NVIDIA porting CUDA to run directly on x86 chips
updated 02:15 pm EDT, Tue September 21, 2010
NVIDIA CUDA to run natively on x86 chips
NVIDIA chief Jen-Hsun Huang today at its GPU Technology Conference said his company would bring its CUDA general-purpose computing language directly to x86 chips. The approach developed with the Portland Group will let systems without NVIDIA cards handle the code. It will work best with multi-core processors and is seen as ideal for servers.
A timetable isn't yet available for when CUDA will be ready to make the switch or which platforms will have direct access to x86. Both Mac OS X and Windows have native CUDA support and may get the added support relatively soon.
The move is potentially a hedge against universal standards. Both Microsoft's DirectCompute and the more universal OpenCL don't require a graphics chipset to run. Concepts of general-purpose computing languages came about to help calculate physics and other advanced math in parallel rather than depend solely on a single stream of code.
Fresh-Faced Recruit
Joined: Aug 2001
So wait...
CUDA is a language that (basically) lets general purpose tasks run on a GPU. Now you're going to take that and port it to a general purpose processor? That doesn't seem to make a whole lot of sense...