The CMU task parallel program suite

Abstract

The idea of exploiting both task and data parallelism in programs is appealing. However, identifying realistic, yet manageable example programs that can benefit from such a mix of task and data parallelism is a major problem for researchers. We address this problem by describing a suite of five applications from the domains of scientific, signal, and image processing that are of reasonable size, are representative of real codes, and can benefit from exploiting task and data parallelism. The suite includes fast Fourier transforms, narrowband tracking radar, multibaseline stereo imaging, and airshed simulation. Complete source code for each example program is available from the authors. This research was sponsored by the Advanced Research Projects Agency/CSTO monitored by SPAWAR under contract N00039-93-C-0152, and also by the Air Force Office of Scientific Research under contract F49620--92--J--0131. The views and conclusions contained in this document are those of the authors and shoul...