Principles and capabilities of 3-D, E-M particle simulations

Abstract Presently available computing power is just sufficient for carrying out some meaningful three-dimensional, fully electro-magnetic particle simulations. The background for building the appropriate codes is reviewed in this article, and the modest achievements of one such code, SPLASH, (operating on a CDC 7600) are presented. A more ambitious code for the CRAY1, still under construction, is based on similar principles and algorithms. Data management emerges as a crucial element, in addition to the physics and numerics, for determining code structure.