2020-04-15 CANCELLED: Introduction to the LRZ HPC Infrastructure (HIHI1S20)

In this introductory course we will give an overview of the High Performance Computing (HPC) systems operated by the Leibniz Supercomputing Centre (LRZ).

First, a general introduction to HPC systems as well as a brief discussion of historical developments and current trends will be given. Then, the different systems at LRZ will be presented in detail. While touching upon the world leading supercomputer SuperMUC-NG, the focus will be directed at different cluster and storage systems as part of the LRZ Linux Cluster. These systems will be further explored in a dedicated hands-on session. This will prepare participants to succesfully run compute jobs on LRZ HPC systems and will cover important system components like the environment module system and the Slurm Workload Manager. Additonal supporting software services will also be presented.

The material will be presented as a combination of lectures, demos and hands-on sessions, with a focus on the latter. There will be breaks during the session.

Participants will gain the general understanding and skills necessary to efficiently utilize the LRZ supercomputing infrastructure for their own projects.

Schedule (tentative):

• Introduction to the LRZ (B)
• Developments in high performance computing (Moore's law, parallel processing, LRZ systems) (B)
• Basics of high performance computing (system characteristics and cluster topology, levels of parallelism) (B)
• SuperMUC-NG: system overview (B)
• Overview of HPC systems for Bavarian universities: Linux Cluster, DGX-1 and GPU systems( and mentioning the Compute Cloud) (B)
• Criteria and process for system access (B)

15 minute break

• Data storage options (B)
• Supporting software services/collaboration tools (B)
• RStudio Server as web-based frontend to the Linux Cluster (B)
• Hands-on Linux Cluster: connecting and (file) system overview (B)

15 minute break

• Environment modules (module list, show, available, load, unload) (B)
• Basics of the Slurm Workload Manager and interactive usage (sinfo, salloc, srun) (B)
• Job scripts and processing (sbatch) (B)/(I)
• Job queue and accounting (squeue, sacct) (B)/(I)

• Introduction to GNU/Linux and SSH (or comparable previous knowledge)

All participants are expected to bring their own laptops.