Parallel programming technologies OpenMP and MPI


Aleksander Antonov Ph.D., leading researcher RCC MSU
Bakhtin Vladimir , Ph.D., Head of department of the Institute of Applied Mathematics. Keldysh Sciences, Assistant Professor of System Programming Faculty of Computational Mathematics and Cybernetics, Moscow Lomonosov University.


This track will be the functionality of parallel programming MPI and OpenMP, extending the core courses According to technology. It is assumed that students already have a basic knowledge and skills in this area. In the MPI will be considered: work with groups and communicators, virtual topology, sending different types of data, one-way communications, error handling, and other issues. In OpenMP will consider all modes of distribution and synchronization, as well as additional auxiliary procedures. Special attention will be given to the sharing of parallel programming in MPI and OpenMP hybrid model. The track includes a large number of case studies on the supercomputer "Lomonosov".