Adding GPU Acceleration to Electromagnetic Simulations in RSoft FullWAVE FDTD
Application Notes
In the design of modern photonic devices, the Finite Difference Time Domain method (FDTD) method is an important method for electromagnetic field simulation.
However, if you make the compute grid smaller or the simulation range larger to increase accuracy, the simulation can take days or weeks to complete with traditional CPU multi-process methods. To solve this problem, RSoft FullWAVE FDTD supports GPU acceleration. The parallel computing power of NVIDIA GPUs significantly reduces simulation time compared to CPU-based calculations.