Proc. Asilomar Conference Signal, Systems and Computers, Nov. 7-10, 2010, Pacific Grove, CA USA.

Multi-core Sonar Beamforming with Computational Process Networks

John F. Bridgman, Gregory E. Allen and Brian L. Evans

Department of Electrical and Computer Engineering, The University of Texas at Austin, Austin, TX 78712 USA
gallen@arlut.utexas.edu - bevans@ece.utexas.edu

Paper - Poster

Software Release

Process Networks Research at UT Austin

Abstract

This paper evaluates the scalability with respect to processor cores of a three-dimensional sonar beamforming kernel implemented on a multi-core workstation. Beamforming is an example of an extremely parallelizable problem. This implementation is instrumented with OpenMP to exploit multi-core computer systems. However, when executed on a 16-core machine, this kernel scales much less than expected. We implement this beamformer system within the scalable framework of Computational Process Networks to achieve additional performance and processor utilization for a larger number of cores. On our benchmark machine, the implementation with Computational Process Networks obtains a throughput speedup of more than two times over OpenMP with the default settings, and 13% improvement in throughput over OpenMP with optimized settings.


COPYRIGHT NOTICE: All the documents on this server have been submitted by their authors to scholarly journals or conferences as indicated, for the purpose of non-commercial dissemination of scientific work. The manuscripts are put on-line to facilitate this purpose. These manuscripts are copyrighted by the authors or the journals in which they were published. You may copy a manuscript for scholarly, non-commercial purposes, such as research or instruction, provided that you agree to respect these copyrights.


Last Updated 04/23/07.