README.md

LU Decomposition of Matrix using Row-Pivoting

• This repository contains Sequential and Parallel implementations of LU decomposition using row-pivoting that use Gaussian elimination to factor a dense N x N matrix into an upper-triangular one and a lower-triangular one, in C++.
• Includes analysis of comparisions of sequential, pthread and openmp implementations.

Instructions to run the code

1. To clean log files and executables: make clean
2. To change problem size and number of threads:
   • Change the value of N and PTHREAD_COUNT from constants.h accordingly
3. To compile the different modes of LU decomposition:
   • Sequential Implementation: make seq
   • Parallel Implementation using pthreads: make pth
   • Parallel Implementation using openmp: make omp
   • All implementations compile: make all
4. Run the different codes:
   • ./sequential
   • ./pth_impl
   • ./omp_impl
5. You can view the time taken by different implementations in log.txt