Advanced Imaging on PC Clusters
J. Bee Bednar, Advanced Data Solutions

This paper discusses advanced imaging on massively parallel PC clusters. It reviews the kinds of algorithms that are suitable for cluster computation and provides some insight into alternative parallelization strategies. It attempts to compare differences between SMP and cluster environments. It focuses on parameter estimation in acoustic, elastic, and anisotropic media, as well as the technology pertinent to image data over such media. Real and synthetic examples demonstrate that the additional computational expense of using more accurate methods is well worth the cost. When implemented on the current generation of attractively priced-high-performance hardware, these emerging methods are shown to be eminently feasible and cost effective.

Biography

J. Bee Bednar received the B.S from Southwest Texas State University, and the M.A. and Ph.D. degrees in Mathematics from The University of Texas at Austin in 1962 and 1968 respectively. His experience includes research in abstract functional analysis, Anti-Submarine Warfare, Digital Signal Processing, Seismic Imaging, and Cluster Computing. At Drexel University in Philadelphia and the University of Tulsa he taught both Mathematics and Geophysics. In managerial positions at Cities Service and Amerada Hess, Bee was responsible for the management of research in seismic processing and the development of advanced cluster computing technologies. Under his direction, Amerada was one of the first large energy related companies to accept and expand cluster computing concepts. Bee has enjoyed numerous consulting opportunities at a variety of large oil companies. He is currently Vice President of Research and Development at Advanced Data Solutions in Houston, TX.