high performance computing on graphics processing units: hgpu.org

This aim of this project was to port the FFT routines of LLRP to CUDA, which was done successfully. This success is quantified as the FFT portions of the program executing in a much shorter time than the FFTW transforms. The project shows that GPUs are certainly viable for use in numerical codes such as this which hopefully will encourage further research in this area. However, due to the difficulty of porting some of the code to CUDA, the performance was severely limited by memory transfers. If better normalisation code were to be written, even stronger performance gains could be realised.