Abstract: We present a parallel implementation of the fast boundary element method (BEM) for the Helmholtz equation. After a brief description of BEM, vectorization of the computationally most demanding kernels, and shared memory parallelization, we focus on the distributed memory parallelization using a new approach for distribution of the fast BEM system matrices among computational nodes. Moreover, we present a modification of the adaptive cross approximation (ACA) method capable of dealing with BEM matrices containing zero blocks, which usually lead to problems for the original ACA algorithm. Numerical experiments demonstrating the performance of the implementation were carried out using several Intel architectures.
Best Poster Finalist (BP): no
Poster summary: PDF
Reproducibility Description Appendix: PDF
Back to Poster Archive Listing