"Workload Characterization for Media Processor Design"

Prof. Lizy K. John
UT Austin

Friday, April 9th, 2:00 PM, ENS 302

ljohn@ece.utexas.edu


Abstract

Media Processing has emerged to be a dominant computing workload even for general purpose processors. Despite the recent attention given to media processing and media instruction set extensions, the understanding of multimedia applications is fairly limited. Many benchmarking studies on media processors have used media and signal processing kernels. General purpose computer designers have understood the dangers of designing machines based on kernels versus full applications. The nonavailability of compiler infrastructure to fully utilize the media processing extensions has severely limited adequate analysis of the media processing extensions. This talk describes several issues in the design of a high performance media processor. Has sufficient workload characterization been done for this class of applications? Are approaches used in general purpose processor design equally valid for digital signal processor design? This talk attempts to analyze some of these issues.

Biography

Dr. Lizy Kurian John is an assistant professor in the Department of Electrical and Computer Engineering at the University of Texas at Austin. She received her Ph. D in Computer Engineering from Penn State in August 1993. Her research interests include high performance processor and memory architectures, superscalar and superpipelined processors, cache memories, program behavior studies, compiler optimization for high performance processors, rapid prototyping, Field Programmable Gate Arrays etc. Her research is supported by the National Science Foundation, the State of Texas Advanced Technology program, IBM, AMD and Intel. She is recipient of an NSF CAREER award and a Junior Faculty Enhancement Award from Oak Ridge Associated Universities. She is a member of IEEE and its Computer Society and ACM and ACM SIGARCH. She is also a member of Eta Kappa Nu, Tau Beta Pi and Phi Kappa Phi.


A list of Signal and Image Processing Seminars is available at from the ECE department Web pages under "Seminars". The Web address for the Signal and Image Processing Seminars is http://anchovy.ece.utexas.edu/seminars