U.S. finally reclaims world’s top supercomputer throne with 16 petaflop DOE cluster

Sorry, Japan, America is once again home to the world’s fastest supercomputer — for the first time since November 2009.

The newly created “Sequoia” cluster at the Department of Energy’s Lawrence Livermore National Laboratory scored a whopping 16.32 petaflops with the Linpack benchmark, according to a new list released by Top 500 Supercomputer Sites today. The cluster ousts Fujitsu’s “K Computer” in Japan from the number one spot. That computer now sits at the number two spot with 10.51 petaflops.

(Read more on what petaflops are all about in a related story about Intel’s new 50-core Xeon Phi chip.)

Sequoia is an IBM BlueGene/Q system sporting 1,572,854 cores across 96 racks and 98,304 computer nodes. The Linux-based  cluster hogs 1.6 petabytes of memory and takes up 4,500 square feet. In short, Sequoia is a beast — and that’s a good thing, as it could help us to avoid a bleak nuclear future.

“Sequoia will provide for necessary advances in the understanding of weapons science as well as vastly improved estimates of uncertainty in predictions of weapons behavior. Both are vital to ensuring confidence in our nuclear deterrent as these weapons age decades beyond their intended deployment life,” Bruce Goodwin, Principal Associate Director for Weapons and Complex Integration at LLNL, said in a 2009 interview. “The advances in applications and the hardware on which Sequoia will be built represent a giant step forward for the predictive simulations that are at the heart of certification of the nuclear stockpile.”

Via Ars Technica; Photo via LLNL

blog comments powered by Disqus