In this webinar, Satnam Singh, Professor of Reconfigurable Computing, School of Computer Science, University of Birmingham (UK), will demonstrate data-parallel programming with Microsoft’s Accelerator system. The system provides a language neutral library for expressing whole-array computations which can be dynamically compiled into code for execution on GPUs, as well as code running on multiple processor cores using SSE instructions. Professor Singh will introduce the Accelerator API and data-structures which are exposed as a domain specific library through the use of overloaded operations with specific examples in C++ and F#.
Webinar: Heterogeneous Data-Parallel Programming, Nov. 16
November 3, 2011 by