Practical Course on Parallel Computing (SS2019)

High performance computing is a very important topic in the scientific area. It allows to distribute work over parallel and multiple instances to reduce computational time and utilized available resource in the most efficient way. Within the practical course on parallel computing, different techniques and methods as well as specific parallel architectures are considered. Utilizing libraries that are specially implemented for parallel task distribution, the course shows how to parallelize applications. The following concepts and programming models related to the different parallel architectures will be considered:

1. Distributed memory architectures
1. Cluster computing with Torque PBS
2. Message Passing Interface (MPI)
3. MapReduce
4. Spark
2. Shared Memory architectures
1. OpenMP
2. Pthreads
3. Heterogeneous parallelism (GPU, CUDA, etc.)
1. CUDA

The course is taught in English.

For questions and comments, please contact Johannes Erbel.