The series gains both extra performance and energy saving capability

Oct 1, 2013 14:55 GMT  ·  By

Intel keeps saying supercomputers are nothing without x86 chips, and NVIDIA says that the parallel processing power of GPU compute accelerators is becoming even more essential, but the truth, as seen by most of those with a stake here, is that the best idea is to use both.

Case in point, global supercomputer builder Cray Inc. has updated the XC30 series with both Intel Xeon Phi coprocessors (PCI Express accelerators based on x86 chips) and NVIDIA Tesla K20X GPU accelerators (also on PCIe).

With this, the Cray XC30 and Cray XC30-AC systems gain better application performance and much improved energy efficiency.

Cray XC30 series was previously named Cascade and was the line where Intel's Xeon Phi coprocessors made their debut.

Now that NVIDIA's Tesla line is in the game, things will only get better. I can only imagine what the next Top500 supercomputer list will look like.