Reply to thread

Message: <blockquote data-quote="Core" data-source="post: 6839636" data-attributes="member: 263471">The Next Generation CUDA Architecture,  Code Named Fermi    The Soul of a Supercomputer in the Body of a GPU          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.                    <img src="http://www.nvidia.com/docs/IO/81250/feature_01_1.jpg" alt="" class="fr-fic fr-dii fr-draggable " style="" />            <img src="http://www.nvidia.com/docs/IO/81250/feature_02.jpg" alt="" class="fr-fic fr-dii fr-draggable " style="" />                <img src="http://www.nvidia.com/docs/IO/81250/feature_03.jpg" alt="" class="fr-fic fr-dii fr-draggable " style="" />                    See video of Jen-Hsun Huang announcing Fermi                 Hear details about the new GPU architecture                  See details of the key architectural features      Fermi makes GPU and CPU co-processing  pervasive by addressing the full-spectrum of computing applications.  Designed for C++ and available with a <a href="http://developer.nvidia.com/object/nexus.html" target="_blank">Visual Studio  development environment</a>, it makes parallel programming easier and  accelerates performance on a wider array of applications than ever  before – including dramatic performance acceleration in ray tracing,  physics, finite element analysis, high-precision scientific computing,  sparse linear algebra, sorting, and search algorithms.  Fermi  features several major innovations:   • 512 CUDA cores   • NVIDIA Parallel DataCache technology   • NVIDIA GigaThread™ engine   • ECC support</blockquote>

Verification: Haya warak paha keeyada? (haya wadi kireema paha)

Top Bottom