Acceleration of Time-Domain Finite Element Method (TD-FEM) Using Graphics Processor Units (GPU)

Abstract

The time-domain finite element method is used extensively in areas of microwave engineering. However, TD-FEM runs too slow for some simulations to be practical, especially when run on standard personal computers. The propriety of hardware for the acceleration of TD-FEM computations has been investigated. It is well known that the key of TD-FEM is matrix operation. There always be an obstacle that the speedup of dense matrices operations in Graphics Processor Units fields. This situation will be improved in next generation. Even though the presence of the obstacle, the standard consumer Graphics Processor Units (GPUs) can be used to accelerate TD-FEM simulations by a factor of close to two yet, relative to an Intel CPU of similar technology generation. With OpenGL as the Application Programming Interface (API), a standard commercial graphics card has been programmed to solve a 2-D electromagnetic radiation problem.