NVIDIA Scientist Promises ExaScale Machine in 2017
NVIDIA has been trying to make everyone believe the combination of CPU and GPU computing will deliver the maximum computing ability. Bill Dally, NVIDIA’s Chief Scientist has recently revealed that the company’s working on something powerful called “NVIDIA ExaScale Machine” which is scheduled to release in 2017.
According to the slide leaked by hpctech.com, the ExaScale GPU will pack 2400 throughput cores (7200 FPUs) and 16 CPUs on a single chip with a TDP of 300W, delivering up to 40TFLOPs of single-precision floating-point processing power or 13TDLOPs of double-precision floating-point processing capabilities. Besides, each node of the chip will feature 128GB of memory, 2TB/s bandwidth, 512GB Phase-change/Flash for checkpoint and scratch.
October 29th, 2009 at 10:28 pm
Nvidia didn’t develop this setup, they are a software company. Anyone can make a GPU, this isn’t impressive at all. lol ;P
October 30th, 2009 at 2:39 am
SOFTWARE FUCK YEAH
November 1st, 2009 at 11:28 pm
By the time AMD and Intel will have fester “traditional” systems