Plasma Simulation on Networks of Workstations using the Bulk-Synchronous Parallel Model

Computationally intensive applications with frequent communication and synchronization require careful design for efficient execution on networks of workstations. We describe a Bulk-Synchronous Processing (BSP) model implementation of a plasma simulation and use of BSP analysis techniques for tuning the program for arbitrary architectures. In addition, we compare the performance of the BSP implementation with a version using MPI. Our results indicate that the BSP model, serving as a basis for an efficient implementation, compares favorably with MPI. Keywords: BSP Model, Networks of Workstations, Parallel Computing, Plasma 1 Introduction Complex scientific applications such as plasma simulation rely on parallel processing for efficiency. The General Concurrent Particle In Cell (GCPIC) method [2] partitions the particle and field data among the processors to distribute the computational load. However, this method leads to frequent communication and synchronization among the component p..