Introduction to GPU Computing on Raven
- Date: Jul 21, 2021
- Time: 02:00 PM - 04:00 PM (Local Time Germany)
- Speaker: Members of the MPCDF Application group
- Location: online via Zoom
- Host: MPCDF
- Contact: training@mpcdf.mpg.de
The new supercomputer of the Max Planck Society, Raven, is fully operational and comprises a total of 1592 CPU compute nodes with 72 Intel Xeon IceLake-SP cores each. A set of 192 nodes is equipped with 768 NVIDIA A100 GPUs (with 4 GPUs per node) which deliver a major fraction of the system's total compute power, ranking the GPU partition 47th on the June 2021 editions of the TOP 500 list of fastest supercomputers, and 12th on Green 500 list of the most energy-efficient machines.
This online tutorial gives an introduction specifically to GPU computing on the Raven system, aimed at HPC users interested in leveraging the powerful GPU resources. The following topics will be covered:
- hardware configuration
- GPU software environment for HPC and AI/data analytics: applications, libraries, compilers, tools
- batch system: submitting GPU jobs
Follow-up courses in September (GPU bootcamp Nvidia/MPCDF) and November (Advanced HPC workshop at MPCDF) will focus on GPU application development.
The tutorial is given by application experts of the MPCDF. There is the possibility to ask questions via the chat, and during a dedicated Q&A session after the presentation.
see also: Raven user guide