A supercomputer known as Jaguar has finally bested IBM's Roadrunner supercomputer in the biannual TOP500 list, but researchers have already begun looking into exascale supercomputers that consist of 100 million cores and run 1,000 times faster than Jaguar. Computerworld reports that the U.S. Department of Energy has begun holding workshops on the new supercomputers to run high-res climate models, develop smart grids and aid fusion energy design.
The current supercomputer race measures success in one quadrillion calculations per second, or one petaflop. Roadrunner has dominated the list since it became the world's first petaflop supercomputer in 2008, and has tackled everything from dark matter to HIV family tree modeling at the same time. But a repartitioning of the system dropped is performance from 1.105 petaflops per second to just 1.04 petaflop/s, and allowed the Cray XT5 supercomputer at the Department of Energy's Oak Ridge laboratory to seize the lead with a performance of 1..75 petaflop/s.Another Cray XT5 supercomputer known as Kraken grabbed third place with a processing speed of 832 teraflop/s (trillions of calculations per second). IBM's BlueGene/P supercomputer in Germany took fourth place with 825.5 teraflop/s.
China's new Tianhe-1 supercomputer, whose name translates as "River in the Sky," rounded up the top-five list. The Chinese supercomputer at the National Super Computer Center in Tianjin, China currently tackles issues such as oil exploration and simulating aircraft designs. Several companies may also count themselves winners in this supercomputer race. IBM and Hewlett-Packard claimed 186 and 210 systems of the top 500, and Intel processors power 402 of the 500 systems.
As for national geek pride, the U.S. maintains a healthy lead, with 277 out of the 500 systems. The UK leads Europe in this category, with Germany and France tied for second. China also holds dominance over the Asian share, followed by Japan and India.
Five amazing, clean technologies that will set us free, in this month's energy-focused issue. Also: how to build a better bomb detector, the robotic toys that are raising your children, a human catapult, the world's smallest arcade, and much more.


Online Content Director: Suzanne LaBarre | Email
Senior Editor: Paul Adams | Email
Associate Editor: Dan Nosowitz | Email
Assistant Editor: Colin Lecher | Email
Assistant Editor: Rose Pastore | Email
Contributing Writers:
Rebecca Boyle | Email
Kelsey D. Atherton | Email
Francie Diep | Email
Shaunacy Ferro | Email
I gotta get me one of those!
I see where the editor dithered about how to represent the Oak Ridge 1.075 petaflops as 1..75
(I can hear him/her saying ... "Yeah that looks like more that 1.04 pflops to me!")
A decimal point here ... a decimal point there ... pretty soon it adds up to real FLOPS!
Now that you notice it I do see where he tried to up the importance of Jaguar vs Roadrunner. Although me being new to the petaflops and terathings, wouldnt have noticed a seemingly small difference of .675. Shame on me...
It might be something more sinister. Since the decimal and the zero are side-by-side on the numeric keypad, could it possibly be...no...not a typo.
"a repartitioning of the system dropped is performance from 1.105 petaflops per second to just 1.04 petaflops, and allowed the Cray XT5 supercomputer at the Department of Energy's Oak Ridge laboratory to seize the lead with a performance of 1..75 petaflops."
I agree that the number should be 1.075, not 1..75. But the whole discussion rings as false as the old Univac claim that IBM's multi-tasking operating system added "overhead." The fact that two programs could run concurrently in less time than serially was conveniently ignored. In the case of Jaguar, repartitioning was not likely have been done to reduce performance. Presumably there were dual-partition gains that out-weighed the benefit of having one large processor. The discussion is therefore one about how the machine is being used rather than maximum capability.
Does it run Flash?