Multiplying 2D Matrices with a 3D approach

In this paper, I explain a previously published three-dimensional algorithm for multiplying two two-dimensional matrices. The basic concept of this approach is to split the matrices into sub-matrices which are intelligently distributed across all of the available processes. Then, each process computes a smaller ma-trix multiplication subproblem. These intermediate results are combined to create the final answer. This algorithm takes ad-vantage of the parallel aspects of matrix multiplication while minimizing the required communication. A model has been generated to estimate the execution time of the algorithm given various necessary system parameters and the size of the input matrices. Finally, the results from the original paper are shown and discussed.