Hybrid Parallel FEM on PC Cluster with Multi-core Processor

We developed a 3-D hybrid parallel finite element method with the process parallelization by MPI and the thread parallelization by OpenMP for the purpose of improving parallel performance using PCs having many core CPUs. In this paper, we clarify the performance of block multicolor ordering preprocessing as a thread parallelization method through the torque analysis of an IPM motor using the 3-D hybrid parallel finite element method on a PC cluster with a 12-core processor.