This Summer, one of our Intel Black Belt Software Developers, Jim Dempsey, published a multi-part article on the Parallel Programming Community to help developers to enhance their Parallel Programming Skills: Superscalar Programming 101 (Matrix Multiply). Jim Dempsey discusses how to optimally tune a well known algorithm. He takes this algorithm and showcases the common method for parallelizing this algorithm and then he outlines several different approaches to parallelizing this algorithm. The final method produces a fully cache sensitized approach to parallelizing this algorithm. Jim’s article is very detailed with some great code examples that I think you will find very useful. He has gotten some good reviews by developers that have read this article. I’d like to recommend reading this multi-part article.
Let me know what you think of Jim’s article, Superscalar Programming 101 (Matrix Multiply).
