Abstract
In this paper, we introduce one novel hardware acceleration technique based on a vector unit built in a regular CPU for high performance electromagnetic simulation. We investigate the performance of parallel FDTD method on the Intel and AMD processors accelerated by the Vector Arithmetic Logic Unit (VALU), high performance cluster, and Graphics Processing Unit (GPU). The FDTD method that is parallel in nature is one of the most popular numerical methods to simulate various electromagnetic problems and phenomena. Several examples are employed to demonstrate the engineering applications of parallel conformal FDTD method based on the VALU acceleration.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
