Enhancing Multicore System Performance Using Parallel Computing with MATLAB

Enabling you to select the programming paradigm to work well in a multicore system

Description

A visit to the neighborhood PC retail store provides ample proof that we are in the multicore era. The key differentiator among manufacturers today is the number of cores that they pack onto a single chip. The clock frequency of commodity processors has reached its limit, however, and is likely to stay below 4 GHz for years to come. As a result, adding cores is not synonymous with increasing computational power. To take full advantage of the performance enhancements offered by the new multicore hardware, a corresponding shift must take place in the software infrastructure—a shift to parallel computing.
MATLAB® and Parallel Computing Toolbox™ address the challenge of getting code to work well in a multicore system by enabling you to select the programming paradigm most appropriate to your application. Using a typical numerical computing problem as an example, this article describes how to use the two most basic of these paradigms: threads and parallel for-loops.

By Piotr Luszczek, The MathWorks

This article was published in MATLAB Digest, September 2008, which you can read at http://www.mathworks.com/company/newsletters/?s_cid=nws_flex

• MATLAB release: MATLAB 7.6 (R2008a)