Stefan Mileschin

Kicking off today is the annual International Conference for High Performance Computing, Networking, Storage, and Analysis, better known as SC. For NVIDIA, next to their annual GPU Technology Conference, SC is their second biggest GPU compute conference, and is typically the venue for NVIDIA’s summer/fall announcements. Earlier we covered the announcement of NVIDIA’s role in the US Department of Energy’s latest supercomputer acquisitions, and today we’ll be taking a look at NVIDIA’s latest Tesla GPU compute card, Tesla K80.

At SC’13 NVIDIA introduced the Tesla K40, the first “fully enabled” Kepler Tesla card. Based on the GK110B variant of NVIDA’s GPU, this was the first Tesla product to ship with all 2880 CUDA cores enabled. Typically one would think that with a fully enabled GK110 based board that the Kepler Tesla lineup would have reached its apex, but for SC’14 NVIDIA will be pushing the performance envelope a bit harder in order to produce their fastest (and densest) Tesla card yet.

http://www.anandtech.com/show/8729/n...-k80-gk210-gpu