Jen Hsun Huang, Nvidia's graphics technology conference today heads his company will bring its language CUDA general purpose computers directly chips the x86. Portland will be developed by the team's strategy for NVIDIA cards without code so that the system to handle. It will work better with multi-core processors, and are ideal for servers.
But when the time to whip to replace or x86 platforms that will have direct access is not available to be ready for. Both Macintosh and Windows, and native support for the whip, and perhaps also gain support and fairly fast.
Will a potential defense step against global standards. Together, Microsoft and more to the Global Direct Compute OpenCL sounds not need to run graphics. Come concepts of language, for general purpose computing to help the code to rely solely on a single stream but also other developing countries, parallel computing in physics and mathematics.
But when the time to whip to replace or x86 platforms that will have direct access is not available to be ready for. Both Macintosh and Windows, and native support for the whip, and perhaps also gain support and fairly fast.
Will a potential defense step against global standards. Together, Microsoft and more to the Global Direct Compute OpenCL sounds not need to run graphics. Come concepts of language, for general purpose computing to help the code to rely solely on a single stream but also other developing countries, parallel computing in physics and mathematics.
No comments:
Post a Comment