Swiss Supercomputer Edges US Out of Top Spot (bbc.com)
There have only been two times in the last 24 years where the U.S. has been edged out of the top spot of the world's most powerful supercomputers. Now is one of those times. "An upgrade to a Swiss supercomputer has bumped the U.S. Department of Energy's Cray XK7 to number four on the list rating these machines," reports the BBC. "The only other time the U.S. fell out of the top three was in 1996." The top two slots are occupied by Chinese supercomputers. From the report. The U.S. machine has been supplanted by Switzerland's Piz Daint system, which is installed at the country's national supercomputer center. The upgrade boosted its performance from 9.8 petaflops to 19.6. The machine is named after a peak in the Grison region of Switzerland. One petaflop is equal to one thousand trillion operations per second. A "flop" (floating point operation) can be thought of as a step in a calculation. The performance improvement meant it surpassed the 17.6 petaflop capacity of the DoE machine, located at the Oak Ridge National Laboratory in Tennessee. The U.S. is well represented lower down in the list, as currently half of all the machines in the top 10 of the list are based in North America. And the Oak Ridge National Laboratory looks set to return to the top three later this year, when its Summit supercomputer comes online. This is expected to have a peak performance of more than 100 petaflops.
Read the summary, it was bumped out of the top 3 to become #4.
How many nodes and cores per node? How much memory is available on each node? What kind of interconnect is used? And what type of file system is used? In my experience, Lustre can be quite a bottleneck for many applications. It's also great that there's that much power, but it will only be used efficiently in parallel with good interconnections between nodes. Scalability of jobs matters a lot.
scroll text?
SAD!
is it "top spot" or "top three spots"?!
the Swiss just copied Calipari
FTA:
With the two Chinese supercomputers and one Swiss system occupying the top of the rankings, this is the second time in the 24-year history of the TOP500 list that the United States has failed to secure any of the top three positions. The only other time this occurred was in November 1996, when three Japanese systems captured the top three spots.
Happiness in intelligent people is the rarest thing I know.
Ernest Hemingway
Something tells me NSA or other state entities have faster computing power, and when they are done with it they destroy them rather than surplus them.
NSA had no comment.
Those bankers must be going all-out on encryption.
IBM's currently working on the 100 petaFlop system using POWER9@14nm which will come online sometime in 2017, and POWER10@7nm, and POWER12@5nm. IBM has always worked on three successive generations of POWER processors at a time. POWER9 completed in 2016 so POWERx+1 arrives five years later so will see POWER10@7nm in 2021, and about three years later will get POWERx+2 or POWER11@5nm in 2024. Every POWER processors development cycles is about 7-10years long. IBM first started working on POWER11 in 2013 once POWER8 was completed.
Their operating system is full of holes.
Trump's home PC that he browses Fox "News" on for non-fake news is more powerful than this Swiss PoS. His PC's the greatest, has the bestest Interwebs and most ultraflops.
So they must have really fancy weather forecasts now.
Top spot != Top three. Is it too much to ask for an accurate title? Clickbait trumps accuracy? I can't think of another reason how this title could have slipped through.
This Piz Daint is efficient, too, consuming less than 30% as much power as the Opteron-based Titan. TaihuLight is still very, very impressive, though the new one at Oak Ridge should top it.
This is a hacked account, for which the owner can not be held responsible.
If the GOP/Trump plan to gut the DOE happens you can forgot about supercomputering in the US.
The same supercomputer center has 2 redundant clusters dedicated to meteo forecasts, yes: http://www.cscs.ch/computers/k...
And yes, the forecasts are fancy.
Yeah yeah I know, named after a mountain.
But I would like to see a supercomputer named Deez Nuts.
The name of the supercomputer "Piz Daint" is extremely offensive in many Europen languages.
But it looks like the owners don't care.
I don't get it, the US is still in the top 5%.
I don't get it, the US is still in the top 10%.
I don't get it, the US is still in the top 20%.
I don't get it, the US is still in the top quarter of all sovereign states.
I don't get it, the US is still in the top third of all sovereign states.
I don't get it, the US is still in the top half of all 190 (or is it 206?) sovereign states!
I don't get it, the US is still in the top two thirds of all sovereign states.
I don't get it, the US is still in the top three quarters of all sovereign states.
I don't get it, the US is still in the top 80%.
I don't get it, the US is still in the top 90%.
I don't get it, the US is still in the top 95%.
but you still have to suck up the fact that you're losing. After that you can move on.
I think a far more interesting article would be what are these systems being used for on a day-to-day basis?
My beliefs do not require that you agree with them.
See you in Switzerland!
I think we'll be okay. The XK7 will run AOL just fine for Rick Perry.
Neither GOOG nor NSA discloses the size of their supercomputing clusters.
We can get some idea of how big the former is by China's application to expand their "fastest" supercomputer. They were thwarted when usgov denied Intel an export license. If the threshold of feeling threatened enough to deny license is crossed only when doubling the size of the "fastest", then US military computing must be somewhat faster than the "fastest".
For Google, you can get some idea from watts. Tianhe-2 20MW (seems to have lower cooling efficiency than designs published by Facebook), GOOG ~1500MW (globally, not in one datacenter, but single-location figures are still higher than Tianhe-2).
pattern recognition and AI development and think IBM selling you a watson software system you will or could need a super computer to allow it to really grow.