Parallelization of hydrocodes on the Intel hypercube
Darrell L. Hicks, Lorie M. Liebrock, Vincent A. Mousseau, G.A. Mortensen · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 1987
This report describes the Intel hydrocode parallelization project. The project consists of running four solution methods on two test problems. The code used is a modified version of one written for a project on the Heterogeneous Element Processor (HEP). In the course of running these test problems, timings were done to determine the optimal number of processors to use in solving each problem. The final conclusion drawn from this experiment is that while the hypercube architecture is appropriate for this class of problems the high communication overhead combined with too little memory makes the Intel hypercube a poor implementation target.