CSE Colloquium Series
Data Prefetching: Taxonomy and a New Approach
Friday, March 21, 2008
Talk: 11:00 am - 12:00 pm
Host: Matt Mutka
Data prefetching has been considered as an effective way to mask data access delay and to bridge the performance gap between processor and memory. With hardware and/or software support, data prefetching brings data closer to the processor before it is actually needed. Many prefetching techniques have been developed to reduce data access latency. In this talk, we first introduce a taxonomy that characterizes and classifies various design concerns in developing a prefetching strategy. We then propose a new prefetching approach; the server-push based prefetching. Finally, we present our ongoing NSF HECUR project, the Server-Push I/O Architecture for High End Computing. In this research a File Access Server is designed and developed to actively push data to memory before an I/O is called.
Professor Xian-He Sun received his Ph.D. in computer science from Michigan State University in 1990. He currently is a professor of computer science and the director of the Scalable Computing Software laboratory at Illinois Institute of Technology (IIT), and is a guest faculty in the Mathematics and Computer Science Division and Computing Division at the Argonne and Fermi National Laboratory, respectively. Before joining IIT, he worked at DoE Ames National Laboratory, at ICASE, NASA Langley Research Center, at Louisiana State University, Baton Rouge, and was an ASEE fellow at Navy Research Laboratories. Dr. Sun's research interests include parallel and distributed processing, software systems, performance evaluation, and high-end computing.