Big Mac achieves around 14 TFlops with 128 Nodes 307
mzs writes "The Virginia Tech G5 cluster has achieved around 80% of its peak performance in preliminary Linpack testing with 128 nodes according to Jack Dongarra at the Top 500. "They're getting about 80 percent of the theoretical peak," Dongarra said. "If it holds, and it's unclear if it will, it has the potential to be the world's second most powerful machine." Typically getting 60% of peak in the Top 500 lists is quite good. If the Big Mac cluster achieves 60% of peak it would displace the 2,300 2.4 GHz Xeon cluster at LLNL for the number three spot on the current list."
Mmm Big mac (Score:2, Funny)
Re:Mmm Big mac (Score:2)
Re:Mmm Big mac (Score:2)
I remember the driver sauce, hme is surely worth reading. sunhme.c [linux.no] sunbmac.c [linux.no]
nope... (Score:3, Informative)
Title is wrong - they get 80% efficiency on 128 nodes. The 14 TFlops number is if that efficiency is held through the full size of the machine (2000+ processors).
Re:nope... (Score:2)
Re:nope... (Score:3, Informative)
Re:nope... (Score:2)
I was going to tack on something like "...which it won't." to the end of my post, but I thought it would get interpreted as an opinion as opposed to pointed out that the submitter was on crack.
Re:nope... (Score:2)
Re:nope... (Score:2, Insightful)
I was going to say that something seemed terribly wrong... because when I read that I interpreted it as 128 nodes were pumping out 80% of it's peak at 14Tflops...
if that were the case then the last 972 nodes would almost be a complete waste if they only gave a 20% preformance increase.
good call on that Durinia!
What happened to the federal controls? (Score:3, Interesting)
Re:What happened to the federal controls? (Score:2)
If a nation could not export the computers needed for a cluster, they could always build the cluster here in the US and utalize the thing over the Internet. In fact, they could probably contract a domestic company to do it for[THE REMAINDER OF THIS POST HAS BEEN CENSORED BY ECHELON]
Re:What happened to the federal controls? (Score:2)
Keep in mind that one of the first major uses for supercomputers (or any computers for that matter) was for breaking cryptography (ie, German codes during WW2). I'd imagine the restriction was done in the name of national security.
Re:What happened to the federal controls? (Score:3, Informative)
Testing/developing a nuke program is much easier on a supercomputer than attempting live tests like the USSR, USA, France, Red China, and the UK did back in the 40-70's. Infact, a test detonation by a "unknown" would be sufficent grounds for a beat down by those countries or your neighbors (see Israel vs. Iraq, 1981) India and Pakistan have been allowed nukes mainly as a local deterent to keep 1-2 billion people from dying in South-Central Asia in the Indo-Muslim war that has
Re:Simulating Nuclear Weapons is more the CPU cycl (Score:2)
Re:What happened to the federal controls? (Score:2)
you see, they were seen as tools for producing weapons and breaking cryptography, and they thought that they could stop the passing of time and limit technological advancement(or something, anyways, it must have been a short term decision because in long term it just doesn't hold up, because the supercomputers of today are desktop computer parts tomorrow.)
anyways, it's OK to export weapons(tanks, missile systems
Re:What happened to the federal controls? (Score:2)
What gave you that idea? I would love for everybody to have the same access to knowledge and the same access to the opportunity to contribute and move humanity forward.
Re:What happened to the federal controls? (Score:2)
Well, that statement gives me the "warm fuzzies" too, for a second, until I wake up to reality and remember that enough people think that the best opportunity to "move forward" is for their country/terrorist organization to join the nuclear club. Just look at Pakistan. Their country has innumerable problems with underdevelopment, but building a nuclear bomb somehow
Re:What happened to the federal controls? (Score:2)
G4, G5 (Score:3, Funny)
Re:G4, G5 (Score:2)
Don't you mean a giant heatsink?
Re:G4, G5 (Score:2)
I think he means this [home-evolution.com].
it's a giant heatsink... but (Score:2)
I agree (Score:2)
Of course... (Score:2, Funny)
There are weenies that will say "Psstt.. you know that #2 computer in the Big 500? It only has one button on the mouse!"
Re:Of course... (Score:2)
-WS
Re:Of course... (Score:2)
There are weenies that will say "Psstt.. you know that #2 computer in the Big 500? It only has one button on the mouse!"
Technically, it has 1100 buttons...damn I bet that is one unweieldy mouse.
Mass storage? (Score:2)
Re:Mass storage? (Score:2)
Thats one fast Mac (Score:2)
This is tremendous advertising for apple, but what about clusters of Power4's and 5's? why wouldnt they out-perform this cluster? at 14tflops... would 2200 macs be fairly equal to the earth simulator?
And lastly with IBM seeing their G5's at 3Ghz in 8 months or so.... d
Re:Thats one fast Mac (Score:2)
The Power Mac version of the IBM POWER chips probably scales better with cost/performance. POWER/UltraSPARC/Itanic are really best suited to high-reliability workhorse servers. Clusters probably don't need ultra-high reliability if they can program around it.
Re:Thats one fast Mac (Score:2)
Re:Thats one fast Mac (Score:4, Informative)
Re:Thats one fast Mac (Score:2)
Re:Thats one fast Mac (Score:2)
Re:Thats one fast Mac (Score:2)
Re:Thats one fast Mac (Score:2)
Re:FOR THE LAST TIME: IT RUNS MACOSX (Score:2)
Re:FOR THE LAST TIME: IT RUNS MACOSX (Score:2)
Re:Thats one fast Mac (Score:2, Informative)
That's nothing... (Score:5, Funny)
yeah but (Score:2)
Big Mac? (Score:2)
Big Mac cluster? What's that do, display in real time the flow of special sauce molecules over the surface of a flame broiled patty?
So what? (Score:2)
Yes, this is a troll post. I am pissed the submission after the Napster 2 one wasn't the itunes4Windows announcement.
rejected my story! (Score:2)
AAAGH!
Re:rejected my story! (Score:2)
Topic Icon (Score:3, Informative)
Like this for example > http://www.apple.com/g5processor/ [apple.com]
Re:Topic Icon (Score:2)
Even *after* I sent a new one. Three times.
I'm thinking that the G4 icon will be there when the G7 Virgin Tech cluster comes out.
Re:Topic Icon (US Flag) (Score:2)
Hey - You're right! It looked a little odd - so I zoomed it up a bunch and sure enough: 12 stripes! I wonder which colony they dropped?
(Reminds me of the time I was driving home from work and some guy had painted a US flag on the side windows of his Chevy Blazer. As a former teacher, it saddened me that this "patriot" figured we only had 36 states - six rows of six stars. Also, in his world, there were 15
Re:Topic Icon (Score:2)
>>>>>>>>>>>
That's a pretty accurate description of the school, based on the people I know that go there
Re:Topic Icon (Score:2)
Now I know what I have to do... (Score:2)
Re:Now I know what I have to do... (Score:2)
Great feat - IBM! (Score:2)
Why is everyone so obsesses with Apple when it comes to G5? What does Apple have to do with the G5?! This is the kiddie version of IBM's success CPU named Power4, if I'm not totally incorrect.
Apple, to me, is a group of cosmetologist hangarounds.
Only 64 Nodes (Score:2)
What? (Score:2)
All this time the answer to my supercomputing needs was at McDonalds. What a burger!
Re:What? (Score:2)
My thought: "Yes, I would like fries with that."
--RJ
Efficiency to Burn (Score:2)
Re:Efficiency to Burn (Score:2)
What compiler are they using? (Score:2)
What OS are they using? (Score:2)
Re:What OS are they using? (Score:2)
Re:What? IS it OSX they are using? (Score:2)
Well that changes my perception. at first I just considered it a big blue supercomputer but ... the 2nd fastest supercomputer on the planet runs OSX!
Premature (Score:3, Insightful)
I wonder what they could do with a WHOPPER array! (Score:2)
Theoretical peak of the LLNL Xeon cluster? (Score:2)
This article reports that the 2300-noder operates at 7.6 teraflops, but i was wondering what percentage of its theoretical peak that is.
Re:Theoretical peak of the LLNL Xeon cluster? (Score:2)
Rmax is the best it's attained; Rpeak is the theoretical peak in a perfect world with 100% multiprocessing efficiency.
One of the really amazing things about the NEC Earth Simulator (aside from its sheer power) is that its Rmax is over 85% of its Rpeak. A lot of other systems only get Rmax of 60-70% of Rpeak.
I've read that PowerPC chips are pretty efficient in SMP scenarios, and if
Re:Theoretical peak of the LLNL Xeon cluster? (Score:2)
Re:Theoretical peak of the LLNL Xeon cluster? (Score:2)
SETI array (Score:2, Insightful)
Although they don't run Linpack, and therefore can't be considered on Top500 the same way, it's still cool to know that SETI would still place second on the supercomputing list. Back in 2001, they were averaging a very large number of teraflops as well, (>10TF) the figure is on the internet somewhere. In 2001, that was greater than the top three supercomputing sites comb
Pretty cheap (Score:3, Insightful)
At $5.2 mil for 1100 machines, I think they paid full market price; that's over $4,500 per machine, and currently Apple is selling dual 2 Ghz G5's for ~$3000. And that's with lots of extras that they wouldn't want in a cluster (ATI 9600, CDRW, etc), which hopefully they convinced Apple they didn't need... (else they've got a whole lot of Mac keyboards sitting around!)
I wonder how much of the cost was the actual machines, and how much was infrastructure and networking stuff (I can just see 1,100 Macs all powered off one extension cord and a bunch of surge protectors).
Re:Pretty cheap (Score:2)
Re:Pretty cheap (Score:2)
They plan to use the GPU on the 9600 in the future for extra compute cycles. It's a cheap coprocessor.
What will the effect of (Score:2)
cost (Score:2)
I pulled the modem out of the default dual 2ghz g5 mac and upped the ram to 4gigs, as I vaguely remember the big mac node config being from some other article somewhere. Cost 5,320. Tried to up the quatity to 1100 in my cart but the web form would only allow 3 digits, so I did 110. Got 563,200. That looks like a volume discout because my
OOPS! (Score:2)
Sorry folks. I'll sheepishly retreat into the corner now
Shouldn't the icon read G5? (Score:3, Insightful)
Re:Shouldn't the icon read G5? (Score:2)
The article title is demonstrably wrong... (Score:2)
Here's how I figure that:
A G5 proc can do 4 64-bit FP ops per clock cycle (2 FPUs with each capable of doing a multiply/add op), so that's 8 GFlops/s per 2GHz proc, or 16 GFlop/s per dual-proc node. For 128 nodes, that's a little over 2.048 TFlop/s peak. Dongerra said they were getting 80% of peak, which would be 1.638 TFlop/s.
The thing
Re:The article title is demonstrably wrong... (Score:2)
Re:The article title is demonstrably wrong... (Score:2)
Stop the obfuscation ... (Score:2)
If there are too many nodes the error rate goes up super-linearly (since interconnect errors have to be factored in.) There's no getting around this. If they are just running *without* ECC and hoping the results just come out correctly, then this is worthless -- its amounts t
Re:Stop the obfuscation ... (Score:2)
ECC is overrated.
Re:Stop the obfuscation ... (Score:2)
Not necessarily. Know what ECC stands for? One of the C's is for Correction. Given enough extra bits, you can arrange the coding so that valid memory words all differ from one another by two or more bit flips. Given a single bit flip, and assuming it was only a single bit flip, you can uniquely determine the corresponding valid memory word. If the hardware is functioning properly (e.g., it was just a cosmic ray hitting a RAM cell), then you can rewrite the valid combinat
Re:Stop the obfuscation ... (Score:2)
Re:Stop the obfuscation ... (Score:2)
2. Step away from you computer and walk out side for a bit. Try not to be scared by all the blue sky and that really bright yellow ball up there.
3. Come back in after sudden exposure to intense light causes immediate sunburn.
4. Read about Deja Vu here [chaosmint.com].
Wow (Score:3, Funny)
Must be the "special" bun in the middle.
oh yeah?? (Score:2)
Gigaflops question (Score:2)
more hype (Score:2)
Uh, huh. So, if those statements were true, the G5 would be 12 times faster than a Xeon. Quite a feat of engineering that would be. But I don't think so. The folks at VT are extrapolating their 64 node performance to a 2300 node cluster.
This tells us two things:
Good comparison.. (Score:2)
I have submitted this as story, but it evidently never news like the G5 cluster... Linux and AMD no longer our favorite 'underdogs' anymore?
I wanted to like iTunes... (Score:2)
For me,
But it also won't download or let you create MP3 files from music you download from them. I've got a car player with a 40GB hard drive that only supports MP3
Re:To quote G. W. Bush: "Enough is Enough" (Score:4, Funny)
Re:To quote G. W. Bush: "Enough is Enough" (Score:2)
Why VirtualPC doesn't work on G5s (Score:2)
Re:hello... (Score:2)
Re:hello... (Score:2)
That's a huge factor. Getting that factor 20 scale is hard and is a large part of the usefulness of the underlying software and networking architecture.
Since it hasn't been demonstrated, the Top 500 people are right to withhold judgement.
Re:I strongly doubt (Score:2)
No. They would only have to be the same 80%. What this test is saying is that 128 nodes together are producing 80% of their theoretical peak (around 1.6 of 2 TFlops). If this efficiency is replicated through the rest of the nodes the final output would be around 80% of its peak (around 14 of 17.6 TFlops).
Re:How Many... (Score:2)
Volkswagens! (Score:2)
Also, I don't trust it until it comes with a 'Performance Rating' from Cyrix, How many Pentiums is this beast?
Re:How Many... (Score:2)