor). This allows a developer writing code for the platform to run in a number of different modes: all Opteron, all Cell, or something in between. The first of these (all Opteron) may constitute a significant amount of the early work on the machine by practitioners, as they can simply compile legacy codes to the platform and ignore the Cell processors. Of course, to reap the full benefit of the machine, developers will exploit both the Cell chips and the Opteron chips.
  59. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    The military's done a far better job of restraining itself than the Bush Administration, at least as far as secret prisons, rash wars, and torture goes. Far better than Congress, too, as well as the courts.

    The job that JAG's done in opposing Bush and Yoo--and defending both the Gitmo detainees and the Constitution--at the cost of tens, if not hundreds, of individual careers--is nothing less than heroic. It makes me proud to be an American.

    They follow orders, yes, but they aren't blinded by them.

  60. Imagine a beowolf cluster of these. by Anonymous Coward · · Score: 0

    Lame. Obligatory. Comment.

  61. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    Why is the military so much more progressive (with practical results) than any other institution of government?

    I'd say it's just tradition. If it's progressive you're after, try Japan - not only did the MDGRAPE-3 supercomputer reach the petaflops performance level a couple of years ago already, but it's more or less filled with special-purpose processors for doing chemical, microbiological and pharmaceutical simulations.

    This is so different (and better, I'd argue) from Roadrunner on so many different levels. Honestly, I'm a little ashamed that in this day and age we are actually proud of having stuffed a bunch of mass-market processors in a room to study nukes.

  62. more information by meta+coder · · Score: 1

    Name: roadrunner (teraflopus excesus)

  63. Hand Held calculators and TurboCharchers? by Anonymous Coward · · Score: 0

    Wtf is wrong with these people?
    Either they think that the general population is retarded (I can't argue against that fact), Or
    They are retarded, (I can't argue against that fact either), or both (A fact that I also can not argue against)!.

    They are comparing these "Chips" to Handheld calculators and, get this, Turbochargers! , wow, yeah, the super computer is full of chips, turbochargers and handheld calculators, fucking genius!!!

  64. NOT MILITARY! by serviscope_minor · · Score: 3, Interesting

    It's the department of energy, not the military. Specifically, it is at Los Alamos, which is not a military base.

    --
    SJW n. One who posts facts.
    1. Re:NOT MILITARY! by mrbluze · · Score: 2, Funny

      It's the department of energy, not the military. Specifically, it is at Los Alamos, which is not a military base. What, you mean hundreds of moderator points were wasted on above military-related posts? Woohoo!! Bingo baby!
      --
      Do it yourself, because no one else will do it yourself. [beta blockade 10-17 Feb]
    2. Re:NOT MILITARY! by x2A · · Score: 1

      So how come the reason the machine didn't come with a shutdown command installed was "when have we ever had an exit strategy?"?

      --
      The revolution will not be televised... but it will have a page on Wikipedia
  65. Does that make them a bunch of.... by Anonymous Coward · · Score: 0

    Petaphiles??

  66. I misread and expected cell phone supercomputers by witherstaff · · Score: 1

    My first thought when I saw the article title was a cellphone based networked super computer. Something along the lines of Rainbow's End or Halting State.

    I wonder how many iphones would be needed to do a cellphone petaflop computer.

  67. Re:exaflop, zettaflop, the yottaflop and the xeraf by StormyWeather · · Score: 1

    They are more progressive because lives are at stake in an immediate fashion. Who wants to be the guy that says "yea we could have built a supercomputer to test nuclear mishandling, but we decided it would take too long and be too expensive.. oh, yea sorry about turning the Dallas metroplex into a cherry red ember."

  68. Re:exaflop, zettaflop, the yottaflop and the xeraf by religious+freak · · Score: 1

    Who's excited? I'm not saying if it's good or bad, I'm saying this is the truth.

    --
    If you can read this... 01110101 01110010 00100000 01100001 00100000 01100111 01100101 01100101 01101011
  69. Allow me to oblige ... by Zero__Kelvin · · Score: 4, Funny

    Take everyone on earth, and put them each in a different Ferrari Testarossa with no engine, no gas in the tank, and no ignition system. That is how fast this thing moves.

    Some other equally useful analogies:

    Take the same aforementioned people, and give them a OLPC. The amount of time it takes them all to calculate their degree of separation from Kevin Bacon, and divide by a googolplex. , then round up. That is the number of people that think the calculator analogy in the article was a good one.

    Take the inverse of the clock frequency and multiply it by the number of instructions required for Windows to boot far enough to attempt to obtain an IP Address dynamically. Add to that the time it takes for the DHCP request to reach your Billion made router. That is the amount of time it takes for it to hose your router. Take the inverse of the clock frequency and multiply it by the number of instructions it takes to apply a service pack. Add it to the boot time, calculated as described above. That is the amount of time it takes to achieve a BSOD.

    HTH,

    - Thomas P. D'Agostino

    --
    Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
    1. Re:Allow me to oblige ... by Idbar · · Score: 1

      I think the most insightful comment on this thread, and it's the funniest one too.

  70. I've spoken about this before... by zappepcs · · Score: 1

    Finding cheap and powerful processors like the Cell processor is an amazing step forward for everyone. Not because it can be part of a supercomputer, but because you can fit 5 of them on your desktop and have a supercomputer at home.

    If you can put 5 PS3s inside your desktop system, and use the big CPU you normally have as the admin CPU, you could have a virtual world as your desktop UI. Imagine walking your little avatar to your home office, clicking on a file cabinet, opening the folder for your budget, clicking on the latest date report folder, and it opens the right file for you. It would allow people to use computers and their tools in the same way that they operate in normal life, just VR.

    Then you walk to the kitchen of your desktop interface and click on the cook book and it opens a web browser loaded with links to cooking recipes and information. There are many more ways to use it, that's just an example. There are many more.

    If such power is made available to the home user, it will get used, and the uses will be amazing :) IMO anyway.

    1. Re:I've spoken about this before... by Anonymous Coward · · Score: 0

      Ok, if you're going to take a jab at Microsoft then at least be fair about it. We know Microsoft BOB sucked a bunch, but lets not pretend it was that greedy on system resources!

    2. Re:I've spoken about this before... by x2A · · Score: 1

      "Imagine walking your little avatar to your home office, clicking on a file cabinet, opening the folder for your budget, clicking on the latest date report folder, and it opens the right file for you"

      I'd rather a couple of shortcut keys. "Walking your little avatar to your home office...." sounds like I may as well be walking up to an actual file cabinet and looking for a folder, whole point of computers is being able to get them to do the fetching things for you.

      --
      The revolution will not be televised... but it will have a page on Wikipedia
    3. Re:I've spoken about this before... by hyades1 · · Score: 1

      "Imagine walking your little avatar to your home office, clicking on a file cabinet, opening the folder for your budget, clicking on the latest date report folder, and it opens the right file for you. It would allow people to use computers and their tools in the same way that they operate in normal life, just VR."

      And then your avatar walks over to its virtual desk, goes on-line and downloads some virtual pr0n. At which point, it performs various unwholesome activities that will land your real world computer in the shop and cause you to become a "person of interest" with the FBI.

      Having successfully trashed your life, the little bastard will offer a virtual snicker as armed members of the constabulary shove you into a car that will ultimately deliver you to a whole new life as Bubba's girlfriend. Your last coherent thought before the beatings start is that the Terminator movies got the whole Skynet/Rise of the Machines scenario just a wee bit wrong.

      --
      I've calculated my velocity with such exquisite precision that I have no idea where I am.
    4. Re:I've spoken about this before... by cnettel · · Score: 1

      Platform games could be done on a NES. Asheron's Call and other 3D MMORPGS are turning ten soon. A somewhat immersive virtual 3D world could be done years ago. Making a nice UI out of it will not require 5 Cells. (Really good speech recognition might need it. Current approaches do not get that much better by throwing CPU power at them, but it at least seems possible. Walking around a virtual world matching the real one in all the bad ways to actually get stuff done seems stupid. The best aspects of the desktop UI analogy were not, actually, that it was a tremendously realistic simulation of working at an actual office desk... it wasn't. It was close enough and provided an efficient working environment.)

  71. 133 million dollars by Profane+MuthaFucka · · Score: 1

    That got me thinking...what is the most expensive computer anyway? How does the fastest computer rank with the costliest computers?

    Google Google Google, and presto. The Japanese Earth Simulator supercomputer cost $400 million dollars.

    --
    Fascism trolls keeping me up every night. When I starts a preachin', he HITS ME WITH HIS REICH!
    1. Re:133 million dollars by Ecuador · · Score: 1

      Which explains why it is still a respectable #30 in the list after all these years. For reference, the oldest entries on the 2007/11 list are from 2002, and the Earth Simulator is one of them (by far the fastest). Also, when the Earth Simulator was installed it was an astonishing 5 times faster than the second entry.

      --
      Violence is the last refuge of the incompetent. Polar Scope Align for iOS
    2. Re:133 million dollars by Anonymous Coward · · Score: 0

      Also, it's a proprietary vector computer. Much expensive, but should be far easier to use than BG.

  72. In bitter cold Sweden by this+great+guy · · Score: 1

    old women use supercomputers to dry laundry !

  73. Re:I misread and expected cell phone supercomputer by genericpoweruser · · Score: 1

    Wouldn't that be murder to battery life?

    --
    A fool and his lamb are worth two in the bush.
  74. Re:exaflop, zettaflop, the yottaflop and the xeraf by BoneKracker · · Score: 1

    No analysis is required to see the many, many open source projects funded by the military. For example, neither UNIX nor the Internet would have come to be without DARPA funding (not to mention OpenBSD, Reiserfs, etc., etc., etc.) No analysis is required to see the many, many public works performed in the U.S. and world wide by the Army Corps of Engineers. No analysis is required to see the dozen or two humanitarian aid missions being carried out at any given time by the U.S. military. No analysis is necessary to see the preferential treatment given by the U.S. military in awarding tens of billions of dollars in contracts (as a matter of policy) to small businesses, minority-owned businesses, women-owned businesses, and businesses operating in economically depressed communities. Just because those socially-beneficial activities aren't the kind of sensationalist blather (like "OMG secrit prisonz") that NBC or CNN like to pump into "inquiring minds" like yours does not mean the information is not readily available and widely-known.

  75. Data Collection anomalies by drew30319 · · Score: 3, Interesting
    Are you familiar with this site? http://www.surfacestations.org/ There are 1221 USHCN climate stations in the USA and (according to the website) no photographic database of these stations exists. It includes examples like:

    Forest Grove, OR (air conditioner exhaust blowing directly on temperature shelter, nearby asphalt and buildings)
    and

    Aberdeen, WA ( large heat source: sewage treatment plant - nearby parking, hot automobile radiators )
    I could see how these would be of concern to anybody interested in data quality.
    --
    JAGga.me ----> Producing video games addressing emotional health and wellness issues affecting teens.
  76. Unit Conversions by LordHatrus · · Score: 1

    >> "six billion people on earth used hand calculators
    >> and performed calculations 24 hours a day and seven
    >> days a week, it would take them 46 years to do what
    >> the Roadrunner can in one day."

    Who the hell is sitting out there submitting /. stories with the most crazy-ass units?!!
    6-billion-people-hand-calculating-24/7-years?
    Libraries of Congress?
    Rods per hogshead?

  77. But will it run.. by Anonymous Coward · · Score: 1, Funny

    Coyotos?

    [they do get a lot of military funding, iirc.]

  78. I know what can catch this "Roadrunner" by wildem · · Score: 1

    I am in the process of building and ACME "Coyote" supercomputer cluster made of used toasters, open cans of beans and bottle caps.

  79. Re:Not in perspective - this is a media number by x2A · · Score: 1

    Well I'm definitely gonna start my password with a 'z'. People who start it with the letter 'a' have their passwords cracked by brute force in one 26th of the time!

    --
    The revolution will not be televised... but it will have a page on Wikipedia
  80. NOT 46! by Provocateur · · Score: 1

    42!

    --
    WARNING: Smartphones have side effects--most of them undocumented.
  81. Re:exaflop, zettaflop, the yottaflop and the xeraf by iwein · · Score: 1

    Military taking the lead on computing as usual. Why is the military so much more progressive (with practical results) than any other institution of government
    Just a wild guess: budget?

    --
    Show a man some news, distract him for an hour. Show a man some mod points, distract him for the rest of his life.
  82. Re:exaflop, zettaflop, the yottaflop and the xeraf by GigaplexNZ · · Score: 2, Insightful

    The job that JAG's done in opposing Bush and Yoo--and defending both the Gitmo detainees and the Constitution--at the cost of tens, if not hundreds, of individual careers--is nothing less than heroic. It makes me proud to be an American. Yeah, I would feel really proud too if the leader of my country was an illiterate moron who violated human rights and caused honest people to lose their careers because they chose to do the right thing.
  83. Jonathan Richman - Roadrunnner by permaculture · · Score: 1

    One two three four five six!

    Roadrunner, roadrunner
    Going faster miles an hour
    Gonna drive past the Stop 'n' Shop
    With the radio on
    I'm in love with Massachusetts
    And the neon when it's cold outside
    And the highway when it's late at night
    Got the radio on
    I'm like the roadrunner

    Alright
    I'm in love with modern moonlight
    128 when it's dark outside
    I'm in love with Massachusetts
    I'm in love with the radio on
    It helps me from being alone late at night
    It helps me from being lonely late at night
    I don't feel so bad now in the car
    Don't feel so alone, got the radio on
    Like the roadrunner
    That's right

    Said welcome to the spirit of 1956
    Patient in the bushes next to '57
    The highway is your girlfriend as you go by quick
    Suburban trees, suburban speed
    And it smells like heaven(thunder)
    And I say roadrunner once
    Roadrunner twice
    I'm in love with rock & roll and I'll be out all night
    Roadrunner
    That's right

    Well now
    Roadrunner, roadrunner
    Going faster miles an hour
    Gonna drive to the Stop 'n' Shop
    With the radio on at night
    And me in love with modern moonlight
    Me in love with modern rock & roll
    Modern girls and modern rock & roll
    Don't feel so alone, got the radio on
    Like the roadrunner
    O.K., now you sing Modern Lovers

    (Radio On!)
    I got the AM
    (Radio On!)
    Got the car, got the AM
    (Radio On!)
    Got the AM sound, got the
    (Radio On!)
    Got the rockin' modern neon sound
    (Radio On!)
    I got the car from Massachusetts, got the
    (Radio On!)
    I got the power of Massachusetts when it's late at night
    (Radio On!)
    I got the modern sounds of modern Massachusetts
    I've got the world, got the turnpike, got the
    I've got the, got the power of the AM
    Got the, late at night, rock & roll late at night
    The factories and the auto signs got the power of modern sounds
    Alright

    --
    Environmentalism is the new Victorianism. Everyone ties on a green corset and pretends we're virtuous.
  84. Re:Ummm by jacquesm · · Score: 2, Funny

    maybe he should get 6 billion hand calculators and mail them out ?

    That's probably a new contender for the stupidest metric ever, it beats 'libraries of congress per second' hands down.

  85. Obligatory by 4D6963 · · Score: 1

    Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day.

    But how many Libraries of Congress of data can it handle?

    --
    You just got troll'd!
  86. Yeah, but by Anonymous Coward · · Score: 0

    It will also take all these 6 billion people armed with calculators 46 years to write "Hello World" that uses half of the computing power of the beast.

  87. what a stupid statistic by daviddickson1 · · Score: 1

    ...if all six billion people ... what a stupid statistic ... If all six hundred million grandmothers on earth tried to program a video recorder (sorry about this reference to an archaic machine for the very young) every hour, 24 hours a day and seven days a week, it would take them 46 years to record what I can in one day.

  88. Re:exaflop, zettaflop, the yottaflop and the xeraf by jejones · · Score: 1

    Looks like they use the standard deceptive practice amongst such groups of only including discretionary spending in the pie chart; since discretionary spending is less than half the spending and shrinking quickly as the "entitlements" grow at an insane rate, the pie chart is worse than useless; it's lying with statistics, to borrow the title of Darrell Huff's classic book.

  89. and its first simulation result was .... by thedbp · · Score: 1

    ... Get the hell out of Iraq.

  90. So in 15 years... by Crookdotter · · Score: 1

    ...we will have this kind of power on our desktop? Wikipedia tells me that a current desktop PC is equivalent to the top supercomputer of 15 years ago in terms of operations. I can't imagine having this amount of computing power in 15 years time, but it makes me think of the possible programs we will run.

  91. Cell processor by rshimizu12 · · Score: 1

    One thing to remember is that there is various iterations of the cell processor. The Xbox is a 3 core version The Playstation. I believe the Playstation is a 6 core processor. The roadrunner will use a 8 core processor. IBM originally discussed having a 16 core processor. There was not much talk about it afterwards. My guess is that there was significant bus contention issues. The original Power4's shut down one of their cores while running at full speed to avoid contention. The Power6 was designed to overcome these issues.

    1. Re:Cell processor by jacquesm · · Score: 1

      it makes you wonder what kind of magic HP and SUN use in their 'big iron', 64 or more cpus in a machine is one thing, to have them all working in a meaningful way on a problem without starvation issues is no mean feat.

    2. Re:Cell processor by BiggerIsBetter · · Score: 4, Informative

      One thing to remember is that there is various iterations of the cell processor. The Xbox is a 3 core version The Playstation. I believe the Playstation is a 6 core processor. The roadrunner will use a 8 core processor. IBM originally discussed having a 16 core processor. There was not much talk about it afterwards. My guess is that there was significant bus contention issues. The original Power4's shut down one of their cores while running at full speed to avoid contention. The Power6 was designed to overcome these issues. No, the XBox 360 has a three core PowerPC processor, not a Cell BE processor. The Cell BE in the PS3 has 1 PPE (Power core) and 7 SPEs (that "other" CPU core), while IBM apparently gets to use the fully functional Cells (PPE + 8 SPE) in their more expensive hardware. Those Cells with even fewer functioning SPEs might end up in HDTV TVs or similar.
      --
      Forget thrust, drag, lift and weight. Airplanes fly because of money.
    3. Re:Cell processor by bestinshow · · Score: 3, Informative

      XBox360 has a tri-core in-order PowerPC - each core is actually very similar to the single general purpose PPU in the PS3's Cell.

      Cell in addition has 8 SPUs. 1 is disabled in the PS3 for yield reasons, and another is reserved, so there are 6 available for general purpose computing.

      Both run at 3.2GHz. I think Cell has at least 3x the vector/streaming power of the XBox 360 CPU, but only 1/3rd of the general purpose capability. Figures pulled from thin air, etc.

    4. Re:Cell processor by Nursie · · Score: 1

      Not only that, but they usae the next generation up from Cell BE, Cell PowerX8i. Alledgedly this is five times faster.

    5. Re:Cell processor by king-manic · · Score: 1

      One thing to remember is that there is various iterations of the cell processor. The Xbox is a 3 core version The Playstation. I believe the Playstation is a 6 core processor. The roadrunner will use a 8 core processor. IBM originally discussed having a 16 core processor. There was not much talk about it afterwards. My guess is that there was significant bus contention issues. The original Power4's shut down one of their cores while running at full speed to avoid contention. The Power6 was designed to overcome these issues. The 360 is a 3 slim core PPC chip. It isn't a cell. The PS3 cell has 1 slim PPC core and 8 SPU's (1 turned off for yeild, 1 dedicated to prevent hacking), the SPU's are the innovative thing behind the Cell. They are dumb but fast floating point number crunchers. In effect the PS3's Cell is a 7 Core processor with 6 very specialized cores while the 360's Xenon is 3 slim PPC cores. A slower variant of those PPC cores also powers the wii.
      --
      "There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
  92. Re:Not in perspective - this is a media number by Crookdotter · · Score: 1

    Does a brute force cracker have to go from a to z? When done in a non linear sequence it removes any advantage by starting with a z (or disadvantage starting with an a).

  93. Re:exaflop, zettaflop, the yottaflop and the xeraf by Wyrmy · · Score: 0

    Its the same reason NASA has spurred many developments and improvements in the rest of the civilian world. Does this mean that after a certain number of production types they will lose the blueprints and go back to the abacus?
    --
    Every act of conscious learning requires the willingness to suffer an injury to one's self-esteem.-Thomas Szasz
  94. Re:Not in perspective - this is a media number by Anonymous Coward · · Score: 0

    That's assuming you have to try every single key to break 128-bit SSL encryption. Is that actually the case?

  95. Re:Not in perspective - this is a media number by x2A · · Score: 1

    "Does a brute force cracker have to go from a to z?"

    No absolutely not, and with encryption methods used for ssl, generally randomly created large/prime keys are used rather than passwords, so you'd be searching for numbers rather than words.

    --
    The revolution will not be televised... but it will have a page on Wikipedia
  96. Human Brain by Kifoth · · Score: 3, Interesting
    A quick Google search: http://www.google.com/search?q=human+brain+petaflop, suggests that we're probably in spitting distance of at least matching the processing power of a human brain.

    Not sure about the software though...

    1. Re:Human Brain by HetMes · · Score: 2, Interesting

      There are so many things wrong with this claim!
      1.) The people who claim this, are probably also the ones who benefit if it is believed to be true, in terms of funding, general interest, cool-factor.
      2.) Comparing the human brain and computers is not even comparing apples and oranges, it's apples and ellipses. Both are sort of round, but that's about as far as it goes.
      3.) We have no idea exactly where we are on the road to matching the human brain in any way. So how can we know how long the road ahead is.
      4.) Processing power of the brain is more or less irrelevant to its capabilities. Hundreds of millions of years ago the first insects walked around without falling over every step. We can't even build a decent pair of robotic legs to do that.

      All in all, it's like saying the beach has more processing power than we will ever need, the molecules just need to be rearranged. Maybe so, but until that time, it's just sand incapable of adding 1 + 1.

  97. Imagine Beowulf cluster of this... by Zdzicho00 · · Score: 1

    /Z

  98. Re:exaflop, zettaflop, the yottaflop and the xeraf by Alpha+Whisky · · Score: 1, Funny

    On the other hand, it'll also make it easier to figure out how to make bigger sticks that are lighter and sharper. You mean they are taking Einstein's prediction that World War 4 will be fought with sticks and stones seriously, and want to be ready for it?
    --
    it's = it is

    its = belonging to it

  99. Re:Ummm by Stellian · · Score: 1

    It takes a whole day for this supercomputer to do the same number of calculations as 6 billion hand calculators can do in 24 years? I am not at all impressed.
    If it could do that in something like, umm... a second! yeah, that would be something.

  100. The Military Has Nothing To Offer Gaming? by ireallylovelinux · · Score: 1

    The military has created all the drool over weapons in games like Call of Duty 4, Crisis, and FEAR the military created the whole genre of FPS!

  101. Re:Not in perspective - this is a media number by Anonymous Coward · · Score: 0

    Well, you don't have to try all 2^128 combinations. You could use the general number field sieve to factorize the key exchange. This would probably use much much less time than 10^16 years.

  102. Re:exaflop, zettaflop, the yottaflop and the xeraf by JasterBobaMereel · · Score: 1

    "Why is the military so much more progressive (with practical results) than any other institution of government?"

    They are organised, well funded, and have a definite goal in sight

    --
    Puteulanus fenestra mortis
  103. Re:Why the slashdot downtime? by sherpajohn · · Score: 0, Offtopic

    Genesis turned into a bland pop act 2 albums after Peter Gabriel left. But I think you know that.

    --

    Going on means going far
    Going far means returning
  104. What about actual performance? by impskizzle · · Score: 0

    So what about actual benchmarks? How many FPS do they get in Crysis?

  105. difficult to program. by flaming-opus · · Score: 1

    Yes, this machine is difficult to program, but that is true of all capability class supercomputers. This one will be a little bit more difficult than others, in so much as it is not a simple evolution of an existing design. However, even when you upgrade from 1000 nodes of power5 to 2000 nodes of power6, you still have to do a lot of tuning to your codes to get them to run well. That's the rule, not the exception, in large supercomputer instalations.

    This machine is very expensive. I would suspect that more traditional designs could hit a petaflop for less money. The must believe that this programming model will perform well for several of their critical codes, or they would not have spent so much money.

    Furthermore, since the machine is so expensive, and will be available for open research for only a limited time, one can assume they are only going to run a handful of codes on it. They have probably already taken the time to optimise those codes for the wierd architecture.

  106. Hardware question by Anonymous Coward · · Score: 0

    Does this Road Runner have the hemi or the 440 6 pack?

  107. big iron by flaming-opus · · Score: 1

    from the standpoint of supercomputers, the big SMP systems from SUN/IBM/HP are not big-iron. They are baby sized. They are also machines design differently, to solve a different set of problems.

    The big sun/HP servers are designed to host enterprise-sized databases, supply-chain/ business-intelligence / operations server jobs. They are generally highly parallel transaction processors, not running parallel compute tasks. This doesn't make them easier to design or build, far from it, but it does mean that the requirements are different. You will notice that linux is just now making inroads onto the really large business SMPs. These vendors, as well as the peoplesoft/SAP/oracle/novell/websphere/etc-types have spent decades and billions of dollars developing software that will efficiently use those monster machines. They're pretty amazing systems, but they are not supercomputers.

    1. Re:big iron by jacquesm · · Score: 1

      I realize that they can't do vector ops (or at least not with any speed that would qualify as 'impressive').

      Just for kicks I just priced out a HP machine with only 16 dual core cpus and 64 G of ram and it topped 1 million bucks!

    2. Re:big iron by flaming-opus · · Score: 1

      yes, that's the other thing about supercomputers, they have to compete with commodity clusters, so are quite inexpensive compared to big servers. You'll notice that IBM sells many lines of technical computers including opteron clusters, cell based accelorators, Blue Gene, and also the power6 clusters. The power6 clusters are essentially a bunch of big-iron unix servers ganged together on an interconnect. They work pretty well, but are very expensive.

      It's not really vector ops that makes the difference. The big MPP systems like blue Gene, Cray XT, and SGI altix don't use vector processors either, but they scale to many thousands of processors, not a few dozen.

    3. Re:big iron by jacquesm · · Score: 1

      what's the cause of that huge price increase? I can't seem to figure it out, the premium over my guess of what such a system would cost (about $100 K was my expectation) is enormous.

  108. Greetings, Professor Falken by raguirre · · Score: 1

    "An American military supercomputer, assembled from components originally designed for video game machines"...and nobody made a single WarGames joke? Slashdot user base seems to be getting way too young.

  109. It runs the Moab scheduler by Anonymous Coward · · Score: 0

    FWIW, here's the software it uses:

    XCAT + Warewulf (now named perceus) for imaging/diskless booting
    http://xcat.sourceforge.net/
    http://www.perceus.org/portal/

    Linux 2.6.18 kernel

    Torque as the resource manager : http://www.clusterresources.com/pages/products/torque-resource-manager.php
    Moab Cluster Manager as the scheduler : http://www.clusterresources.com/pages/products/moab-cluster-suite.php

    The Moab scheduler is on four of the top ten super computers as of November 2007 : http://insidehpc.com/2008/01/21/the-411-cluster-resources/

    Source: http://www.lanl.gov/orgs/hpc/roadrunner/rrinfo/RR%20webPDFs/RRSysMgt-with-LAUR.pdf

  110. In related news .... by PPH · · Score: 1

    .... the RoadRunner system has been put to work by the DoD on the problem of developing a successful Iraq strategy. It should be completed by McCain's second term in office.

    --
    Have gnu, will travel.
  111. Re:exaflop, zettaflop, the yottaflop and the xeraf by CAIMLAS · · Score: 1

    Why is the military so much more progressive (with practical results) than any other institution of government?

    The military gets results because, unlike the rest of government, it doesn't exist through simple self-perpetuation, and must attempt to justify its existence. At its core, military serves as the national immune system, protecting the body. If it does not perform, the body it serves ceases to exist.

    In contrast, most government exists to serve itself; a lack of performance is simply indicative of government performing as it was always meant to perform: to serve those in power. The significance of failure or success is irrelevant, provided the politicians are still able to promise something progressive come next election so as to grant them more power.

    --
    ~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
  112. Re:exaflop, zettaflop, the yottaflop and the xeraf by jdgeorge · · Score: 1

    There are numerous other scientific and social areas in which the military advances society, with far more practical results than do-gooders in other government or public institutions. This makes no sense. I suppose there may be some truth to it depending on what you're talking about, but this unfounded blanket statement that the military's "results" are far more practical than that of the unnamed "other government or public institutions" is meaningless.

    I am generally a supporter of the military institutions in the US, but I find the vague assertion that the military advances society more practically than other government institutions ... unlikely. For example, consider the following questions:

    Specifically regarding computing, how are the military's results mre practically beneficial to social or scientific areas than, say, the Department of Energy? How do the results of the military outstrip the results of the Centers for Disease Control and Prevention? In what ways does the military advance science and society more than the National Science Foundation? What results does the military provide that are more practical than those of the Department of the Interior?
  113. Re:Not in perspective - this is a media number by swillden · · Score: 2, Funny

    That's assuming you have to try every single key to break 128-bit SSL encryption. Is that actually the case?

    Unless the 128-bit cipher being used is weak, that is the worst case, and the average case is that it takes half that long.

    --
    Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
  114. How Your Body Works (by A. Nother Doctor) by Rui+del-Negro · · Score: 1

    Thomas P. D'Agostino, the administrator of the National Nuclear Security Administration, said that if all six billion people on earth used hand calculators and performed calculations 24 hours a day and seven days a week, it would take them 46 years to do what the Roadrunner can in one day. The human body is indeed a wonderful thing. Its infinitely complex way of functioning would take a computer, working flat out, day and night, excluding Bank Holidays and Christmas, 3,971 years to work out. The slightest flicker of the eyelid, the smallest movement of the big toe, involves such extraordinarily complex processes that the average man, working flat out, excluding Bank Holidays and Christmas, but including weekends, would take 84,643 light years to work it out. If you can imagine an Airedale terrier jumping in and out of a watering can once every 7 minutes for 12 years you have some idea how long that would take. And that's only one light year.

    Even the most simple process that the body can perform -- like paying the doctor -- would take a piece of asbestos over 9 billion years to work out. If you can imagine a man at a cocktail party congratulating the hostess on the avocado dip 40,000 times every second for 2 1/2 hours twice a week for 28,000 years you can begin to realise what an extraordinarily wonderful thing the human body is.

    To put it even more simply, if you can imagine a doctor leaving his lucrative Harley St. practice to a younger partner, and cruising round the world 4 times a year, drinking 3 bottles of champagne with a friend's wife every afternoon, and writing an article on How Your Body Works once every 96 days, you'll get some idea of why I was struck off the register. Good evening.

    (from "Monty Python's Brand New Papperbok")
  115. Re:exaflop, zettaflop, the yottaflop and the xeraf by anaesthetica · · Score: 1

    The emphasis in my post is on practical results, differentiating the military from other institutions by its consistent focus on mass implementation rather than simply research. Notwithstanding the immense research and development budget controlled by the various military and defense institutions in the U.S., the other agencies you cite are focus principally on research, statutory and regulatory law change, and expert-level implementation of their various advances.

    This is in contrast to the military, whose social and technological innovations require propagation en masse, either throughout military society (which then has knock-on effects to the rest of society) or on a mass usability basis (universal communication frameworks or mass produced technology available to the individual enlisted man or woman on the front).

    Because the military's focus is on the logic of necessity and the logic of life-and-death, they are often more able than other agencies to cut through the Gordian knot of domestic political factions, ideologies, and interests. For example, this allowed them to justify racial integration well ahead of most public and private institutions in U.S. history, despite prevailing political conflicts and racist ideologies.

    So while DoE does energy research and proof-of-concept designs, the military makes the practical push to make alternative energy use a reality, in part because they are driven by a logic of necessity, and in part because they are perceived as less political than the DoE. Similarly, the CDC does a tremendous amount of research on vaccinations, cures, and epidemiological health policy. On the other hand, the military makes the practical push to get easy-to-administer, cheap-to-mass-produce vaccines and treatments available, because they're necessary to the strategic logic of warfare--witness the history of penicillin for example.

    The military certainly is no substitute for the other organizations that you cite. But in terms of practical effect on a mass scale, the military is more often than not on the leading edge.

  116. Re:Not in perspective - this is a media number by swillden · · Score: 1

    Well, you don't have to try all 2^128 combinations. You could use the general number field sieve to factorize the key exchange. This would probably use much much less time than 10^16 years.

    SSL uses two ciphers for each communications link: a public key cipher and a symmetric cipher. A random symmetric key is selected and used to encrypt the actual data. A public key is used to encrypt the selected symmetric key to deliver it securely to the other end.

    The "128-bit" cipher in an SSL exchange is the symmetric cipher, not the public key cipher. A 128-bit public key would be horribly weak, crackable in minutes, if not seconds, using GNFS so 1024-bit and 2048-bit keys are normally used. The 128-bit symmetric cipher (commonly AES) does not use composites of large primes in its keys and is not amenable to attack with the GNFS. The best known attack on AES128 is a brute force search through the keyspace.

    --
    Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
  117. Broken window fallacy by slysithesuperspy · · Score: 1

    Last I heard they blew stuff up. You may see impressive results, but you are looking at what you can see. So, what about all the things that would have been made that consumers were actually willing to pay for? How many things haven't been invented and created in the private sector that would have been if loads of scientists, engineers, money and other resources weren't ploughed into making stuff that blows stuff up? A net loss is hardly progressive, sounds rather regressive to me.

    http://en.wikipedia.org/wiki/Parable_of_the_broken_window

  118. Re:exaflop, zettaflop, the yottaflop and the xeraf by jdgeorge · · Score: 1

    I understand the sense of what you are saying; the military, by necessity, puts things into practice rather than engaging in what one might characterize as "academic research".

    However, you have still not offered any specific evidence that the social or scientific benefits of the practical application of research by the military are greater than the practical applications of research by other government entities (except as a result of scale) or that the investment in academic research by other government entities does not produce substantial scientific or social results.

  119. Re:Not in perspective - this is a media number by Anonymous Coward · · Score: 0

    Ah, but tell me this: How many schoolbuses is it from end to end? How does the size of one transistor compare with the width of a human hair? If the cooling systems fail, will it get many times hotter than a pizza oven?

    I think answers to these questions will go a long way toward helping us understand how this thing works.

  120. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  121. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    "BlueGene/L--first on the TOP500 list of supercomputers with a sustained world-record speed of 478.2 teraFLOPS--is a revolutionary, low-cost machine delivering extraordinary computing power for the nation's Stockpile Stewardship Program."

    "Established in 1995, the Advanced Simulation and Computing (ASC) Program supports the Department of Energy's National Nuclear Security Administration (NNSA) Defense Programs' shift in emphasis from test-based confidence to simulation-based confidence."

    Very funny way to say: "we use this enourmous computers with extremely difficult programming to test our aging nuclear weapons."

  122. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    Military taking the lead on computing as usual. Why is the military so much more progressive (with practical results) than any other institution of government? In this case, because they have 133 million dollars, need much processing power for one program and have only a handful of very big research installations.

    So, it also has Cell-based processors AND Opterons. I wonder what the functional division between the two chip types is? Probably they run housekeeping, I/O and part of the calculations on the Opterons, and only the calculations on the Cells.
  123. Re:exaflop, zettaflop, the yottaflop and the xeraf by Overclocker · · Score: 1

    Military taking the lead on computing as usual. Just for the record, Los Alamos is a civilian laboratory run by the Dept. of Energy. It's not a military laboratory (i.e., run by the DoD) -- despite what the New York Times might say.

  124. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    This computer has nothing to do with practical results. There is no (or almost no) software that will run on it. It is simply a political stunt used to give LANL/DOE a big number; peta-hype. This computer is not going to solve any real world problems, and LANL has put zero resources into the very significant software engineering that would be required to actually make it useful. Go ahead and lust, but its an even bigger piece of crap than past LANL supercomputers like "Q" and "BlueMountain".

  125. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    Well, yes, not as much need to justify activity, or rather, easier to justify because it targets individual and collective self interest directly.

    But the real motivator is strategic:- the first one to the top controls the game.

  126. Re:exaflop, zettaflop, the yottaflop and the xeraf by Anonymous Coward · · Score: 0

    Yikes. I don't see why this link feels the need to inflate American military spending; the actual/factual numbers stand for themselves. tsk, tsk.