This course gives an introduction to the Message Passing Interface (MPI), the dominating distributed-memory programming paradigm in High Performance Computing. The following topics are covered:
- Basic principles of distributed-memory computer architecture and the Message Passing Interface (MPI)
- Blocking and non-blocking point-to-point communication
- Blocking and non-blocking collective communication
- Derived data types
- Subcommunicators, intercommunicators
- Performance issues
The lectures are accompanied by hands-on exercises.
Learning Objectives
At the conclusion of the course, you will be able to:
- understand the principles of HPC cluster architecture and distributed-memory parallel programming on clusters,
- employ the fundamental communication primitives of MPI,
- use derived data types in order to simplify complex comunication requirements,
- employ communicators and subcommunicators,
- understand the most common performance issues with MPI programming and parallel programming in general,
- employ a tracing tool for simple MPI program analysis.
Participation
Participation is free of charge for attendees from German universities, academic computing centers, and research institutions. Please only register for the course if you are really going to attend. No-shows will be blacklisted and excluded from future NHR@FAU events.
Certificates
All actual course attendees will receive a course certificate.
Prerequisites
- Familiarity with one of the standard HPC programming languages (C, C++, or Fortran)
- Ability to handle the Linux command line via a remote connection (editing, compiling)
Instructors
Dr. Alireza Ghasemi, Dr. Georg Hager (NHR@FAU)
This course is organized by Erlangen National High Performance Computing Center (NHR@FAU)
