Following official unveiling of Nvidia’s next-gen Fermi architecture, Nvidia have notched up a big win in the HPC industry already.

Oak Ridge Natural Laboratory (ORNL) will build a supercomputer using Nvidia’s Fermi products, intended to be 10 times more powerful than today’s fastest supercomputer. This is a massive magnitude to overtake what is today world’s most powerful supercomputer.

More details next page.

Following official unveiling of Nvidia’s next-gen Fermi architecture,
Nvidia have notched up a big win in the HPC industry already.

Oak Ridge Natural Laboratory will build a supercomputer using Nvidia’s
Fermi products, intended to be 10 times more powerful than today’s
fastest supercomputer. This is a massive magnitude to overtake what is
today world’s most powerful supercomputer.

Mr. Jeff Nichols, associate director for Computing and Computational Sciences at ORNL, says with Nvidia’s Fermi, substantial scientific breakthroughs would not be possible. By using Fermi’s parallelization as well as its greater programmability compared to other GPUs, scientific computing will be done at a speed many times faster than CPUs, which currently power most supercomputers.

ORNL has already assembled a Hybrid Multicore Consortium, which will program applications to run and utilize Fermi to its potential.

This is undoubtedly a massive boost to Nvidia’s aspirations to be a major player in the HPC industry.

Reference: Nvidia Press Release