Parallel Computing

Introduction

This repository contains a collection of parallel computing algorithms implemented in C++ using MPI.

The implementations have been tested on a Windows 11 Machine, compiled using MSVC running the C++20 standard and MSMPI v10.0.

Preview

Simulation of diffusion of heat in iron.

Programs

Monte Carlo Simulation for Pi

Matrix Multiplication

Heat Simulation | Read the doc before trying to run this one, or you might be a little disappointed.

Running the code

To run the code, you need to have MS-MPI installed on your machine.

After compiling the code, navigate to the output directory and run the following command:

mpiexec [-n <number_of_processes>] <executable_name> <program_name> [<program_arguments>]

For a quick interactive preview without arguments, you can run:

mpiexec [-n <number_of_processes>] <executable_name>

Or if you don't want to run it parallelly, just:

<executable_name>

Though I'm not sure why you'd want to do that.

You can set the number of processes to 1 if you want to compare performance based on number of threads.

To run, say, Monte Carlo Pi estimation with 6 processes and 2e8 iterations:

mpiexec -n 6 ParallelComputing.exe monte_carlo 200000000