A SCALABLE STRATEGY FOR THE PARALLELIZATION OF MULTIPHYSICS UNSTRUCTURED MESH-ITERATIVE CODES ON DISTRIBUTED-MEMORY SYSTEMS

摘要

Realizing scalable performance on high performance computing systems is not straightforward for single-phe- nomenon codes (such as computational fluid dynamics [CFD]). This task is magnified considerably when the tar- get software involves the interactions of a range of phe- nomena that have distinctive solution procedures involv- ing different discretization methods. The problems of ad- dressing the key issues of retaining data integrity and the ordering of the calculation procedures are significant.