About the Linux clusters hosted by the MPCDF

Dedicated Linux compute clusters are operated for more than 25 Max Planck Institutes and working groups. In total this comprises more than 100.000 CPU cores, hundreds of GPUs and more than 50 PB of (net) online storage (as of 2020). The MPCDF supports the Max Planck Institutes during the entire life cycle, starting from the sizing and configuration of the resource, Max-Planck internal procedure ("BAR" proposal), benchmarks, procurement process, to the deployment, operation and optional renewal.

In general, all Linux clusters feature the following hardware structure and software stack:

• SUSE SLES / CentOS operating system
• GPFS as cluster wide parallel file system
• Slurm batch system
• module environment for software deployment
• Intel MPI, Intel compilers and gcc

In order to facilitate migration between the central HPC systems and the dedicated Linux clusters the software stack of the dedicated clusters (including compilers, MPI, numerical libraries) is kept as similar as possible to the HPC systems.

Contact Ch. Guggenberger if you are considering purchasing a compute cluster for your Max Planck Institute. We are happy to support you each step of the way - from requirements gathering through procurement, set-up and operations.