The High-Performance Computing Center Stuttgart (HLRS) will host a two-day online introductory course on OpenMP offloading with AMD GPUs from 21 to 22 October 2025.
OpenMP is a key option for speeding up or offloading computations on modern heterogeneous computer systems with GPUs. This course introduces AMD Instinct GPU and Accelerated Processing Unit (APU) architectures, explaining how GPUs operate and can be utilized for offloading in OpenMP. It also highlights new features in recent OpenMP versions and GPUs, such as the unified memory programming model, which simplifies writing HPC applications across various GPU programming models. Additionally, tools for performance analysis and optimization are covered.
This course is aimed at beginners in GPU programming who have a basic understanding of parallelization with OpenMP and/or MPI on CPUs. After completing this course, you will understand the fundamentals and feel confident in porting your applications from a CPU-only system to systems with discrete GPU accelerators or APUs.
