Skip to content

Latest commit

 

History

History
8 lines (4 loc) · 478 Bytes

README.md

File metadata and controls

8 lines (4 loc) · 478 Bytes

Matrix Multiplication

Matrix multiplication implemented in CUDA C++ with both GPU and CPU versions for comparison. For 1000x1000 matrices the GPU computed the solution 6 orders of magnitude faster than the CPU. (NVIDIA GeForce GTX 970 and Intel Core i5-4690 CPU)

Created using Visual Studio 2019, hence the amount of files. Relevant code is in ./CudaMatrixMultiplicationCustom/kernel.cu

Although it was only tested for square matrices, should work for rectangular ones.