IMPLEMENTATION AND PERFORMANCE EVALUATION OF A PARALLEL OCEAN MODEL

Ocean models have been used to study global phenomena such as the El N ino Southern Oscillation. The effects of El Nino are felt worldwide an d are of importance to all segments of society. High resolution ocean models are being developed to run more sophisticated simulations. The demand for computing power is expected to increase drastically with th e use of high resolution models. Running extended simulations with hig h resolution models on current vector supercomputers will require comp uting power which is not available or prohibitively expensive. We have developed a parallel implementation suitable for running high resolut ion simulations. We studied performance issues such as the communicati on overhead and cache hit rate. Our experiments with grid partitioning and inter-processor communication techniques enabled us to obtain an efficient parallel implementation. Three grid partitioning techniques are evaluated. Two schemes to gather and scatter global grid data amon g processors are compared. A cache model to analyze performance is des cribed. (C) 1998 Published by Elsevier Science B.V. All rights reserve d.