Using coordination to parallelize sparse-grid methods for 3D CFD problems
The good parallel computing properties of sparse-grid solution techniques are investigated. For this, an existing sequential CFD code for a standard 3D problem from computational aerodynamics is restructured into a parallel application. The restructuring is organized according to a master/slave protocol. The coordinator modules developed thereby are implemented in the coordination language Manifold and are generally applicable. Performance results are given for both the sequential and parallel version of the code. The results are promising, the paper contributes to the state-of-the-art in improving the efficiency of large-scale computations. Also a theoretical analysis is made of speed-up through parallelization in a multi-user single-machine environment.