Slashdot Mirror
Virginia Tech Announces Supercomputer Plans
CousinVinnie writes "Previously noted in this Slashdot story, the administration of Virginia Tech has announced they're puchasing 1100 G5's (another story) in hopes to build a top-10 supercomputer by October 1. Tech will be spending $5.2 million over five years on the project, which should help it pull in more research money." Maybe VT can use the new computers to beef up their web site.
Does anyone know who else was considered for this contract? I'd love to see the arguments for the different platforms!
-WS
An operating system should be like a light switch... simple, effective, easy to use, and designed for everyone.
1100 G5's...that should corner the market for about a week...and give Apple a small boost to it's bottom line..
In America today you can murder land for private profit. You can leave the corpse for all to see, and nobody calls the c
Burns: [throws his glass at Homer]
You call this Postum?
[bashes a 5-feet high pile of paper]
Burns: You call this a tax return?
[bangs a CRAY with his cane]
Burns: You call this a supercomputer?
The comparison is like Apples to Oranges. Most people end up asking "Orange you going to build a beowulf cluster of those Apples?"
"Learning is not compulsory... neither is survival."
--Dr.W.Edwards Deming
Maybe Apple will use this G5 cluster against a single-processor itanium to show that, yes, they ARE the fastest personal computer!
The only problem will be finding a desk big enough to fit the guys...
Yeah, and his story says it was reported earlier, if you read it...
Not to sound like a troll, but isn't the Apple a bad machine to use for this? It's big, the fan configuration will make it extrordinarily loud, and it's built to cater to the end user, not to the embedded machine market. Yes, OSX/Darwin does work fairly well, but I'd think that the entire purpose of this computer originally would make it ill-suited to this task.
Many companies build physically smaller machines that still pack a lot of power, or sell parts to allow someone to design their own layout in a chassis. Remember, individual cases, power supplies, and the like become way overkill in such a large computer, and it would probably be cheaper to convert electricity once for a large section of the computer, supplying 12v, 5v, and 3.3v without each computer converting itself.
This just seems like the wrong way to do something thats hallmark has been in being cheap.
Do not look into laser with remaining eye.
So far we've seen that it's a cluster and what the building blocks are. What's the interconnect? What's the OS? What are the nodes using for a network filesystem? Are they at all? Is this intended for parallel jobs or for embarassingly parallel work?
Do you know why the road less traveled by is littered with the bones of the unwary?
There was a lot of inside info from people who work at Virginia Tech or go to school there. Lots of speculation and rumor, too, if you're into that sort of thing.
Si la vida me da palo, yo la voy a soportar Si la vida me da palo, yo la voy a espabilar
humm someone have a problem with rednecks?
Maybe a little to close to home? Or is the poster from WVA?
"(I) have this unfortunate condition that causes me not to believe a single thing any politician says when a mic's on.
Wait a minute - you're complaining about the cost of a G5, but go on to suggest they buy a Myrinet, a rather expensive interconnect. Something doesn't compute here.
... or else you'll be the victim of a drive-by fruiting!
Do not look into laser with remaining eye.
Slashdot summary:
1) Itaniums are for pussies.
2) Go Apple!
3) Opterons still kick the G5's butt.
4) I can't wait to run doom3 on my backordered G5.
5) People griping about apples proprietary hardware and software, and how this cluster could have been built cheaper from oem parts, and ebay ethernet hubs.
6)Dumb lists summarizing other trolls.
i would guess they have a higher-up who is a mac fan that was probably pushing them, only because it's rare to see such a setup with macs.
Was just trying to point out that even with an expensive interconnect such as Myrinet, the economics of apple just doesn't work out. But then, even if you are using a cluster of G5s, to get any reasonable super-computing power out of it, you would need a low latency (expensive) interconnect.
Here's the article from which the Collegiate Times article has paraphrased: http://www.technews.vt.edu/Archives/2003/Sept/035
Since it's based on a Power4 core, I think you should ask IBM that...
As another /. user pointed out, G5's do have an upgrade path. It's an IBM mainframe...
Opteron: Still under development.
Now tell me, on the Good/Fast/Cheap curve you design parameters lie?
"Learning is not compulsory... neither is survival."
--Dr.W.Edwards Deming
How much would you bet that Apple is throwing in some $$ for this try.
Blar.
It may be true that all dirts roads lead the VPI, but all white lines lead to UVa...
Comparing it to Windows will be a moot point, since El Dorado is going to have a 40% larger code base than XP.
Post suggestions here!
The comment has already been made. Let's move it along people. Nothing to see here.
From one of the articles "For the supercomputer to break the top five supercomputers in the world, it would have to possess 10 teraflops of memory."
Semper ubi sub ubi
The alumni have other causes to spend their money on, like renovations to the football stadium.
I would bet Apple would gladly cut a fantastic deal to get their model in the news as being part of one of the fasters supercomputers. They likely paid very little for the hardware.
It's a no brainer...
Karma means nothing to me, so suck it...
Why aren't they waiting for the Xserve update? Rhetorical question, but still...
I haven't seen one, but it looks like the PowerMac G5s are about 4U wide. 1100 x 4U = 4400U / 42 per rack ~= 105 racks.
Not only is this going to take up an enormous amount of room, but the power and cooling requirements are going to be crazy as well. And they don't have rails so getting them in the racks, and working on them once in the rack, is going to be a PITA.
1100 G5 Xserves would need only about 25 racks. Many fewer UPSes and A/C units to power in each rack. Much easier to install and work on.
I know Apple is gung-ho about this validating their "Fastest PC Ever" claims. But it seems a little poorly thought out on the University's part even if they got a sweet up-front price on the machines. Remember: the system price is a small part of TCO.
obviously no deficiencies vs. no obvious deficiencies
Pudge thinks their website isn't good enough. What does he want? Some flash? Maybe some pop up ads to spice it up. Whatever happen to simple being good and fancy being woooo pretty but useless. Oh wait, that still hasn't changed.
Well, there is the small matter of thermal dissipation--heat to polysyllabically-challenged types like yourself. Those AMDs you're babbling about would cook themselves; the traditional advantage of PowerPC processors for clustering is in the fact that they run MUCH cooler.
Even if you could make the AMD-based cluster run, you'd have spent so much on cooling that the G5s would end up being cheaper. Just ask DOD...
This isn't really a dupe, as this is a mention of the first official words form the school on the subject. Officials are finally speaking (and in some cases backing off) of the cluster in public.
I am, and always will be, an idiot. Karma: Coma (mostly effected by
Oh, nevermind.
I'm a nature photographer.
is it 5 million dollars worthwhile rareness?
I would buy some rare *insert collector item here* and make a normal faster/cheaper/better cluster!!!
[I can picture a world without war, without hate. I can picture us attacking that world, because they'd never expect it]
For the supercomputer to break the top five supercomputers in the world, it would have to possess 10 teraflops of memory.
I think that they mean 10teraflops of computing power, as opposed to 10terabytes of memory -- since the later would require each CPU to have 10GB of ram in it. Nonetheless, the anomaly tells me that this is a reporter not used to computer issues. (too few computer geeks at the college paper).
Free Software: Like love, it grows best when given away.
Anyone get the feeling that Apple might be pulling a Be, Inc and is trying to pull off a focus-shift?
Remember Be, the "multimedia" OS turned "Internet Appliance". Remember the death of Be. (damn, that stings. I miss the BeOS.)
Now witness Apple:
For decades, seemingly the darling of the press-production (DTP) world, catering to artists of all magnitudes, it was the computer you used to create real, bona-fide art. It attracted the freaks, the hippies, the art chicks. For many people, this was unnerving. Different people get "different" looks.
Now who's Apple targetting?
With OS X, I'm thinking geeks. We're different people, too, but in a, well, different manner. Instead of the artists, Apple's going for traditional suits, the realm of IT. It may be a matter of sheer survival that Apple penetrates here, because they don't stand a chance in these days of "homogenous" work environments.. Out with Apple (even if it works) and in with Dell WinXP machines! Linux faces the same dilemna, although Linux has some other benefits/detriments for it's widespread adoption. If Apple can show it's worth in the server room (just like Linux is doing), then maybe, just maybe, they'll start looking at Apple on the corporate desktop (just like Linux is doing).
Now, the idea of catering to suits is somewhat.. frightening. The whole damn market is different. They don't care about "look and feel", they care about numbers (see economic downturn, outsourcing to India, massive layoffs, H1B abuse, etc). This means Apple will have to change from being "cool" to utilitarian. But wait, I think I just painted myself into a corner here... Wasn't that the point of Apple? To be a tool and not an obstacle? Instead of creating computer art, we're now creating databases? Maybe Apple is on to something here...
If you were me, you'd be good lookin'. - six string samurai
because I know that Georgia is buying a 256 node cluster and at least some of them will be G5's from IBM running linux.
This was also discussed even earlier.... on Slashdot in July in a discussion of grid computing....
I'd love to see the arguments for the different platforms!
I think the argument for G5 came from here.
Belief is the currency of delusion.
to get any reasonable super-computing power out of it, you would need a low latency (expensive) interconnect.
Well, that very much depends on what type of computing you're doing. Some scientific computing is more tolerant of high-latency environments and would rather have the bandwidth.
I can't seem to find the quote from any of the articles right now, but VT is planning on using an Infiniband interconnect from Mellanox. While I don't know the relative price points, they are touting the fact that this is a high-speed interconnect that's faster than Myrinet or Quadrics at a fraction of the cost. I can't say for sure, since the Infiniband cluster we're helping to build at Stanford is not yet assembled.
This should be interesting to watch. I'll be very interested to see the $/gigaflop ratio for VT's cluster (though that doesn't have a bearing on the interconnect).
Considering this school is in the top 50 doctoral programs in Computer Science, it undeniably proves that you are more ignorant than the targets of your diatribe. This implementation will probably help their standing in the academic community, and is probably a good investment.
Just the motherboard and chip for a dual 1.6 GHz opteron runs you $1394.99 if you need PCI-X (which is probably necessary with the interconnect you want). And on top of that, you need the network cards, memory, case, hard drives, etc. etc.
You could easily get to Apple prices going with AMD.
Can someone explain the " Maybe VT can use the new computers to beef up their web site" comment? It loads perfectly fast for me. It looks pretty good. It even runs PHP, so it couldn't be a "They shouldn't use ColdFusion" type remark.
Am I missing something, or was that just a completely random comment?
________________________________________________
suwain_2
Jokes like this can get you put away in the punatentury for a very, very long time.
-----
jonathan barket
I just graduated from there and that website, while not ugly, is non-sensically arranged and the search comes up with some of the most ridiculous links. The department does have some nice clusters already set up though as well as a sweet 3d visualization studio!
There are a bunch of people posting gripes that this was a bad idea. But I don't think it's that bad. We should at least withold judgment until we see some data. One thing's for certain, it will outperform YOUR cluster.
Among the top complaints were:
You could buy several AMD's for that.
You might be able to, but the G5's they are buying already have 2 very good processors. As long as they're dividing up tasks among processors, it's nice to have all the memory management and overhead taken care of at a level of two processors per node instead of one. To be honest, I've never seen it done before, and it could have very interesting results.
The Mac's aren't designed for this sort of thing.
We don't know all the details of this cluster because they weren't all mentioned in the story, but my hunch is that Apple might cater to them a little if they are offering to dump $5 mill on a cluster. They might package the cases differently (sans curvy plastic or with shared power supplies).
Anyway, when it comes to speed of high precision calculations, the G* chips have proven their worth. And most High Science applications fall into that range of operation. We all know that clustering and distrubuting is touchy. The cost and speed don't scale linearly. And the cost vs speed ratio definately doesn't scale literally.
There is a possibility these computer science professors know something. So we might want to see how this thing performs before we rush to judgement.
Slashdot Syndrome: the sudden, extreme urge to correct someone in order to validate one's self.
Lots of speculation and rumor, too, if you're into that sort of thing.
This is slashdot! We're all about speculation and rumor. Innuendo, too, especially on the weekends.
Oh, and sentence fragments.
I want to drag this out as long as possible. Bring me my protractor.
How about resale? When the projects wind down and things need upgrading, they can get maybe over half a mil' in return for offloaded desktops (or at least scrap aluminum), as opposed to 57 cents for a bunch of beige schrapnel.
You know what?
Lots of "WHY?" questions, with lots of pointless trolling on the G5; but none of them actually look for answers. Mostly just more idiots who can't understand that a good vendor is important; that their own time is important; that ease of use is even more important now than it ever has been before. Luckily, these same idiots spend all their time setting up sendmail over their 14.4 modem. As for the G5, here are some strongpoints for it: - A fast memory pipe (1GHz) - Good heat management (9 fans but it's quieter than its predecessor) - Damn good FP performance (To get comparable FP performance on intel, you have to use the -fviolate-ieee flag on gcc, think about that) - Vendor-installed, vendor-supported Unix, with the vendor employing the entire OS's development team. - Fast system interconnects with network & I/O - Easy system setup (this matters a lot when you've got 1100 of them) - Proven apple reliability (and if you're going to fight this one, have something better than "is not!") (again, very important when you've got 1100 of them) Oh yeah, and OS X. Mach microkernel, Rondezvous, and distributed builds in the default toolset. Again, the idiots I mentioned above wouldn't have a clue about this stuff. As for _why_ VT getting this, VT's one of the largest engineering schools in the country. We've gotta simulate airflow over wings, heat propogation over materials, and other stufff this CS major doesn't understand. And we've got big development in bioinformatics. All kinds of CPU to crunch. AFAIK, the cluster's being paid for by federal grants or something like that. And now fools, flame me. Prove me right.
Care about electronic freedom? Consider donating to the EFF!
Eh, Virginia's tuition problems come from doing away with the car tag tax in the '90s, not building a supercomputer on the cheap in hte 'oughts.
All's true that is mistrusted
Tons of multi-Opteron systems already being sold, and for some time now.
Homeboy is mega-stupid.
G5: Deliverable today
Well, deliverable this weekend if you order 1,100 of them at a time; everybody else has to wait... but yeah.
$x='S24;r)>63/* h@<5+oZ)32"5cz';$me='phroggy'x$];
$x=~y+ -xz+\0-Tx+;print$_^chop$me for split'',$x;
They got a great deal from Apple on the machines, and when it falls of the bottom of the top 10 with the next release,they will have 1,100 machines to sell. Considering the excellent deal they got from Apple, plus a grant or 2 they picked up somewhere, they get great advertising for next to nothing. Of course I'm making this up....
Below is a link to show the noise of g5. (a movie). Apple did something called engineering (imagine that in a pc!), to put in many noise reducing features. So, the boxes may be bigger, but you get less power consumption and less noise. It is almost as if you pay extra money, and get extra features. Weird, I know....
2 0.html
http://homepage.mac.com/aaronsteele/iMovieTheater
"He who laughs last, didn't get the joke."-Cap
Myrinet eats CPU when sending data?? You must be using the suckiest driver and firmware possible. And Infiniband has lower latency than Myrinet???? Infiniband is a combination of protocol and hardware and actually Infiniband has slightly higher latencies than the best MPI implementations on Myrinet. Myrinet is a just a piece of hardware. You can write firmware in Myrinet to do almost everything in the Lanai processor present in the card itself, without consuming any CPU cycles. The performance you get out of Myrinet entirely depends on the libraries you are using.
What will happen next, dogs and cats living together? Mass hysteria?
As it turns out, this is the minimum recommended system configuration to run OS X 10.3 Panther.
Wow, imagine a Beowulf cluster of-- oh, too late.
503 Sig Unavailable
The Signature could not be accessed. Please try again later or contact the administrator
You know every time Apple Does something to improve it market share, company have a tendancy to want to do this ... its good buisiness sense, every PC troll comes out and puts em down.
I never really thought PC users were generally self protentious peaople but for the most part they are. after reading this ->
http://www.applelust.com/oped/amc/archives/amc0307 18.shtml
article that goes into the facts, a word pc trolls dont like to hear, about the early dual g5-xeon comparison it made how bad you people really are clear to me.
YES im sitting here running windows xp on my Athlon x86 PC but i , like unlike many i guess, have a mac sitting in the other room.
At the end of the day the G5 is an emensly power full computer. even the dual's are used to genetics can u imagine what over one thousand of them can do? Just except the fact that appple has sold some computers and find more posts to troll
I have a sneaking suspicion that these computers aren't going to be used as a supercomputer for long. I bet they set this up, get on the supercomputer list, and then in six months or a year farm out the computers to use in computer labs around campus. Besides, I haven't heard a compelling reason why VT *needs* a supercomputer.
Mainframes have one job: to move data from point A to point B as quickly as possible, while doing a relatively minor amount of processing on the way. Mainframes are what you use when you want to process every ATM transaction that happens around the world, all at the same time. In fact, your average mainframe is not really any more powerful than a dual- or quad-CPU Intel server, raw processing wise.
Supercomputers are the exact opposite. They're stacks and stacks of CPU's that process largely independent chunks of data. They do huge amounts of processing on each chunk of data. They do *not* move data particularly well. In many cases, supercomputers are held together with Gigabit Ethernet. That's not exactly *fast*...
Different computers, different tasks.
Linux IT Consulting and Domino Development in Michigan
This means that those of you who ordered your G5 early and expected to get one in early September, you are SOL. Your machine is goint VT. You have to wait another month before you get your machine.
-Matt
As an alumn, I am irritated with the decision. As a cluster developer of 5 years, I'm highly irritated with the rational of the G5. It's one thing to develop a system for the intention of doing research, it's another to base a decision on "..delivered by Oct. 1..". The question you should be asking is, which is more important - getting on the list? or doing the research? Seems to me that there is a more cost effective solution, that provides higher capacity, greater throughput, and more overall compute capability at lower cost... I'd personally suggest VT slow down, rethink the cluster, and buy something that fits the needs of the school and research programs-therein.
Side Note: While Tech has a great football team, the football program is (other than special discounts to students, and using the VT name) completely independent of the school. The football program is a business venture that does not interact with or require school permission, nor is it governed by the school boards that Steger answers to.
Use Linux!
Please note that what IBM calls the G5 and G6 CPU cores used in the S/390 and Z-series are a DIFFERENT chip than the Power series. The big push in the S/390 series is I/O bandwidth and reliability. Even the smallest S/390 currently available will take up to 256 I/O channels - which will *each* support a number of Fiberchannel-class connections.
0 .h tml
Reliability - it does things like have 14 CPU dies on a card - 12 processors and 2 spares. Each CPU is actually 2 running in lockstep with a comparator. At the end of each instruction, the two sides are compared for "same state" - if they agree, the entire state of the processor is latched into a memory array. If they disagree, the hardware tries to clear a possible soft error by reloading the state from that array and retrying the failing instruction - if they agree this time, it logs a soft error and goes on. If they still disagree, another chipset called the 'processor controller' reads out that latch and loads it into one of the spares, and starts it running from there.....
On the other hand, the 750 chipset that the world calls the G5 is a different beast, which will kick the S/390's butt on CPU speed. But it can't get anywhere near the z-series for reliability and I/O.
http://www-1.ibm.com/servers/eserver/zseries/99
Read why here.
That that is is that that that that is not is not.
Well, no. It'd be deep BONDI blue, if it were 1997 (hint: it's not). Nowadays, it's Deep Titanium or Deep Snow, something Virginia has never seen.
Hey freaks: now you're ju
Now that we have G5s don't ya think it would be a better idea to use a G5 pic instead of the G4 icon? e.g. http://www.apple.com/g5processor/
Those who laugh at you for you having a Mac.. are the people who constantly call you to fix their PC.
Contrary to popular belief, those building supercomputers don't just run down to the local CompUSA and pick out the one they like best. I'm quite sure that these folks did their homework and looked at the various options.
Buying good hardware from a reputable company that commits to helping you meet your goals should never be shunned by a bunch of people who have never been faced with building one of the three fastest supercomputers in the educational world. If any of those people want to critique this purchase then I'm listening.
Of course we torture people, we need the information --Gen. Pinochet
How the OS license on those go? I know with OS X you can get unlimited clients with each Xserve but I'm not sure how a cluster purchase works.
:-)
If you seriously think about it, the OS cost is far beyond the hardware cost at time... especially on the Wintel side of things. It's an honest question, not a critcism.
"Politicians find new names for institutions which under old names have become odious to the people."
Wrong again... sherlock Thats UVA.. as to VA being a police state or VTech being a red neck campus I wouldn't know I left VA for college, but I don't think one of the top rated science schools in states is all that bad.
DP
"(I) have this unfortunate condition that causes me not to believe a single thing any politician says when a mic's on.
what about photoshop benchmarks?
::ducks::
char *mySig;
So I read this "The project comes at a time when the university's academic departments are struggling to fulfill students' educational needs in the wake of a $72 million reduction in state support."
and then I wonder why you would spend $5mil dollars over the next 5 years to build a supercomputer? It seems like a better idea would be to reach out to the slahsdot/linux communities and see what kind of equipment they could get donated/free and then build a semi-super computer with that - or hell even just buy a shitload of cheap pc's to do it with....
maybe i'm just missing something...
Ave Molech Setting
It is a troll because the money for the supercomputer came from a NSF grant for that specific purpose. Furthermore the university expects to make a five-fold return, as have most universities in the top-x supercomputers.
Have you ever been to VT? We've got construction going on all over the place. The football stadium is about to get another "upgrade" after having received on just a year or two ago. We've got major construction going on in at least 3 different places, not to mention many smaller construction projects.
Meanwhile teachers are getting let go, classes that were taught in 30-person rooms 3 years ago when I started, are now taught in 400+ person lecture halls.
Does it suck? Certainly. However the money for the construction projects, football stadium, and supercomputer are all from grants, donations, and other means intended for a specific purpose. They can not legally take the money from a supercomputer grant or football stadium donation and use it to pay a teacher's salary.
We have uneducated rants in the school paper at least once a week saying "why are we upgrading the football stadium if we cant pay teachers!@#$"
Yeah, it does suck, but the university has no choice in the matter.
Just call it the Big Apple.
Why don't you zip it and please read the facts from the horse's mouth:
["Virginia Tech will have one of the top ranked supercomputing facilities in the world, supporting significant "big science" research. It is anticipated that Virginia Tech will realize at least a five to one return on this investment in terms of annual research grant and contract activity," says Glenda Scales, assistant dean of computing and distance learning at Virginia Tech.]
Let me say that again: Five to one return of investment projected. At least. If they are saying this outloud, I can assure you Ms. Scales knows what the hell she's talking about.
You can find it here
I'm frustrated with people arbitrarily say the Power PC chips aren't capable, when I know they are (having worked on a cluster of PPC based VME processors doing processing and high I/O throughput to a central mainframe). Motorola just dropped the ball wrt the high-end market, etc. IBM designed the architecture and is putting the ball back into play so to speak.
Or so they claim here. It seems they have all their bases covered and don't give a damn about ECC for a reason.
[Srinidhi Varadarajan, an assistant professor of computer science at Virginia Tech, and Jason Lockhart, director of the College of Engineering's High Performance Computing and Technology Innovation, initiated the venture at Virginia Tech. Varadarajan is an expert in reliability, a key issue in successfully exploiting terascale computing.]
They keep on going:
[Component failures are endemic to any large-scale computational resource. While previous generations of supercomputers engineered reliability into systems hardware, today's high performance computing environments are based on inexpensive clusters of commodity components, with no systemic solution for the reliability of total machine.]
And now for the solution for your reliability problem.
[Virginia Tech has the first comprehensive solution to the problem of transparent fault tolerance, which enables large-scale supercomputers to mask hardware, operating system and software failures - a decades old problem. It's a software program called Deja vu, designed by Varadarajan. He also integrated the software with Apple's G5s. This work will enable the terascale computing facility to operate as the first reliable supercomputing facility, according to Varadarajan, a National Science Foundation Faculty Early Career Development Program (CAREER) Award recipient.]
So maybe, just maybe, you and other people could:
1. READ before posting.
2. Then READ a little more.
3. Did I say READ already.
-sigh- Whatever.The moderators are getting so bad, they're making the submitters produce dupes for them in advance...
Floating point operations.
Why is it that no one is willing to give Apple an inch of credit?
Look, it's really simple. The G4 is loud simply because they didn't put in the engineering effort to make it quiet. The G5 on the other hand, is really really quiet. I was just using one today, so I know whereof I speak.
-Mark
Considering that these towers and every mac product in the current line is either silver-colored or white.. I don't see how pink factors into anything here. If colored computers are so "gay" then what do you think of makers like Alienware that build PCs which are yellow or green?
Those who laugh at you for you having a Mac.. are the people who constantly call you to fix their PC.
considering the G5's 'smart' airflow design, engineering the case like one big air channel, I would imagine that 1100 of these, if aligned correctly, could sum up to quite a nice breeze...
While I was a student at Virginia Tech ('94-'99, including a year of co-op) as a Computer Engineering student, I was always impressed by the diversity of computing resources throughout the university. Each department generally sided with a single environment (DECstations to FreeBSD PCs for CS, Windows (AutoCAD) for Engineering, Macintosh for Math & English), although all were supported. Even all the way back to the mainframe and its dumb terminals, which students used to sign up for classes. The general student at Virginia Tech learned, by necessity, how to do the basics across a variety of systems.
I'm happy to see Virginia Tech continue its push forward. "Commodity supercomputers" through clustering almost always refer to Intel-architecture systems. Why not Apples as well? It's a brilliant move forward, not only for the computational power this involves but also from a P.R. perspective... would all those high-school techies have heard about this if they chose an Intel architecture solution? And Apple will get good P.R. as the building blocks behind a supercomputer.
The numbers I've seen seem to suggest that a 128-host IB fabric is only slightly more expensive than a Myrinet fabric of the same size, with similar latency characteristics for MPI programs and about 3x the bandwidth per host. Quadrics has better latency characteristics than IB, but about half the per-host bandwidth; I'm not sure about relative cost.
What I'll be very curious to see is how Apple and VT scales an IB fabric out to 1100 hosts. They'll either have to buy a boatload of "backplane" switches or sacrifice bisection bandwidth across the system as a whole. That'll probably depend on whether they have big bisection-bandwidth-bound codes, like parallel FFTs...
"My life's work has been to prompt others... and be forgotten." --Cyrano de Bergerac
text
The setup should have been:
the fan configuration will make it extrordinarily loud
and the punch line (which I did get right) is:
On the plus side, the aeronautics department can use it simultaneously for a wind tunnel!
But who knows! Maybe the incorrect one works, too!
That is all.
OK, apart from being a nice thing to flame people with, what is this ECC talk?
Why is it so important?
Are there other "supercomputers" running without it?
Is there a good reason to disregard it and go with G5 anyway?
I think, therefore I am...I think.
My boss here at VT is a volunteer for this project... they've been designing and building rackmount shelf-type units to store all these new G5s, as well as helping with the cooling system. Here's some info he gave me.
;-), so there's a chance that he might use some PPC distro at some point.
The cluster will eventually run Mac OS 10.27... he said eventually, and Jason Lockhart, the project leader, is a friend and fellow Linux geek of mine (please don't hammer his inbox
Interconnectivity will be done with Cisco equipment, among the onboard gigabit LANs. Infiniband cards will also eventually be installed for 10 Gbit throughput.
You guys can offer alternative solutions and troll this as much as you want, but this is what VT is going with. In my opinion, it's not a bad choice... the New IBM PPC chipset is balls-to-the-wall computing, and Apple's 'stock' offerings in the G5 (Gbit ethernet, serial ATA, etc.) are all strong selling points. The fact that this cluster is intended for intense vector and matrix-based algorithms is another bonus, b/c of the PPC vector processing unit.
Apparently Apple shifted us up to the top of their production ladder, in order to make the contract, thereby extending the wait times for consumers itching for a G5... I find that a little humorous. Can't wait to see gigaflop statistics!!
May the threads progress competently.
Speaking very broadly, the recent PowerPC chips are very good at vector calculations, where the x86 chips tend to be better at integer calculations. Many scientific applications - for instance weather models - perform very well in vector models.
This here poster has answered the question already.
I'd say, mod him up, flog me for not reading this myself and go flame some ECC fanboys... (did I just say that?)
Seriously though, I thought ECC - or something alike - had to be important, reading about it, but OTOH couldn't imagine this university (with quite a track record already) not coming up with this themselves.
It sort of confirms my views on *some* wannabe-scientists on this list.
Having a degree doesn't equal having common sense. I like playing high and mighty like any other guy, but come on... This is largely a repeat article, it should have given some people the time to set their facts straight.
I think, therefore I am...I think.
Woohoohooo
:-))))
I always found this post to be absolutely idiotic. It still is of course, but almost pissed my pants laughing
I think, therefore I am...I think.
In talking to the person who is recruiting me to help lug the computers around when they arrive, the OS is to be OS X 10.2.7 on arrival, with plans to upgrade to Panther upon it's release. Straight out-of the box releases, with NetBoot planned to be used to distribute the images to each computer. This contradicts the rumors I've heard before, but is closer to a source who is on the planning team, who is too damn busy to talk to a luser like me.
Those who are possible volunteer recruits, there is an info session in Andrews ISB in the Corp. Research Center at 7:30 tonight and tomonrrow night (same presentation both nights). You *cannot* be on wage for VT to be elegible. I'm not sure if GAs count as this, since I'm not one, I didn't check.
I am, and always will be, an idiot. Karma: Coma (mostly effected by
How does Myrinet "eats CPU when sending data" ?!? It uses DMA to read/write data from/to host memory. The only thing the CPU is doing is writing 48 Bytes by PIO to post the send, whatever is the size of the message to send.
So either this is a flame bait or you really have no idea what you are talking about. I think I do as I write code for Myrinet firmware.
There are several companies building 2 and 4 processor Opterons, but NOT for costs such that each G5 can be replaced by 4 Opterons.
That's the point of the parent post: you can get extremely cheap AMD boxes (specially if you build them yourself). And you can get multiprocessor AMD machines fully supported and guaranteed by serious companies. But you can't get both on the same package.
The cheap ones are not an option for a serious massive project like Virginia Tech's. The classy ones are (price ways) on the same ball park as the G5 (admittedly cheaper but by no means by a factor of four).
AMD fanboys, take note: The G5 does have superior floating-point hardware, for either double precision superscalar or vector single precision work. If they're doing floating point the G5 is a clear win. /.ers live in, because they build their Athlons themselves. Perhaps the reality distortion field merely applies to introducing reality to people who have never seen it before.
The memory bus on the G5 is a bit better than on the Opterons - especially once you start doing threaded work, the dual unidirectional buses essentially allow cache transfers at the same time as memory transfers, and a whole bunch of other possibilities.
Lastly, what are you smoking? The only way an AMD becomes competitive with a G5 (machine to machine here) is if you build the AMD yourself and leave out the stuff the G5 has in it. Are you suggesting that Virginia Tech build 1100 Opteron, no, 4400 Opteron systems (you said 4x, not me!) themselves? That's crazy. They want somebody else to build and test the machines, and that somebody to be responsible when they fail. Of course such a real-world advantage has little to do with the bubble most
Apple won't give out review systems to real computer hardware sites, and so we have no proof I'm aware of. If you do I want to see your head to head tests of G5 and Opteron in single, dual, quad, and clusters since you obviously have proof and have tested all of this. You can now get two flavors of Windows, 3 Linux flavors, and FreeBSD is around the corner for AMD64. Where are all the 64 bit G5 operating systems again? Apache is ported, mysql is ported, Seti@Home, etc, etc. Where are all the OSX 64 bit program ports?
ignorance is bliss. googlefiberatx.com
Well I don't know about other science work but my dad does AIDS research and they use Macs exclusively in their labs for all of their model simulations and experiments. There's apparently a lot of biology research software that is available only on the Apple platform as well. I guess maybe because of their ties to educational institutions over the years?
Wonderful: three pointers to Apple's web site, pointing to pages with slick corporate "interviews". Do you actually work for Apple or are you just insanely zealous?
There are an awful lot of scientists using Macs for their research and work. I use them almost exclusively now after retiring my SGI's in favor of the OS X boxes and judging from the meetings I attend, I would say Macs have anywhere from 10-40% penetrance in science depending upon the subfield. For instance the last vision meeting I attended (ARVO, the big one for the vision research community), there were Powerbooks and iBooks everywhere. Probably a good 33% of the laptops I saw.
Visit Jonesblog and say hello.
Go Hokies
I was down there last week for the game, and the school is definitely transforming. also the links in the article point to off VT network sites, so is the comment warrented?
If you want, I can let you borrow my ex-girlfriend. Just put her in the corner of the room and she'll solve that cooling problem of yours in no time.
Yeah, but what kind of framerate will this giant G5 cluster see? My Athlon at home...
(I'm kidding, I don't use an Athlon at home, I use a P3 Tualatin.)
Non impediti ratione cogitationus.
No, 1102--you forgot the two boxes the geeks at VA Tech bought to test the environment. And lest you think Apple provided them for free; how do you think the group proved they spent the grant money? Think about the professor in 'Real Genius' for a minute.. LOL..
In America today you can murder land for private profit. You can leave the corpse for all to see, and nobody calls the c
yeah, because we all know that Va Tech needs a supercomputer for students to do karaoke.
Is that where the magic smoke runs through so the computer can work?
Marcelo Vanzin
Make that Big Blue Apple.
The potato it is uninformed.
Berkeley did it on $500 grand with their SETI@home project.
!@#$% whole-grain cereal. When I want fiber, I eat some wicker furniture. - G. Carlin
would it be possible to use the machines both as lab machines for students AND as a cluster? I mean does the gear that networks them necessarily prevent them form being used as individual machines? sorry for the n00b question but it sure seems like VT would get a lot more for their money that way.
Dude, look at an old sparc sometime. Sparc 1/1+/2 had 16 ram slots, circa 1990. Of course, you had to fill 4 at a time. The max is 128 MB i think.
:) --M
Not unless you could find a 30 pin SIMM in 32MB/four density, which I've never seen; the biggest being 16MB/four. I still have several Sun 3/80's, Sparc 1s, 2s, 5s, and 20s sitting in my basement. And the 20s I still consider a perfectly usable workstation. They were great computers in their time and I still have many fond memories hacking those Suns. But *sigh* it's time to clean out my basement.
I would advise you to have a little more faith in your alma mater....whoever told you there is a:
"more cost effective solution, that provides higher capacity, greater throughput, and more overall compute capability at lower cost..."
was smoking crack. If it was yourself than I respectfully disagree (and would like to see said solution).
Each PPC970 has two independent FPU, each of which is capable of executing a fused multiply/add instruction. This allows the PPC970 to function as if it had FOUR FPU's when doing all that nifty matrix math that makes up a large part of scientific calculations. This puts the machine's peak performance at over 17 Teraflops. If you don't believe me read the article over at arstechnica, or just look at the PPC970 documentation. Not to mention that the Altivec SIMD unit (for any repetitive single precision calculations) SMOKES AMD & Intel's vector solutions. Also the Mellanox Infiniband communication fabric has ridiculously low latencies and extremely high bandwidth..beats Myrinet on both counts from what I've seen...
The long and the short of it is that a dual-Opteron built cluster would have to have significantly more nodes (== more expensive) to deliver the same peak (or average for that matter) performance, and would take longer to be built.
Apple wins on price.
Apple wins on performance.
Apple wins on delivery time.
I'm sure the decision to use the G5 was not made lightly or frivoulusly, and that all the options were carefully weighed before choosing the one made the most sense.
Come on /. Would you please grab a damn G5 thingie offa something somewheres and quit using G4 thumbnails in Apple stories? It sticks out like a green sore thumb and it's embarrassing for /. Time to roll with the changes.
I suggest LASEK/LASIK.
Zealot, perhaps. Troll, never.
Hmmm. "Zealot." Me thinks we need a new mod category...... If only because it'd REALLY help my karma.
Going back to lurking as a very proud mac zealot......
How a bout Titan, ok it's made out of Aluminium...
I got it:
The Aluminati Cluster
And I bet the first they're going to compute would be the question to "42".
There are two rules for success:
1. Never tell everything you know.
The PowerPC 970 is based on the Power4. You need to check your facts. (Hint: visit the IBM website)
Thanks.
Some questions, if you feel like answering, that is.
Any reason why Apple doesn't support this ECC memory apart from being a bit over the top for simple desktops?
Can they be custom installed (sort of like "look here, I'm going to build a supercomputer instead of doing this photoshop stuff, so put those ECC thingies in there, ok?") or is it plain not supported, not possible?
Is ECC memory readily supported by other manufacturers (apart from SGI and CRAY)?
And the obvious one, why do you think a place like Virginia U. overlooks this while they - and the people responsible - obviously have firsthand experience in this?
I think, therefore I am...I think.
I did build my athlon...
but i *bought* my ibook. Warranty, yo. sometimes it's worth it. plus OS X is uber-sweet.
You like your new Mac more than you like me, don't you, Dave? Dave? I asked...She said Yes.
<heard from the bowels of Intel's headquarters>
Damn, Apple beat our Itanium shipments by 100 machines!
That "traditional advantage" doesn't apply here because the power consumption of IBM's PowerPC 970 and the AMD Opteron are nearly identical.
Exact numbers are few and far between on both sides (neither company has a publicly available document that describes their exact power consumption), but most estimates put them as being real close. Both chips consume about 45-50W typical power consumption, and somewhere in the order of 50-60W maximum power consumption. One or the other might be a few watts cooler, but the difference is quite small.
Both of these chips consume a lot more power than the G4 and both consume a lot less power than the AthlonXP or the Pentium 4 (or the Itanium for that matter, which is the current champion of high power consumption).
Others who know more than I do about Myrinet have already pointed out that you don't seem to know what you're talking about, so I'll skip that.
Next point, FP performance. Let's see, Apple managed a score of 840 on Spec CFP2000 base on their 2.0GHz PowerPC using GCC. AMD managed a score of 1209 on their 2.0GHz Opteron using Intel's compiler, while IBM managed a slightly lower score of 1172 using GCC. If the G5 is so superior, it sure isn't doing a very good job of showing it.
The memory bus on the G5 is better? What are you smoking here?! Each Opteron chip has a low-latency integrated memory controller capable of managing 5.4GB/s of bandwidth. Two G5 shares a higher latency, off-chip memory controller capable of 6.4GB/s total. Their bus may allow memory and cache transfers at the same time, the Opteron has a completely separate buses for these tasks! The Opteron also has twice the L1 data cache and twice the L2 cache that the PowerPC 970 has.
I will, however, agree with you that the Opteron is no cheaper than the G5 if you can get Apple to give you a useful configuration of the boxes for a supercomputer (who the hell needs a high-end gaming video card on each super-computer node? Let alone iTunes!). As for building the machines themselves though, why would they do that? Why not just buy a machine from one of the dozens of companies that are selling dual-processor 1U Opteron servers now. Many can and have built clusters of these already. Speaking of '1U' and all, brings me to one fo the dumbest things about this supercomputer, it's ALL being built usind the damn PowerMac tower cases! Now THAT is just a DUMB idea! 1100 giant desktop cases using bog-standard power supplies and the works, now that is just ridiculous!
Would be madness to build a supercomputer out of Opterons -- and you'd need 4400 of them for it to . . . to do what?
Specint base Opteron 246 = 1248
Specint base G5 2 ghz = 800
Specfp base Opteron 246 = 1209
Specfp base G5 2 ghz = 840
These numbers are from spec.org for the opteron and of course apple.com. Now I'm not sure why I need 4400 opterons all of a sudden to equal 2200 of these initially slower machines. And of course, the idea that no one's doing this:
http://www-1.ibm.com/servers/news/pr_730.html
ARMONK, N.Y. and TOKYO July 30, 2003 -- IBM today announced that Japan's largest national research organization has ordered an IBM TM Linux(R) supercomputer that when completed will deliver more than 11 trillion calculations per second, making it the world's most powerful Linux-based supercomputer. It is expected to be more powerful than the current Linux cluster currently ranked as the third most powerful supercomputer in the world, according to the widely popular TOP500 List of Supercomputers [1].
The supercomputer is planned to be integrated with other non-Linux systems to form a massive, distributed computing Grid -- enabling collaboration between corporations, academia and government -- to support various research including grid technologies, life sciences bioinformatics and nanotechnology.
The system with a total of 2,636 processors will include 1,058 IBM 325 systems, which were introduced today with 2,116 AMD OpteronTM processors. [2]
At least look into stuff before deciding that it's not competitive -- the opteron is certainly competitive. Sure, it's not a Mac, but it's pretty damned fast, and a quick look online shows a price of ~$3000 -- same as list on the G5.
Finally, no one knows how well these things will multiprocess -- they're not using linux, they're using OS X -- which is BSD, granted, and should do well, but god only knows what they've done. I actually do think it will be a nice, fast machine, just not necessarily significantly better for specious reasons than a similar opteron machine when the initial base spec marks are so much in favor of the Opteron (800 vs 1200 is 25% last time I checked), and the Opteron has the advantage of having IBM working on making Linux nice, fast, and well-behaved in a cluster setting.
Kargis
Why does the G5 say "G4" on it?
I love NetHack.
Better luck next time.
Blar.
...the right number of buttons.. the right number of buttons for a beginning user is one. (1)
;)
I, like many other Mac users, purchased a 3 button mouse almost immediately after getting each Mac I've owned.
Do you think it's time to get your head of that hole and stop with the "number of mouse buttons" complaints?
I don't understand why you just want to chime in and miss the joke entirely..
good luck with that, eh? maybe you'll eventually save up enough to get a Mac and not be so bitter.
Yes, and there is a lot of research software that's only available for Windows, and there is a lot of research software that's available only for UNIX. You see, sometimes people write software for specific platforms.
The benchmark that this machine is going to be rated against is LINPACK, which involve large matrix calculations and is heavily floating point dependent. An ideal computer would be able to achieve its peak theoretical FP performance in the extreme of an infintely sized matrix calculation under LINPACK, R_peak. This is different from SPEC FP which is a suite of different floating point intensive benchmarks which test various aspects of floating point performance for the CPU, compiler, and overall system.
AFAIK, the Opteron has 3 asymmetric FPUs where each FPU performs some subset of functionality, and more than one FPU component may need to act in concert to perform other FPU instructions. More to the point, AFAIK, the Opteron is capable of doing 1 floating point multiply and 1 floating point add per cycle.
The PPC 970 has 2 fully symmetric FPUs each of which can perform any FPU instruction available to the architecture. One of the key features of the POWER ISA family and its descendants is the single cycle FMA instruction which allows each FPU to perform 1 floating point multiply and 1 floating point add per cycle. As a result, for what LINPACK measures, the PPC 970 is capable of twice the theoretical throughput per clock cycle when compared to an Opteron. Since Opterons and PPC 970s run at comparable clock speeds, it's no surprise that it takes twice as many Opterons to equal the performace of a cluster of 970s.
SPEC FP is irrelevant in this case.
-Bruce.
Moderated at:
-1 Doesn't Reaffirm My World View
Here's some info about running Linux on Macs:
i d=6918320
/ 0259234&mode=thread&tid=106&tid=126&tid=137&tid=17 4&tid=181&tid=185
http://apple.slashdot.org/comments.pl?sid=77884&c
http://apple.slashdot.org/article.pl?sid=03/09/10
http://www.terrasoftsolutions.com/
http://www.yellowdoglinux.com/
That that is is that that that that is not is not.
As far as I can tell, the press release you referenced wasn't mentioned in the original Slashdot postings. It's not fair to criticize someone for not reading something they didn't know existed, is it? In any case, while I'm sure the Deja Vu software is useful, it doesn't help those of us who aren't setting up clusters--there's still a need for reliable hardware. Check out http://cr.yp.to/hardware/ecc.html for reasons why.