• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Nvidia Reportedly Preps Competitor for AMD Opteron and Intel Xeon

jpiniero said:
Funny, I thought Project Denver was vaporware.
Click to expand...
They were going to do x86, then they were going to do JIT x86 (like Transmeta), then Intel forced them to not do x86 at all. They've had several set-backs.

Note that the supercomputers using Teslas also use fast x86 CPUs, and quite a bit of them. The Quadros-sans-outputs are dumb worker units, and are only good for a small (albeit very hot) portion of the code.

If they do it right, they could both get rid of those pesky x86 bits, profit more per unit, and bring high-FLOPS HPC chips to customers at lower total system prices in the process (Tesla as a full SBC!).

It could be fake, but it would very much surprise me if NVidia weren't doing this. ARM's cores must be highly compatible, and are not tuned to any particular layouts. Like Qualcomm, NVidia would have the opportunity to do things like design their own LSUs and buffers for the CPU based on the cache hierarchy of the GPU. Lots of work for little improvements here and there, that we take for granted when we see a higher performing Intel or ADM CPU, but that most SoC makers can't do much of.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top