Matrix-Inversion-with-CUDA

I implemented a parallel algorithm for matrix inversion based on Gauss-Jordan elimination.

In this homework, the algorithm should be implemented with CUDA programs with competitive performance, which should also be compared with equivalent CPU implementations with the serial algorithm. The computed result should be verified by a matrix multiplication to get an identify matrix. For more information on Gaussian elimination and the related algorithm for finding the inverse matrix, please check the webpage at: https://en.wikipedia.org/wiki/Gaussian_elimination.

Hand in: A CUDA project containing all the necessary source files, as well as a document describing how your parallel algorithm is designed and implemented, with a graph illustrating performance gain over CPU equivalence with respect to the increasing size of the matrices.