CUDA Based Fast Implementation of Very Large Matrix Computation | IEEE Conference Publication | IEEE Xplore