Acceleration of a physical-optics simulator using CUDA

This paper reports on the current state of a work-in-progress porting of a physical-optics simulation tool onto NVIDIA's CUDA platform. Current accelerator APIs are shortly presented. Our choice for the CUDA platform is explained, as well as the data flow of the simulation tool. The current state of the implementation of the port is presented, as are first run time measurements. The results are promising; however, the implementation is quite näıve at the moment. Further work will address the mentioned issues and increase the performance of the simulation tool.