HPC Courses Training Organization and Experiences in Supercomputing Luxembourg EuroCC: National Competence Centre (NCC)

Volume 15, Issue 2 (November 2024), pp. 16–23

Download PDF

@article{jocse-15-2-4,
  author={Ezhilmathi Krishnasamy and Pascal Bouvry},
  title={HPC Courses Training Organization and Experiences in Supercomputing Luxembourg EuroCC: National Competence Centre (NCC)},
  journal={The Journal of Computational Science Education},
  year=2024,
  month=nov,
  volume=15,
  issue=2,
  pages={16--23},
  doi={https://doi.org/10.22369/issn.2153-4136/15/2/4}
}

High performance computing (HPC) is a crucial field in science and engineering. Although HPC is often viewed as a pure field of computer science or a subset of it, it actually serves as a tool that enables us to achieve exceptional results in science and engineering. Since early on, computers have been primarily utilized for extensive arithmetic computations. However, recent advancements in electronics have also made edge computing integral to high performance computing. Additionally, we have witnessed remarkable growth in computer architecture, leading to the development of powerful HPC machines, with supercomputers now reaching exaflop powers. Nevertheless, there are still challenges in utilizing these powerful machines due to the lack of knowledge in integrating physics and mathematics into HPC. Furthermore, complications with the software stack and common parallel programming models that target exascale computing (heterogeneous computing) persist. In this context, we present our effective course design for HPC training, focusing on CUDA, OpenACC, and OpenMP courses, which aim to equip STEM graduates with HPC knowledge. We also discuss how our training stands out in comparison to other NCC training frameworks in the EuroCC context and promotes lifelong learning.