Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Hardware Software Linux

SAPAC Unveils New Australian Supercomputer 227

Sean Burford writes "The South Australian Partnership for Advanced Computing (SAPAC) has unveiled its new AU$1.7 Million supercomputer named Hydra. It is an IBM 1350 Linux cluster with 126 compute nodes (xSeries 335), 1 head node (xSeries 335), 1 storage node (xSeries 345) and 1 managment node (xSeries 345). Hydra has a peak theoretical performance of 1.2 Teraflops, and has currently benchmarked at 682 Gigaflops. The current benchmark places it in the fastest three supercomputers in Australia and equivalent to the current number 80 in the world. The cluster has a total of 258 2.4Ghz Intel Xeon processors and 258GB of RAM. SAPAC expects to achieve a benchmark closer to 700 Gigaflops with further tuning. Hydra is hosted at The University Of Adelaide, who already host a 40 node cluster of Sun e420 machines."
This discussion has been archived. No new comments can be posted.

SAPAC Unveils New Australian Supercomputer

Comments Filter:
  • Nahhh, that's not a computer. Now this [top500.org], this is a computer.

    Mike
  • by Anonymous Coward
    as powerful as "the Gibson"?
  • "The cluster has a total of 258 2.4Ghz Intel Xeon processors and 258GB of RAM."
    258, hm? Izzat metric or sump'n? Maybe it's like that feet/metres thing with that Mars probe. Or is it like how AMD numbers their chip speeds? Is it a Southern hemisphere localised effect perhaps?

    When someone explains this to me I'm going to feel mighty small. Possibly 1/258 of my current stature.
    • How to Speak Australian:

      In Australia, even 2^8 is bigger.
    • by Anonymous Coward
      129 nodes * 2 CPUs = 258 cpus
      129 nodes * 2GB = 258GB (actually 129 * 2048MB)
    • by sould ( 301844 ) on Friday June 06, 2003 @01:08AM (#6129744) Homepage
      Dunno about the processors - I dont see any reason why you'd need a power of two for them. Perhaps we've got two arrays of 2^7 processors with a controller processor each (=128+1 *2 = 258)

      But I suspect as far as the ram goes that the 258 gigs is 256 - but counting 1k as 1000 instead of 1024. (or possibly 1M as 10^6 instead of 1048576)

      Haven't you noticed the difference between what a vendor says is the size of a HDD compared to how many gigs you actually get when you put it in your PC?
      • Thanks, I can picture that now. I was trying to picture the configuration but couldn't.

        I think I'll stick to posting during the daytime from now on. A number of other people responding right now might need some sleep as much as I do - people sure get cranky in these wee hours!

        Yeah I'm aware of the 10^3 versus 2^10 difference. My 80Gb drive fromatted to 76Gb or so.

        Anyway, thanks for the explanation.
    • *It is an IBM 1350 Linux cluster with 126 compute nodes (xSeries 335), 1 head node (xSeries 335), 1 storage node (xSeries 345) and 1 managment node (xSeries 345).*

      126 Nodes + 1 Head Node + 1 Storage Node + 1 Management Node = 129 Nodes total

      129 Nodes total * 2Gb ram per node = 258Gb ram for the cluster
  • Is this news? (Score:5, Interesting)

    by KFury ( 19522 ) * on Friday June 06, 2003 @12:47AM (#6129652) Homepage
    Not to flame or troll, but considering that over 90% of the top 80 came out in the last 30 months, how big a deal is this? Third fastest computer in Australia? Sheesh.

    A computer faster than this is born every two weeks.
    • Hey, if you compare this to the state of our television talent (ie. Steve Irwin...bleh) then this is world class stuff...
    • Re:Is this news? (Score:5, Interesting)

      by PerryMason ( 535019 ) on Friday June 06, 2003 @01:57AM (#6129906)
      Its really nothing huge, but you also need to consider the cost involved. AU$1.7 million is about US$1.1 million. So for about a million US$1 you could get in the top 100 supercomputers in the world.

      Looking at the latest top 500 list [top500.org] this would put it as the third most powerful 'self-made' system in the world. For that reason I think it deserves at least a mention and add the relatively low cost and you've got a /. story.
      • Re:Is this news? (Score:4, Informative)

        by KFury ( 19522 ) * on Friday June 06, 2003 @02:30AM (#6129993) Homepage
        Looking at the latest top 500 list this would put it as the third most powerful 'self-made' system in the world.

        I disagree. Hydra would qualify as "IBM made" as it uses a standard network of standard IBM machines. For comparison, the top 'self-made' computer, Sandia's Cplant Cluster, was built by the lab from off the shelf components, wired together by custom drivers written by the engineers at Sandia.

        If the thrust of this story was an amazing $AU/tflop ratio, it didn't come across at all in the summary.
      • Re:Is this news? (Score:3, Insightful)

        by sql*kitten ( 1359 )
        So for about a million US$1 you could get in the top 100 supercomputers in the world.

        For $1M you could buy a load of computers, sure. But what is the cost of a building to put them in? A while ago I did some consulting for a major telco/colo provider. Their single biggest expense was electricity to run their air conditioning and dehumidifiers. It cost more than renting the building. They were seriously considering buying a utility company to get a better rate on electricity. Also on the cards was a reloca
  • This is nice, but it's in Adelaide, South Australia. The same state that was first to decide that uploading adult content to the Internet, by FILM STANDARDS, is a jailable offense. Ie, upload a Mills and Boon-style book, go to jail.
  • by ewhenn ( 647989 ) on Friday June 06, 2003 @12:48AM (#6129658)
    A spokesperson for SAPAC in a recent press interview has stated that the intent is to create the largest SIMS server on the internet.
  • So this is what the Simpsons meant by Australia's giant boot.

    --Pat

    • by Anonymous Coward
      An American cartoon makes a joke about the Australian government booting a kid in the arse.

      Yet, Australia has outlawed any form of corporal or capital punishment, but the US still lets teachers hit kids and kills people with death sentences. This isn't the pot calling the kettle black, it's the pot calling the fine cutlery black.

      Australia has sane, civilised laws. The USA kills people, hits people, and arms its citizens to the teeth with guns.

      Ah, the irony of it all.
      • Wait a minute... these days I don't think anyone tolerates kids being hit (if not for moral reasons, because they don't want to end up on the news or on NBC's Extra with an expose' or some shit). Also the U.S. doesn't actively "arm" people, it just provides a 230-year old ambiguously-worded law that seems to allow everyone the right to own firearms. Eh, what can ya do.

        But yeah, we do cook people.

        [/completely offtopic]

      • An American cartoon makes a joke about the Australian government booting a kid in the arse.

        Geez, it was a cartoon. A parody. A caricature. The great thing about the Simpsons is that no group is spared from their biting satire.

        Remember this is the same cartoon that shows that all nuclear workers are inept and cause meltdowns all the time. The same show whose police force can't find sand on a beach, where the male father character spends most of his time drinking in a bar, and school children eat exerci

  • by Anonymous Coward on Friday June 06, 2003 @12:50AM (#6129664)
    be very, very careful around this one mate.
    It is very, very dangerous.

    Look at the size of the heatsink on that one!

    And this buggers attck fast. And I mean real fast.

    Crikey!
  • GB vs. MB (Score:3, Interesting)

    by lingqi ( 577227 ) on Friday June 06, 2003 @12:50AM (#6129667) Journal
    ...and 258GB of RAM...

    Is it me or anyone else misread it as "256MB"?

    Actually, misreading it lead me to think about a mainframe at my college, which was an SGI with 12 processors and 512MB of memory.

    The thing is, though - when I first went in the college, we were all like "WOW that's a lot of system resources." When I got out four years later I was carrying that much memory on my laptop...

    breakneck speeds, man.

    However, regardless - (with all due respect) why is this such a big deal that australia limped to #80 on the fastest computer list? didn't other linux clusters break teraflops quite a long time ago? EarthSim was neat because it put THAT much more distance between another country and the US (and nearly nobody saw it coming) - but this seems to me hardly news, besides the possible "one of the fastest computer in australia runs linux," or something...

    • i don't know man, 12 procs still sounds like a lot of resources to me :p

      and if you think about it, most "normal" people (and i use the term loosely) only carry about 256MB or that there RAM.
  • How long... (Score:5, Funny)

    by Esion Modnar ( 632431 ) on Friday June 06, 2003 @12:51AM (#6129674)
    ...til somebody asks it, and it replies:

    42.

    • by Anonymous Coward
      I would think the Earth Simulator would be the better one to ask ;)
    • So few understand the humor in this.
    • ...when you multiply 6 times 9?
    • That happened long ago, actually, you're currently living on the supercomputer (earth) that is computing the actual question ;)
  • I would have thought there had been a T-Mobile Sidekick cluster running Linux by now? ::sigh::

    Dolemite
    ____________________
  • yeah but (Score:5, Funny)

    by toddhunter ( 659837 ) on Friday June 06, 2003 @12:53AM (#6129682)
    They are still going to have to upgrade when Doom 3 comes out
  • by MrCreosote ( 34188 ) on Friday June 06, 2003 @12:55AM (#6129691)
    At least now there is one thing that goes fast in Adelaide.

    • Explanation (Score:1, Informative)

      by Anonymous Coward
      For those outsode Australia:

      The state government where this is housed has decided that anyone doing over 3km/hour over the posted speed limit, is fair game for speed cameras and fines.
  • Its time... (Score:2, Informative)

    by thogard ( 43403 )
    Here's the museum peice [abc.net.au] its going to replace.
  • by leereyno ( 32197 ) on Friday June 06, 2003 @01:06AM (#6129736) Homepage Journal
    It really is too bad they can't use Athlons.

    The per-clock performance on an Athlon is much better than what you'll get from a P4 based Xeon, and that is just on integer. When it comes to floating-point performance a lower clocked Athlon will meet or beat the performance of a higher-clocked P4.

    Right now the only SMP chipset for the Athlons is the 761, which is several years old and lacks dual-channel capability. It also requires the use of registered ECC memory. If the Athlon's had an SMP chipset comparable to the NForce2 or Intel's 775 then it would be a very different story.

    Right now the going rate on pricewatch for an Athlon 3000 is only $10 more than a 2.4 Ghz Xeon, and it would spank that Xeon on floating point which is exactly what is important for a supercomputer.

    I hope that the clustering technology they're using makes good use of SMP systems because if it doesn't then they may very well have misspent their money.

    Lee
    • by Anonymous Coward on Friday June 06, 2003 @01:11AM (#6129755)
      They must be worried about Global Warming.
    • by sjames ( 1099 ) on Friday June 06, 2003 @01:33AM (#6129833) Homepage Journal

      Actually, Athlon is not faster than the P4 Xeons at double precision floating point. Double precision is frequently required for scientific computing and is used in the Linpack benchmark for the top500.

      The current dual Athlon chipset is the 760MPX. The Intel i7501 is the preferred chipset for dual P4. It supports dual channel ECC DDR ram and the 533MHz FSB. These days, nobody wants non-ECC ram for a top 500 cluster. It's not that much more expensive these days compared to decently rated non-ECC ram.

      Then there's the cost of air conditioning 258 Athlons...

      I'd say they made the right decision.

    • by MetricT ( 128876 ) on Friday June 06, 2003 @01:37AM (#6129850)

      We're just a few days from bringing up a 300 processor cluster [vanderbilt.edu] of the exact same type of computers they are using, so maybe I can shed some light. There are several reasons for picking Xeons over Athlons at the moment.

      1. If your app uses double precision floating point, and you can recompile your app using SSE2, an Intel will easily beat the AMD. AMD does scalar floating point operations faster per clock. Intel does vector flops faster. Most interesting real-world problems use vector flops.

      2. Memory bandwidth. Most chipsets can only deliver a fraction of their theoretical bandwidth. I've seen speed differences of 25% running code on identically configured machines, one having Intel E7500 and the other with a ServerWorks GC-LE (the ServerWorks smokes...) And those are *good* chipsets. I have yet to see an Athlon chipset that wasn't crap.

      3. Managability. The x335's are pretty damned slick. I *love* the built-in KVM switch and remote diagnostics. You can daisy chain north of 21 nodes together (I think 35!) and you just have one cable coming off of them.

      4. Total cost of ownership. Our previous p3 cluster was assembled (before I arrived) from Pricewatch parts. We initially experienced a 25% failure rate on memory, and spend an inordinate amount of time fixing random problems. 40 of the p3 nodes takes more than three times as much administrator time as 160 IBM x335's. Spending an extra $50,000 on good, quality parts is cheaper than hiring a competent sysadmin. Don't "efficient" yourself to death.

      Having said all that, I'm *really* looking forward to Opteron. We're getting some in a week or so. 64 bit + SSE2 support is going hard to beat.

      • Having said all that, I'm *really* looking forward to Opteron. We're getting some in a week or so. 64 bit + SSE2 support is going hard to beat.
        This is -specifically- why I'm waiting until the next Opteron stepping before specing out our next cluster.
        The current rig is P4 Xeons (1.8GHz), and is "ok" performance wise, but that's ok, since it was more a "proof of concept" type of thing than a real install *heh*
    • by mc6809e ( 214243 ) on Friday June 06, 2003 @02:30AM (#6129992)
      It really is too bad they can't use Athlons.

      The per-clock performance on an Athlon is much better than what you'll get from a P4 based Xeon, and that is just on integer. When it comes to floating-point performance a lower clocked Athlon will meet or beat the performance of a higher-clocked P4.


      What you see as an advantage for the Athlon is actually a disadvantage.

      The Athlon is trying to do too much per clock and this limits its maximum clock rate. What matters is realized performance. Right now, less work at a higher clock seems to be pulling ahead.

      It really comes down to how large you can make the product (work x clock rate). Less per clock isn't bad if it means you can greatly increase clock rate.

    • It really is too bad they can't use Athlons.

      They were IBM xSeries boxes - they don't (that I'm aware of) offer a x86 version with AMD. Had this been sponsered by someone who did do AMD servers, you would have seen Athlons in there. This was an IBM gig - and they had xeon's in inventory. Not a bad bit of kit, btw...

      You are correct about the floating point, however. For chipsets, look for the AMD-8131 in boards like this [msi.com.tw]. Most of the new workstation class boards from AMD will be based on the Opteron
  • by teamhasnoi ( 554944 ) <teamhasnoi@CURIE ... minus physicist> on Friday June 06, 2003 @01:10AM (#6129752) Journal
    they *still* can't get sound to work, and their window manager crashes every time they play TuxRacer.
  • So what's the difference between the Head Node and the Management node?
    • Management never uses its Head. ;-)
    • Head Node:
      Cluster users connect to this node to kick off their jobs onto the rest of the cluster. Basically, it's the enduser accessible machine.

      Management node: cluster monitoring tools, image server, etc.

      On a -small- cluster (16 or 32 nodes, although 32 is pushing it, IMHO), the head and management nodes are typically the same machine. There really isn't a need for there to be more than that, as it's fairly quick and easy to monitor all of 32 machines. On larger clusters (64/128/etc), it starts becom
  • by femto ( 459605 )
    For Immediate Release:

    IBM reports that the Univeristy of Adelaide has returned its recently purchased IBM 1350 Linux cluster.

    According to Mr. Ian B. Myers, an IBM technician, the cluster landed on the doorstep at White Plains, NY with a note attached reading "This bloody thing doesn't work". On testing, it was discovered that every node had been named bruce...

  • by lachlancs ( 662690 ) on Friday June 06, 2003 @01:18AM (#6129789)
    The cluster is running IBM Cluster Systems Management, not Beowulf, and is using Myrinet Networking.
  • How far third-world countries like Argentina are from the real world. Or what I want it to be the real world. Chord.wav, the sound of innevitability.
  • by Anonymous Coward
    They didn't mention which ISL interconnect they were using for the (presumably) Linpack benchmark number of 682 Gigaflops, but it would be interesting to see a full description of their equipment. I run an identical 1350 system, 126 x335's with 2 x345's for management, with a Myrinet (http://www.myri.com) switch and I've been averaging in the low 600's. And IBM signed off on that as being the "practical" maximum for the cluster. Hmm. (Reaching for the phone...)
  • Looks like the Aussies were looking for a way to try out the slashdot effect on slashdot.

    Either that or they're loaning it to New Zealand to study and see if banning intercourse with sheep increases their country's population.

    I somehow suspect the latter.

  • I wonder what kind of frame rate they are getting for Unreal Tournament...
  • even more! (Score:2, Interesting)

    by golgotha007 ( 62687 )
    an even more interesting supercomputer!

    Japan's Earth Simulator Center [jamstec.go.jp] has 10 TB of main memory and the theoretical performance of 40Tflops!

    quite a system and definetely worth checking out.
  • Coming in second is being the first loser, so what is coming in third?
  • Isn't that what an impotent Bin Laden would suffer? Or is that any impotent person? Maybe its fear of 'flopsy'?
    *sigh*
    its the end of the week and I'm going home in four minutes. What is this crap? In other news, Germany has built the 90th largest ship in the world, the US has finished construction on the thirteenth smallest... you get the idea.
  • When did a cluster become a supercomputer? It may be a powerful cluster, but that doesn't make it a computer. In singularis, see?

    Put Apache on it and we'll /. it in mere seconds.

  • it's cooled by beer.. and is powered by splitting the beer atom [imdb.com]
  • by CaptainPotato ( 191411 ) on Friday June 06, 2003 @03:48AM (#6130175) Homepage
    ...will do almost anything for a bit of publicity at the moment. As the eighth university of the so-called 'Group of Eight' premier league institutions in the country, the University of Adelaide is a mess at present. There is little leadership, poor strategic planning, a recent Government survey that slammed it in many areas, and so on. The announcement of this new system is meant to show how technologically advanced the place is, and how it leads the way.

    This couldn't be further from the truth. Ask the postgraduates, who have Mac LCIIs and 486-DX33s on their desks (I kid you not). Ask the academics, who have been retrenched in recent years (in some facultis, 25% of academic staff lost their jobs because of the university's financial problems), ask the users of its library, which has HUGE funding problems.

    Whilst the new machine may be very nice and have some power, the University of Adelaide really sees it only as a PR campaign (hell, it even made it to Slashdot!), rather than anything significant for the sake of scientific advancement - okay, the researchers, who will use it may have a different opinion, but not the University iself.

    • In Heaven music is English, girls Australian, beer German.In Hell beer is Australian, girls English, music German

      Hey, I take offense. Ausatralian beer is OK, english girls are fucking awesome (what is that cute kurvy cook-girl we see on the ABC?), and german music is great (Alex deLarge, Ramstein)
  • is the APAC Alphaserver SC located at the Australian National University in Canberra and is 63rd on the top500.org list at 825.50 GFlops Rmax http://nf.apac.edu.au/facilities/
  • by Mr_Glooby ( 673458 ) on Friday June 06, 2003 @04:32AM (#6130274)
    Well, I live in Australia, and note the confusion caused by the number 258, so here is why...
    The number between 5 and 7 may not be spoken here because of its similarity (when spoken with a New Zealand accent) with a certain act often carried out between mammals.
    In fact a whole new mathematical system is being developed (based on pictograms) to avoid political insensitivities in the Land where legislation has been passed to the effect that children access the internet, and so the internet must not contain content (including numbers) that might corrupt the young.
    The pictogram for the number that dare not speak its name is an image of two trees. (tree and tree is s.x).
    Similarly, three trees with 'watermarks' (evidence of the recent passing of puppydogs with full bladders) represents the number 99. (dirty tree + dirty tree + dirty tree = 99).
    The same pictogram with underscores (here called doggy doo-doos) represents 100.
    (dirty tree and a turd, dirty tree and a turd, dirty tree and a turd...)

    Besides, our new supercomputer sure beats the 286 we've had to share for the last 10 years! Them Y2K problems are getting to be a real pain!
  • Supercomputers (Score:2, Interesting)

    by [cx] ( 181186 )
    When will the speed stop mattering?
    What is the theoretical speed of 0 latency for computations?

    This may seem like a stupid question, but I never heard once in star trek them saying our computer is such and such fast. They must have reached a limit that allowed them almost instant computation.

    So what would that be in our measured terms however primitive they might be in the longterm outcome of our computers?

    10000 Ghz? 1 Million ghz?

    I dont know,
    any biters on this bait?
    • Supercomputers are usually rated in FLOPS like this one is in the article. FLOPS = FLoating Point Operations per Second. So, it's the number of calculations it can do per second. More processors means more theoretical FLOPS. The speed will stop mattering when we can simulate the entire universe in real-time. At which point, I imagine, the universe will stop mattering.

      This has nothing to do with Ghz. Processor speed is irrelevant insofar as the speed of calculations is what really matters.

  • Go Oz... (Score:2, Funny)

    by Cackmobile ( 182667 )
    Aim high, we build this whiz bang thing and its only No. 80 in the world. What happened to Aussies innovation. We used to be able to build these things out of old fosters cans.

    • Re:Go Oz... (Score:2, Informative)

      by mallie_mcg ( 161403 )
      Aim high, we build this whiz bang thing and its only No. 80 in the world. What happened to Aussies innovation. We used to be able to build these things out of old fosters cans.

      I Call you out, you are not an Aussie, for fucks sake mate, no aussie will willingly drink fosters beer, let alone in cans, and those that do, well they revoked their citizenship.

  • Supercomupters and Olympic Arnold Schwarzeneger stadiums (in Graz).

  • CDAC of India already has developed 1T flops super computers [business-standard.com] based on SUN Ultrasparc-II cpus. The system has primary storage capacity of 5 TB. The communication backbone can be any of CDAC's own PARAMNet at a peak bandwidth of 50 MB/s bi-directional, Myrinet at 160 MB/s, ATM at 155/622 Mb/s, or Fast Ethernet.

    Currently work is going on to make a 10T Flop grid across country linking all premiere research and educational institutes and industrial establishments.

  • The record holder is 35 teras Japan Earth Simulator. IBM ASCII-Q will be 100 teras soon. NOAA turned on a 7.3 tera compter today (NY Times). Last week the DOE hooked together a bunch of game-cubes to make a $50K tera computer.
  • Bruce: Hey, Bruce, we need something to count all these sheep.

    Bruce: Yeah, Bruce - how about this bonzer cluster of Xeons over here?

    Bruce: Nah - it'll never work - it's only 1.2 teraflops, and the cooling system sucks.

    Bruce: Yeah - you're right there, mate - we'd better go with the Cray again. All that Freon sure keeps the tinnies cool.

    Bruce: Now that's what I call a sheepacomputer.

Eureka! -- Archimedes

Working...