In as far as the tug of war between nVidia and AMD for the GPU “king of the hill” title, it has been a close contest these past years. Currently, with AMD/ATI’s HD400o series now done and nVidia’s GT200 project all over as well, both are priming for their new releases.
nVidia, in a bid to upstage ATI’s new 5000 series, is teasing the public with the promise of a GPU that has the processing logic close to a CPU — the new GT300 project, code named FERMI. This translates to smarter data processing and faster speeds down the line. Read more about FERMI from nVidia’s site:
The next generation CUDA architecture, code named “Fermi”, is the most advanced GPU computing architecture ever built. With over three billion transistors and featuring up to 512 CUDA cores, Fermi delivers supercomputing features and performance at 1/10th the cost and 1/20th the power of traditional CPU-only servers.
Fermi makes GPU and CPU co-processing pervasive by addressing the full-spectrum of computing applications.
Visit nVidia’s site for the full write-up on FERMI.
NVIDIA CEO Jen-Hsun Huang has already “teased” the public with his keynote at the GPU Technology Conference in San Jose, California. Overclockersclub.com reports on this, and has the first pictures of a FERMI card — the public’s first glimpse of it.
With ATI having already released their 5870 card and the 5850 almost there, this should build up to a close and exciting race for the current gen GPU wars.
