High Performance Computing
Content: Scientific calculations require more and more computing power. To provide this, clusters use a large number of CPU cores. Creating programs that take advantage of such architectures requires programming techniques that go beyond the capabilities of serial programming languages. The course offers an introduction to the common programming paradigms OpenMP and MPI, which can be used to parallelize C/C++ and Fortran programs. Customized tools for error analysis and performance bottlenecks for parallel programming are also covered. In addition, knowledge for the use of the PALMA cluster will be taught. This includes the batch system and the monitoring tools. The following topics are covered in the course:
- Using the PALMA cluster
- Programming with OpenMP
- Programming with MPI
- Python and HPC
Course format: In-class lectures with practical exercises and oral final exam.
Target Group: Basic programming knowledge in C/C++ as well as the confident handling of the Linux command line are required.
Credit Points: 2 CP within the General Studies in case of successful participation in the course and the final exam.