tcollins6049/parallel-computing-final

Parallel-Computing-Final

By: Tyler Collins and James Canterbury

---

Introduction

In this project we implemented matrix multiplication using OpenMP and Cuda technology. We gathered data on how long it took the programs to run and compared them to a naive matrix multiplication program. We then visualized the results.

---

Contents

• The graphs folder contains a python file (ompGraphs.py), which is used to create the visualizations, and 6 csv files which contain the collected data.
• The matrixCuda folder contains all files needed to run the Cuda technology implementation of matrix mulitiplication.
• The matrixOpenMP folder contains all files needed to run the OpenMP implementation of matrix multiplication.
• The matrixP2 folder contains other implementations of matrix multiplication such as the naive version.
• Matrix Multiply with OpenMP and Cuda presentation.
• README.md: Contains project description and information needed to run the programs.

---

How to run

• To run the OpenMP implementation, add the files from the matrixOpenMP folder to your directory. Make sure you are on the student2 machine and run with the command "./matrixMult".
• To run the Cuda implementation, add the files from the matrixCuda folder to your directory. Make sure you are on the Cuda machine and run using the command "./matrixMult".
• To see the visualizations, add the files from the graphs folder to your repository and run ompGraphs.py.

---