MPI is the dominant programming model used to parallelize applications on all large-scale parallel machines, as well as on Linux and Windows clusters of all sizes.Content:

    Introduction to MPI and environment

    Point-to-point communications

    Collective communications

    One-sided communications

    Derived datatypes

    Optimizations

    Communicators

    MPI-IO

    Hands-on sessions 

Learning outcomes
After this course, participants should know how to parallelize efficiently an application with the MPI library, using basic (P2P or collective communications …) or more advanced features (derived datatypes, overlapping of computation and communication…) of MPI.Prerequisites
A good knowledge of C or Fortran programming language.

    Parallel programming knowledge is not required.

https://events.prace-ri.eu/event/53/