IBM and Nvidia have announced plans to work on accelerated GPUs for use on IBM Power Systems.
That means Nvidia chips will be used in enterprise level data centres.
Both companies will work to integrated Tesla GPUs with IBM Power CPUs and means Nvidia will have a share in providing techn for big data analytics and high performance supercomputing.
Ian Buck, VP of accelerated computing at Nvidia said that the move allow existing supercomputing customers with technologies for energy efficient systems used in scientific applications and in data analytics.
As part of the deal, IBM will provide its Rational enerprise software development tools to supercomputing developers.
And in other news, Nvidia said today it is launching the Tesla K40 GPU accerlerator which it claims will give double the memory and 40 percent higher performance than its K20X GPU.
Features include 12GB of GDDR5 memory, 2,800 CUDA parallel processing cores, dynamic parallelism and it will support PCIe Gen-3 interconnect support.