The Supercomputers and Parallel Data Processing course at the CMC Faculty of Lomonosov MSU is held for the 3rd grade students.

During the course students are offered to parallelize some algorithm using two technologies: OpenMP and MPI. Then they run their algorithms on MSU supercomputers: IBM Polus and IBM Bluegene. The result of the work is a report (with 3D-graphics) on the effectiveness of the proposed parallelization.

I was offered to parallelize 1-dimensional jacobi algorithm. Initial realization of the algorithm you can see in jacobi-1d.c.

In the directories OMP and MPI you can see jacobi-algorithm parallelization using each techology. Both solutions are written on C. Graphics are built using Python Matplotlib library.