#mpi workloads performance on #mapr Data Platform Part 2 — Matrix Multiplication
▻https://hackernoon.com/mpi-workloads-performance-on-mapr-data-platform-part-2-matrix-multiplica
MPI workloads performance on MapR Data Platform Part 2 — Matrix MultiplicationIn the first part of this series, we showed how we can use MPI on top of the MapR Data Platform to successfully find prime numbers within a rather large range. Also, we compare our Sieve of Eratosthenes implementation in MPI and the same algorithm in #spark just to discover how they both behave while looking at some interesting implementation details.In the second part of the series, we are going to implement Matrix Multiplication in both, MPI and Apache Spark. Again, we will look at how each implementation behaves when running on MapR. However, our MPI implementation will be based on Cannon Algorithm while in Spark we will use the MLlib BlockMatrix functions for multiplying matrices.Cannon AlgorithmThere is (...)