This module aims to provide students with the foundations of high-performance scientific computing (HPC). Firstly, it covers program optimization aspects by balancing memory accesses with computational portions. Secondly, parallel computing is addressed through the shared memory paradigm (openMP) as well as the distributed memory paradigm (MPI). Lastly, students must apply the methods learned to accelerate calculation codes related to fluid mechanics.
The students are graded through two independent projects, during which they must optimize and parallelize calculation codes in fluid mechanics.