Parallel Matrix Multiplication Algorithms ⚡

Parallel and Serial Algorithms for matrix multiplication implemented in C++ for GNU/Linux. This implementation executes a serial algorithm along with two parallel algorithms: OpenMP and POSIX Threads. Both matrices of the .txt files contained within the repository have a total of 1,048,576 elements.

The default number of threads to be used by the parallel algorithms is 64. You can customize the number of threads to be used inside the algorithms.hpp file:

constexpr int number_threads {64};

To run this program first make sure you have OpenMP installed. You can install it on Debian-based distributions with the following command:

sudo apt install libomp-dev

Now clone the repository:

git clone https://github.com/jc-juarez/parallel_matrix_multiplication

Then simply use the makefile provided within the repository:

make

And finally run the output executable:

./output

Tested in Ubuntu 20.04 LTS.