Simulation of compressible viscous flows on a variety of MPPs: computational algorithms for unstructured dynamic meshes and performance results Original Research Article

We report here on our effort in simulating unsteady viscous flows on the iPSC-860, the CM-5 and the KSR-1 MPPs (Massively Parallel Processors), using a Monotonic Upwind Scheme for Conservation Laws finite volume/finite element method on fully unstructured fixed and moving grids. We advocate mesh partitioning with message passing as a portable paradigm for parallel processing. We present and discuss several performance results obtained on all three MPP systems in terms of interprocessor communication costs, I/O, scability and sheer performance.