High Scalability FMM-FFT Electromagnetic Solver for Supercomputer Systems

Abstract

An efficient parallel implementation of the Fast Multipole Method (FMM) combined with the fast Fourier transform (FFT) is presented in this paper. The good scaling properties of the FMM-FFT, combined with a smart parallelization strategy, has been shown to be very effective when using large parallel supercomputers. A challenging problem with more than 150 million unknowns has been solved, demonstrating that the proposed implementation of the FMM-FFT constitutes a real alternative to the more frequently used multilevel approaches, such as the Multilevel FMM (MLFMA). Even more importantly, we have achieved a high efficiency with 1,024 parallel processors, which indeed constitutes one of the better scalability results ever reached for a rigorous integral-equation electromagnetic solver.