March 2022 / MPI introduction
Part of Scientific Computing in Practice lecture series at Aalto University.
Audience: Employees and students looking for the extensive intro into MPI programming model.
About the course: In this introduction to MPI programming we are going to cover all the basic functionalities of the API, and see how these are applied to real problems. Starting from simple, well-established algorithms, we will build up towards more advanced and realistic examples commonly found in computational physics and chemistry. Finally, we will explore how the strengths of OpenMP multithreading and MPI can be combined with hybrid parallel programming.
The course is focused on practical aspects of parallel programming applied to problem solving. The language of choice is C, but few words will be spent on how to use MPI in FORTRAN if required by the students. Students are expected to have a basic understanding of conventional programming, and some familiarity with C.
Lecturer: Filippo Federici, D. Sc., Department of Applied Physics, Aalto University
Time, date:
Date |
Time |
|---|---|
Tue 8.3 |
10:00-13:00 |
Fri 11.3 |
10:00-13:00 |
Tue 15.3 |
10:00-13:00 |
Fri 18.3 |
10:00-13:00 |
Place: We are trying to arrange this as hybrid workshop. Join us in Otaniemi or via streaming link provided to registered participant.
Cost: Free of charge for FGCI consortium members and Aalto employees, students.
Registration: Click here to register
Credits (opintopisteet): Credits available for the Aalto students and course certificate can be provided on request for the outsiders. Since the main focus of the course is learning through experience, the students will be required to perform simulations for homework and hand in result report. Credits/certificate require full time participation and handling assignments. Full course hours correspond roughly to 1 ECTS.
Other comments: Participants from Aalto university will be provided with access to Triton for running examples. We will use SSH for Triton connection. One can follow the theoretical part without access to the Triton cluster, but we highly recommend doing the tutorials together. If you cannot access the Triton HPC cluster, you need to install in your system an MPI library and a reasonable C compiler (GCC>5.4). All the functions we use should be available in any MPI implementation.
Additional course info at: filippo.federici -at- aalto.fi / ivan.degtyarenko -at- aalto.fi