FPGA-based Monte-Carlo computation of the electric potential in homogeneous and non-homogeneous spaces

Abstract

This paper presents a general solution for computing the electric potential in homogeneous and non-homogeneous media using a Monte Carlo-based method. The implementation relies on an original framework that uses FPGAs to improve the computational speed.The calculation process relies on a series of both geometric and electric parameters describing the space, which is modeled by a regular geometrical structure.The computational effort strongly depends on the parameters' variation range, which severely limits the performance of software implementations that rely on general-purpose processors.Compared to the previous software implementations, this approach allows increasing the variation range of all parameters involved in the computation, thus yielding more accurate and faster results. The obtained computation time is more than one order of magnitude less than classical software solutions.The computation of the electric potential can be considered a case study used to validate the method. However, the framework is scalable and can be easily extended to a wide class of similar problems.