Multiple data parallelism with HPF and KeLP

High Performance Fortran (HPF) is an effective language for implementing re gular data parallel applications on distributed memory architectures, but i t is not well suited to irregular, block-structured applications such as mu ltiblock and adaptive mesh methods. A solution to this problem is to use an SPMD program to coordinate multiple concurrent HPF tasks, each operating o n a regular subgrid of the multiblock domain. This paper presents such a sy stem, in which the coordination layer is provided by the C++ class library KeLP. We describe the KeLP-HPF implementation and programming model, and sh ow an example KeLP-HPF multiblock solver together with performance results. (C) 1999 Elsevier Science B.V. All rights reserved.