Dynamic Load Balancing of Distributed Memory Parallel Computational Mechanics using Unstructured Meshes for Multi-Physical Modelling

V. Aravinthan, S. P. Johnson, K. McManus, C. Walshaw and M. Cross


As the complexity of parallel applications increase, the performance limitations resulting from computational load imbalance become dominant. Mapping the problem space to the processors in a parallel machine in a manner that balances the workload of each processors will typically reduce the run-time. In many cases the computation time required for a given calculation cannot be pre-determined even at run-time and so static partition of the problem returns poor performance. For problems in which the computational load across the discretisation is dynamic and inhomogeneous, for example multi-physics problems involving fluid and solid mechanics with phase changes, the workload for a static subdomain will change over the course of a computation and cannot be estimated beforehand. For such applications the mapping of loads to processors is required to change dynamically, at run-time in order to maintain reasonable effciency. The issues of dynamic load balancing are examined in the context of PHYSICA, a three dimensional unstructured mesh multi-physics continuum mechanics computational modelling code [1].


M. Cross, P. Chow, C. Bailey, N. J. Croft, J. Ewer, P. F. Leggett, K. McManus, and K. A. Pericleous. PHYSICA - a software environment for the modelling of multiphysics proceedings. In Proc. ICIAM'95, 1996.

Fri Aug 13 13:42:14 BST 2004