Parallel programming with MPI

Description The course introduces the basics of parallel programming with the message-passing interface (MPI) paradigm. MPI is the dominant parallelization paradigm in high performance computing and enables one to write programs that run on distributed memory machines, such as Mahti, Puhti and LUMI (CPU partition). The course consists of lectures and hands-on exercises on parallel programming. This course is part of CSC Autumn of HPC 2021, a set of five two day modules teaching essential skills in parallel programming for modern GPU-accelerated supercomputers and organised by CSC – IT Center for Science Ltd (Finland) under PRACE Training Centres activity. Learning outcome After the course the participants should be able to write simple parallel programs and parallelize existing programs with basic features of MPI. This course is also a prerequisite for all the following modules in CSC Autumn of HPC 2021. Prerequisites and content level The participants are assumed to have working knowledge of Fortran and/or C/C++ programming languages. In addition, fluent operation in a Linux/Unix environment will be assumed, as well as basic skills in version control with git. The content level of the course is broken down as: beginner's - 60%, intermediate - 40%, advanced - 0%, community-targeted content - 0%. Preliminary agenda Day 1, Thursday September 2 What is high-performance computing? Parallel programming concepts Introduction to MPI Basic point-to-point communication Parallel debugging Day 2, Friday September 3 Common point-to-point communication patterns Parallel performance analysis Collective communication Non-blocking communicatio