igli-balla: Implemented CUDA matrix multiplication

[Igli333]

The solution for this assignment contains the naive and tiled implementation of matrix multiplication on GPUs. It tries to optimise the whole process by allowing more work do be done at the same time by the GPU.

• The Dione cluster offered by Abo Akademi and University of Turku was used to run the code.
• The main challenge was fixing a few issues where the tiled multiplication would calculate only a certain part of the matrix and not all the values.