Course Material for the UIBK Parallel Systems Lab will be provided in this repository.
- Shared memory parallelization (OpenMP + more)
- Performance-oriented programming
- Correct benchmarking
- Distributed memory parallelization (MPI)
- Dependence analysis (Real codes)
- Tools (Profilers, sequential and parallel)
All code will be written in C++. (Note: this refers to application code - scripts for e.g. performing benchmark runs may be written in any suitable language)
Exercises should be tracked on some git repository, and access to that repo should be provided to Peter. The exercises need to be completed in time for each weeks' lab.
Exercises should be completed in groups of 2-3 students. Grades are based on the regular presentations of solutions, and Q&A during these presentations.