Seminar on (Performance) Portable Programming of HPC Applications(IN2107)
Prof. Dr. Erwin Laure (TUM and MPDCF)
|Dates:||Kick-off: 26.04.22; workshop: 5.7.22|
2.2.2022, 10:00-12:00, Zoom
|First meeting:||26.04.22, zoom: https://mpcdf-mpg-de.zoom.us/j/97256746786?pwd=Y2wvN3hyT3RCMXR3dzNlWDRKSURIQT09|
Developing applications for massively parallel High Performance Computing (HPC) systems requires specific programming models and tools to express parallelism and data access/movement. For many years, the predominantly used models have been the Message Passing Interface (MPI) and OpenMP. However, with the increasing complexity of HPC hardware, involving vector units, GPUs, and novel memory and storage technologies, new programming models and tools have emerged to efficiently program these components. These new models and tools are however often not easily portable (or even proprietary) and achieving consistent performance across systems is becoming a much more difficult challenge.
The topics that will be covered in this instance of the seminar are:
- Programming for GPUs (CUDA, HIP, OpenMP, OpenACC, …)
- Performance portability frameworks (Kokkos, Raja, …)
- Frameworks for artificial intelligence (AI) applications
- Abstractions and tools for memory or IO-management
The students will select material for the presentation, provide a report, and give a presentation.