Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Linux Software

Linux And Los Lobos Supercomputer 89

Posted by CmdrTaco from the but-the-real-question-is-how-many-bogomips? dept.
DocRea writes: "Using Linux, IBM and the University of New Mexico will connect 256 two-processor IBM Intel-based servers with high-speed Myrinet cards to create a 512-processor machine capable of 375 billion calculations per second. The computer, called Los Lobos, will primarily be used for scientific purposes, but will be adapted by IBM to provide the "cluster" approach to running software for business tasks and e- commerce. "
This discussion has been archived. No new comments can be posted.

Linux And Los Lobos Supercomputer More

Linux And Los Lobos Supercomputer

Comments Filter:

  • don't need SMP to cluster... (Score:1)

    by Anonymous Coward
    As I said, you don't need SMPeabilty to cluster, a bunch of Athlons will do nicely (depending on what you wanna do, maybe you'd want different CPUs: Alphas, PPC, whatever).
    For that matter you can chain anything to anything else with Ethernet and MPI (e.g. LAM or MPICH implementation). I guess for Myrinet (=Gbit-Ethernet) you'd need PCI slots though, I'm not really familiar with that.

    SMP saves
    1) space (+ HW costs)
    2) if you do it right (as in interweaving OpenMP and MPI, quite messy) it saves you latency and gets more bandwidth on communication between nodes. Nothing (as of now) beats the Cray communication technology, certainly not anything having to take the path MEM-> BUS -> PCI -> CABLE (including handshaking) -> PCI -> BUS -> MEM.

    But notice, if you don't need heavy communication, clusters (w./ Ethernet) are fine, such as the NASA computational gravity stuff and cryptocracking e.g.. Other apps, like Lattice QCD (very communication intensive) will have to wait longer, until they get anything comparable to a Cray.

    As for the def of SC: I don't know it, but the definition will be exactly where it matters (as you said): in the law. Anything else is just words, words, words. :)

    cheers,
    Roland

  • If you wanna build a cluster... (Score:1)

    by Anonymous Coward
    How about this [ebay.com]

    This looks like it's cheap enough for the house... And who doesn't wanna run fiber in their house???
  • I'm a-usin' it...and it ain't workin'
  • Oh, cute. *Now* it works :|
  • In the latest episode of "As the karma burns", some 12 or more hours after "Hurrying to post this...." was posted someone apparently couldn't find anything worth moderating up anywhere and went after it instead.
  • Apparently 25 is currently the magic number. Now I've got to find the way to keep *all* my posts from starting at 2 so I don't look like a jerk and attract trolls with mod points.
  • Your comment is 100% rubbish.

    Not scalable? Look at their ccNUMA and CrayLink stuff. Need I say more? Sorry, but your Linux box cannot touch SGI's clustering ability. Maybe in a year or two but not right now.

    Not secure? Again, rubbish. IRIX did have some security problems in the past related to its suid root GUI system administration tools. The simple fix was to not install them. As for newer versions of IRIX, they'll blow the average RedHat box away in terms of security. Go to securityfocus.com and count the exploits for yourself.

    IRIX is not stable? I recently left Ticketmaster Online - CitySearch, where we were ran two large datacenters full of Sun, SGI, and VA Linux systems. I don't think there was a sysadmin among us who would say that our SGI Origin200 servers were not the most stable of the bunch. They've been up for hundreds of days, sending out many gigabytes of web content and running high-traffic Oracle databases.

    Stop being a drone and following your penguin-obsessed friends. Do some comparison shopping and I don't mean read the latest /. posts.

  • Hmm (Score:1)

    by Signal 11 ( 7608 )
    Am I the only one that thought that this was from "Los Lobotomy" labs?
  • Or, better yet. Just buy their parts from Taiwan and avoid the US restrictions altogether.

    (Yes, you can get chips from there.)
  • Now I've got to find the way to keep *all* my posts from starting at 2 so I don't look like a jerk and attract trolls with mod points.

    Actually, there's a checkbox for "No Score +1 Bonus" right above the Submit button. Watch, I'll do it right now...

  • Is this a different story than this one [slashdot.org]?
    --
  • Remember MS's NT clustering software was called "Wolfpack?" Probably unintentional, but an interesting coincidence.
  • Hmm, you might want to go hear Los Lobos before you trash e'm. I saw them less than a month ago, and it was a great show.

    but everything else they've ever done sucks.

    I'd like to see your list of everyting else, unless (and this is my guess) those are the only songs you've ever heard.

    --
  • Are you smoking crack? Los Lobos, the band, rules! The La Bamba soundtrack is the least of their efforts. They only did like three tracks on that one. Their real albums like Will the Wolf Survive and By the Light of the Moon are American classics. They've all but disbanded now, with getting old and the recent tragic murder of Cesar Rosas' (the guitar player) sister. But in their prime, they had total street and star cred. They were pack-the-house headliners and everyone from The Grateful Dead to Bruce Sprinsteen to Sonic Youth wanted them as openers. They rock!

    Hidong
  • What is that +1 limit at now? Has it changed any?
  • Athlon motherboards are already incompatible with Intel processors. The Athlon core just can't do SMP right now, nor does AMD have a SMP chipset.
  • Linux Sombrero Rojo release 6.1 (La Bamba)
    Kernel 0.9 on an i386-washedup
    login:

    Actually, believe it or not, I say Los Lobos last summer. They were pretty good. The did a big jam session that was pretty cool.
  • IBM Creates New Fastest Beowulf Cluster [slashdot.org]

    Too late already posted on March 22
  • You obviously have never seen Hackers or read, and I mean READ, Nineteen Eighty Four.
    Another Anonymous Coward mentioned that aswell, too bad you guys don't have the guts to put a name to your useless posts.
    Like I told him, maybe the movie was not realistic, but if you have ever read the book, you'd know that the quote makes sense.
  • Sorry bout the Dave thing, I am stuck in my Hal mode
  • Dude, that's almost as slanderous as a Microsoft press release! I don't care of it is pro-Linux FUD, it's still FUD, and it isn't welcome. Go play in sid=trolls or sid=trolltalk, or better yet, go play in the street. I'm sure a small city bus will knock the FUD out of you, and if it doesn't, the prolonged hospital stay will keep you away from /. for a while.
  • Lobo is Spanish for wolf, so this name makes some kind of sense for a big Beowulf box. Unless you know what the spanish for Beowulf is. :-o

    Besides that, it's also UNM's mascot.

  • but this was published on wired yesterday... I submitted it about 15 minutes later (which got turned down).

    So it's a repeat of that earlier article, and a late one at that!

    -----
  • I dunno, their Kiko and La Pistola albums were quite pleasing, and they have seemed to draw consistently large crowds...for 25 years. I guess that maybe they're second-rate compared to third-rate bands like the Rolling Stones or Third Eye What-Have-You though.
  • You posted the same story yesterday, here [slashdot.org] on slashdot. It was a different article, but the same supercomputer!

    Check the University of New Mexico press release [unm.edu].
  • We need 4,000 CCs of anesthetics.

    WHat, no anesthetics? Dammnit man, just get a boxer in here! This man needs a Gritectomy, pronto!

    ----
    Don't underestimate the power of peanut brittle
  • Oh yeah I'd like about a dozen 486's sitting around needing tons of space and power. Meanwhile a dual pIII would kill it. Don't forget the cost of gigabit ethernet and switches. Fast ethernet is still no match for craylink or ccNUMA which averages around 600-700meg/sec, and under ideal conditions gigabit is still 120meg/sec. It might be good for cracking seti blocks but moving around tons of data no way.
  • It was "Los Locos" but that was a kick-ass movie.

    "I dunno, man, eet's a... a Meetsubishi!"

    "Am not Mitsubishi! Am Johnny Five!"

    "Collecting mucho input! Urban input!"

    Los Lobos was the name of a Latin band from the eighties; they are responsible for reviving "La Bamba".
  • Hehehe, I thought it was an obscure reference to Short Circuit II, where there was a gang called Los Lobos. Y'know, "Los Lobos kick your ass, Los Lobos kick your face, Los Lobos kick your balls into outer space!".... Still, it just means "The Wolves" so it would be pretty common. Especially reasonable since some people wer saying that its the school team name.
  • Or if no-one asks if it runs Linux, someone is going to ask if one could build a beowulf cluster out of them.

    You can't underestimate stupidity.
  • My Beowulf [amazon.de] DVD (see also here [imdb.com]).
  • Thanks for bringing that to my attenition, i had the site in my bookmarks and never really went to the front page. Sorry, and thanks
  • I dont think these two posts are tlaking about the same box
  • I've been mulling this over for the last couple weeks, ever since I dug through everything I could find online about the Beowulf linux project. Clustering looks like something that's going to revolutionize computing, but wouldn't it be truly interesting if we could find a way to cluster dissimilar equipment efficiently? An office could take all of their old equipment, 486's Pentium's Pentium II's, and throw them all in one gigantic cluster to harness the power of all the obsolete silicon. If this is an impossible dream someone PLEASE wake me up...
  • Just for the record, the UNM mascot is the Lobo. Everything here is `Lobo this' and `Lobo that'.
  • I've already tried to network my 386 and my Macintosh SE into one supercluster but the project flopped.
  • What I don't understand is: Why do they use computers with 2 processors, in stead of four, I mean; every cluster fax states that SMP is much faster than clustering. So why not first exploit SMP to the max and then cluster. so, yes "Les Lobos es Loco".
  • ehh, faq, not fax...sorry for the typo
  • The UNM sports teams are The Lobos, which is Los Lobos in Spanish. New Mexico has a large Hispanic population so they were just trying to fit in.

    But the band was the first thing I thought of too and I live here. :)
  • The first couple of posts after I hit 25 today, that box wasn't there, so I've got a couple of unexplained 2's floating around out there, but Slashdot's been a little flakey overall today (the GUADEC Reports story and the paying bills online story didn't have the i-opener story between them, and then later they did, internal server error messages an hour later, all that fun stuff). It's there now and I'm a-usin' it. : )
  • They are both starting to be made more and more like each other. More to the point, how do you define a supercomputer? This is very important for things like export controls.

    It's amusing to note that Macintosh touted it's G4 multi-processor workstations as violating export restrictions by being considered a 'super-computer'. But it just goes to show how out of date US export law is. Besides, what's to prevent someone from buying a 250 port switch from Cisco, 250 workstations from IBM, and then shipping them seperately and downloading the Beowulf software seperately?

    Export restrictions aren't very restrictive, taken in that light. National security most certainly isn't being enhanced any...

  • I was looking for this on Top500.org(the top 500 super computer site) and i couldnt find this one. Is that site updated daily or what?

    You would've found the answer to your question had you bothered to actually read the front page:

    The TOP500 list has been updated twice a year since June 1993.

    Figuring out why this particular system isn't on the list is left as an exercise to the reader.

  • AMD's processers can do multiprocessing (or could, I vaguely recall a rumor they dropped this with Athlon). The problem is that all the MP motherboards use Intel chipsets which support Intel's style of MP. AMD (and Cyrix et al?) don't have licenses to the Intel MP technology, and came up with their own -- which requires a different mobo chipset, which nobody makes because the boards would be incompatible with Intel CPUs.

    With AMD's recent growth and giving Intel a run for its money on releasing faster CPUs, we may see that change.

    Oh, and a Beowulf is clustered, so doesn't care what manufacturer makes the CPUs, it can even mix them. It's tightly coupled SMP machines that care.
  • If I remember correctly, somebody took 50-75 AIX RS/6ks and clustered them a year or two ago, with pretty nice results... Of course, you could always just get an RS/6000 S80 or a Sun e10000... or go all the way to a mainframe (I'd still love to have one, though a Cray-1 with the couch would be even better...)
  • The easiest way would probably be to post about Natalie's new 8-minute petrified grits, and taunt the moderators repeatedly - that way, you could get back under 25, and not have to worry about that silly 'No Score +1 Bonus' button.

    <anti-karma bait>I'm far enough over that I'd have to do some serious ninja pancake runs, Trolling for Scooby Doo, or M$ advocacy to get back under 25 (though I'm certainly nowhere near the levels of some of the famous posters here)... so I'll post this at +1 and take my chances</anti-karma bait>
  • Yeah they are... unless there's actually two of these at UNM... which would be nice, but little selfish.

    I can see posting "IBM to announce (newfangled technology) whis year" then a month or three later "IBM released (newfangled technology)". But two days of the same announcement is a little tough...
  • But the original articles are on Wired and cnet, but it's about the same system... :)
  • Wouldn't it be great if this really was a whole new story about a whole new Netfinity cluster at UNM? Then they could Beowulf the two clusters together into one of the top ten supercomputers.

    Oh well, just thought I'd mention thought before this thread gets wiped.
  • You can cluster AMD right now! One processor per node, but then again the Intel solutions don't ramp above two/node.

    Ah yes, the practical difference between a Beowulf and a supercomputer! Dig in your pocket and pull out some coins. Nickels are commodity clusters, pennies are SGI boxes and dimes are Cray. They are a collection of the same thing, namely coins/supercomputers. But what differentiates a cluster from a Cray? The intended purpose of the components. All the bits of a T3E were designed for it. In a cluster, the parts were designed for (and in most cases are) low cost workstations, hack rigged in software to run as one monster machine.

  • Stop trolling. Linux isn't touching solaris in the data center market. It doesn't touch it in the scalability department, and the x86 hardware to run it isn't there even if it could.

    If Linux is just killing Sun then why can't they stop making money hand over fist ? The fact of the matter is that Linux scalability plain sucks, and the answer has been "we're working on that" for at least 2 years. It _is_ getting better, but it is simply not up to par with OSes designed explicitly to handle the massive multi-cpu systems made by their vendors.

    Finally, "Beowulf != Scalable". Last time i looked into it, beowulf was originally very much like an MPP and had made some inroads to providing functionality of single-image computing. The applications of linux's "clustering" are still quite narrow, and chances are most people that say they want beowulf clusters don't have any clue what they'd do with them.

    Scalability for most people is something different. "Hey, i've got this webserver and RDBMS that are both thread-aware and thus can benefit from SMP machines..I'll just add more CPUs and RAM to this giant box and they'll automagically get better". Linux is an extremely poor choice for that scenario. The kernel locking is still much too coarse, too many sections are still non-reentrant, and the SMP hardware linux supports isn't particuarly scalable anyhow. Linux is _so_ poor a choice for true SMP computing that that was the area that made those amusing Mindcraft comparisions between Linux and NT realistic (i.e. NOT falsified). If you bother to look, the comparisons were using Multiple CPU boxes with Multiple Network cards. It shouldn't surprise anyone that NT did much better here, as at that time (and perhaps even still today) the linux IP stack was not re-entrant. The test took advantage of the fact that NT handled multiple device instances and SMP extremely well and Linux handled it extremely poorly. NT was designed for SMP from the ground up, with linux its been a progressive hack.

    People who disagree are more then welcome to come up with facts---NOT trolls, to support their arguments. Postings which make unsubstantiated claims and predictions are basically useless.

  • Linux is the true industry leader in regards to scalability and security

    Linux is neither. UNIX like systems are arguably a poor choice for a secure operating system since they are so damn intent on providing service and flexibility. However, even amongst UNIXes Linux is no where _close_ to being the security leader. Try OpenBSD [openbsd.org]. Any mention of "security" that doesn't also include "openBSD is pretty much the most secure UNIX flavor in wide use" is at best unenlightened.

    as far as scalability, see my earlier post. Linux does _not_ lead in scalability because of its poor SMP supprt and the poor scalability of the SMP hardware it can run on.

    Finally, IRIX machines can and do stay up for long periods of time, and there are frankly a hell of alot more mission-critical multi-CPU irix machines than there are _total_ multi-cpu linux boxes.

  • When is AMD going to get off there arses and make their processors friendly to the whole multiprocessor shindig? Translation, when do we get Athlon in a beowolf or SMP? If AMD is serious about wanting to compete with Intel than they need to start working on proof of concept projects like this. Second question is what is going to be the practical difference between a supercomputer and a beowolf cluster. They are both starting to be made more and more like each other. More to the point, how do you define a supercomputer? This is very important for things like export controls.
  • I'd guess the only Solaris this guy has seen is when he run a test installation on his p166... Hey dude! Guess why it was so darn slow?.. I'll give you a hint: 8086... So... Please stick your head out of your ... and take a look at this:

    A midrange machine with up to 30 processors
    http://www.sun.com/servers/midrange/e6500/
    This is what it does when clustered.
    http://www.sun.com/smi/Press/sunflash/9909/sunflas h.990929.1.html

    A machine with up to 64 processors
    http://www.sun.com/servers/highend/10000/
    What is it good for?
    http://www.sun.com/smi/Press/sunflash/2000-01/sunf lash.20000114.1.html

    SUN and clusters? http://www.sun.com/clusters/

    And remeber... If you keep FUDin' around the SUN will never shine on you... ;-F



    "At the end of the journey, all men think that their youth was Arcadia..." -Goethe

    - Where's my karma?
    - On the parking lot son!

  • I suppose now someone's going to ask if it runs Linux... At least it's already clustered, so that throws that question out...
  • Take a look at www.scali.com [scali.com] They provide high performance for beowulf clusters based on linux. They have really some impressive stuff running on their test clusters. It seems that they are getting close to linear scalability of informix. Add a node and the performance increases ~linear.

  • Los Lobos? (Score:3)

    by Otto ( 17870 ) on Thursday March 23, 2000 @08:54AM (#1179501) Homepage Journal
    Why name a supercomputer after a second rate band? I mean, yes, the soundtrack to La Bamba and Desperado, but everything else they've ever done sucks.

    As long as the computer plays that cool guitar intro from Desperado when it boots, I guess it's okay. :-)

    ---

  • Re:Wouldn't it be great... (Score:3)

    by Noryungi ( 70322 ) on Thursday March 23, 2000 @09:12AM (#1179502) Homepage Journal
    It's been done. Check this link. [ornl.gov]

  • Re:Wouldn't it be great... (Score:3)

    by Hardwyred ( 71704 ) on Thursday March 23, 2000 @09:17AM (#1179503) Homepage
    actually.. this is very much a reality. check out The Stone Soup Cluster [ornl.gov]. It uses all kinds of machines from 486's to Pentiums all in one cluster.


    ...and the geek shall inherit the earth...

  • Re:Wouldn't it be great... (Score:3)

    by thelaw ( 100964 ) <spamNO@SPAMcerastes.org> on Thursday March 23, 2000 @08:58AM (#1179504) Homepage
    there already is such a thing, although it's not as closely integrated as a beowolf. it's a process-migration kernel enhancement that allows several machines (whether workstations or clusters) to share cpu loads.

    check it out at mosix.org. [mosix.org] it is especially useful for compiling large projects (i.e. xfree86).

    jon

  • One more time... (Score:4)

    by Tower ( 37395 ) on Thursday March 23, 2000 @09:59AM (#1179505)
    The athlon is SMP capable, there's just not a cheap commerical memory controller that implements it yet. It's the same bus, same SMP structure as the Alphas, and they have done it, but AMD isn't producing chipsets (you'll have to ask VIA).

  • Re:Where's AMD? (Score:5)

    by Hankd ( 166718 ) on Thursday March 23, 2000 @10:27AM (#1179506)
    We have been building AMD PC clusters for several years now, ever since the K6-2. The Athlons are especially impressive. Our latest cluster, KLAT2 (Kentucky Linux Athlon Testbed 2), should have its 66 Athlons chugging away by April. We demo'd our
    first Athlon cluster at SC99 in November 1999.

    Although we have used SMPs as well (e.g., PIII
    quads from Dell), modern processors are memory
    bandwidth starved, and simple SMPs magnify the
    problem. I think a lot of cluster designs try to use SMP nodes to compensate for overspending on the inter-node network. I prefer to do the network carefully and use uniprocessor nodes.

    PS: I'm the author of the Parallel processing HOWTO and my first Linux PC cluster predates
    Beowulf (it was in Feb. 1994)... being good
    and even being first doesn't necessarily give
    you the highest visibility. Remember that when
    you think of AMD's Athlon. ;-)

    PPS: I used to be faculty at Purdue, but have
    recently moved to the University of Kentucky.
    Our new web site is http://aggregate.org/

Slashdot Top Deals

You knew the job was dangerous when you took it, Fred. -- Superchicken

Close