Developed a CUDA version of the FDTD method and achieved a speedup 40x. Implemented on a NVIDIA Quadro FX 3800 GPU, which has 192 SPs, 1GB global memory, and a memory bandwidth of 51.2 GB/s.
Two methods are presented for efficiently computing the eigenvalues of the finite-difference Laplacian. One method embeds the region considered in a rectangle. The other method is applicable when the ...
Results that may be inaccessible to you are currently showing.
Hide inaccessible results