Versions of a program that makes some operation in a matrix and adding it. #define matrix_size 16000

duo_thread.c - single node with 2 threads sharing the task.
quad_thread.c - single node with 4 threads sharing the task.

MPI_2nodes.c - 2 single threaded nodes doing the task.
MPI_4nodes.c - 4 single threaded nodes doing the task.

MPI_Thread.c - 2 nodes with 2 threads sharing the task in each node.

The results are not intuitive, because I wouldn't imagine that using 2 nodes with 2 threads on each would take more time than just using 2 nodes single threaded. But thinking further, We've noticed that the operations are way to easy to compute that the overhead created by the 2 threads take the longest space of time. Interesting article on that : http://www.thequestforoptimality.com/using-more-processors-does-not-necessarily-lead-to-reduced-run-times-on-cplex/