The course is experimental and hands-on, requiring access to advanced GPU compute capacity. All practical work will be conducted on the Norwegian national HPC system Olivia, where course participants will gain user access for the period August–December 2026.
Instructions on how to gain access to Olivia will be given in the kick-off lecture on Monday, August 17. Training in using a supercomputer effectively, including its job scheduling system SLURM, storage management, multi-GPU and multi-node computations, as well as budgeting and accounting of available compute, will be provided through weekly hands-on laboratories and via the GitHub repository for the course.