Parallel Algorithm for the Solution of Nonlinear Poisson Equation of Semiconductor Device Theory and Its Implementation on the MPP

Abstract The solution of the nonlinear Poisson equation of semiconductor device theory is important for the design of submicron devices used in VLSI circuits. The conventional technique applied for the solution of this equation is based on the application of Newton's method to simultaneous discretized equations. This technique requires large computational and storage capacities to handle the fine meshes associated with the modeling of submicron semiconductor devices. A new algorithm for the numerical solution of this equation has recently been developed. One advantage that this algorithm has over conventional techniques is that it is inherently parallel, and is thus well suited to implementation on a computer with large parallelism. This paper describes the initial implementation and testing of this algorithm on NASA's Massively Parallel Processor (MPP).