Solving large systems of linear equations in serial is time consuming and slow process.
Prallelizing the algorithms using suitable parallel constructs can provide the correct solutions with a lesser time complexity.
This project presents parallel and serial implementation of three algorithms for solving system of linear equations:
- Back Substitution,
- Conjugate Gradient and
- Gauss Seidel
In this project, OpenMP is used for parallelizing the algorithms.
For all the 3 algos, serial and parallel implementations is provided. Execution time for each of the 6 functions(serial and parallel implementations of the three algos) is calculated to determine and henceforth compare the efficiency of the three algorithms.
Two code files: required.c and random.c are basically the same implementation, but random.c considered a random matrix input, while required.c considers specific input matrix considering the limitations of the algorithms.