Intel To Power DOE’s New CTS-1 Supercomputing Clusters

[HardOCP News]

Penguin Computing today announced with Intel that the U.S. Department of Energy's National Nuclear Security Administration (NNSA) will install Penguin Computing's Tundra™ Extreme Scale (ES) series, powered by Intel® Xeon® processors, to bolster computing for the NNSA's mission of ensuring the safety, security and reliability of the nation's nuclear stockpile. The systems are being procured under NNSA's tri-laboratory Commodity Technology Systems program, or CTS-1, as part of the NNSA's Advanced Simulation and Computing (ASC) program and will serve Los Alamos, Sandia and Lawrence Livermore national laboratories. The Tundra ES series, an instantiation of the Intel® Scalable System Framework, is based on a high-density Open Compute architecture and features future generation Intel Xeon processors to deliver a peak performance range of 7-9 petaflops. When complete, these supercomputing clusters will be one of the world's largest Open Compute-based installations. For more information, check out the Penguin Computing release.

 

[Armisael]

Just Xeons? Honestly, do people even make supercomputers without humongous GPGPU clusters? I wonder if they're using Xeon Phis and including those under the Xeon name.

(Not calling you out Steve - I don't see any mention of anything else even in Penguin's press release).

 

[rabidz7]

x86 junk. I can't imagine why a supercomputer wouldn't be based on PowerPC or at least a better µarch.

 

[Armisael]

x86 junk. I can't imagine why a supercomputer wouldn't be based on PowerPC or at least a better µarch.

Around 90% of the top 500 are on x86-64. The top 10 is a little more diverse - only half of those are on x86-64.

 

[AVogel]

And all that supercomputing power will still be hackable by some 13 year old British kid because the software sucks.