New AMD Opteron 6200 CPUs Start Supercomputing Too

Since NVIDIA got to brag more than once about its presence on the supercomputing front, AMD decided to do something similar, even getting its chips adopted by the same HPC project as the one NVIDIA spoke about.

In this case, though, the two companies don't find themselves on opposing sides.

Though they both got contracts for the Blue Waters supercomputing project, AMD will be supplying the x86 CPUs while NVIDIA has the GPU side covered (Tesla computing modules, for parallel processing).

According to the new press release from Advanced Micro Devices, the Blue Waters Project, belonging to the National Science Foundation and based at The University of Illinois' National Center for Supercomputing Applications (NCSA) will use AMD Opteron 6200 processors.

“The world's most technically savvy minds have recognized the price-performance advantages AMD has to offer the HPC community as evident by the Blue Waters supercomputer deployment and the HPCwire Readers' and Editors' Choice Awards,” said Paul Struhsaker, corporate vice president and general manager, Commercial Business at AMD.

“The HPC community demands scalable and strong floating point performance, high memory throughput and more cores for highly-threaded applications. It's these exact requirements that helped influence the design of the new AMD Opteron processors.”

For those who don't know, the Opteron 6200 and 4200 were launched just the other day.

Blue Waters will be made of about 253 Cray XE6 cabinets and will employ about 50,000 AMD Opteron 6200 processors.

These, along with the Tesla GPU computing modules from NVIDIA, will lead to a performance of over one petaflop.

Once ready, research will be carried out on tornado formation, the way viruses enter cells, the evolution of the universe since the Big Bang, etc.

On a related note, AMD has a total of two million AMD Opteron cores working as part of multiple supercomputers in the TOP500 List.