Distributed Memory Parallelization for Fast Volume Integral Equation

In this paper, we show distributed memory parallelization for the Method of Moment fast volume integral equation using the task bucket algorithm. The fast volume integral equation is computed using Adaptive Cross Approximation based multi-level H-Matrix. The near-field and far-field block matrix is distributed using a task bucket. Individual MPI process computes the near and far-field matrices in each task bucket, and after computation, the matrices can be used to compute parallel matrix-vector products for iterative solutions.