Last update: 20230616

Hardware

Example comparisons

sciCORE https://scicore.ch/using-scicore/hpc-resources/

Cluster BioMedIT

• Total nodes 15
• Total cores 976 total, 200 user
• Total RAM 6 TB
• Total GPUs 8
• Inter-connect Eth 100G
• Total Disk 900 TB

Cluster sciCORE

• Total nodes 215
• Total cores 13632
• Total RAM 73 TB
• Total GPUs 80
• Inter-connect Eth 100G, Infiniband
• Total Disk 11 PB

EPFL https://www.epfl.ch/research/facilities/scitas/jed/

• Peak performance Rpeak: 2’322 TFLOPs
• Total RAM: 233,5 TB
• Storage: 350 TB
• The cluster is composed of 419 compute nodes, each with
  • 2 Intel(R) Xeon(R) Platinum 8360Y processors running at 2.4 GHz, with 36 cores each (72 cores per machine),
  • 3 TB of SSD disk
• for a total of 30’240 cores (including the frontend node)
• 375 nodes have 512 GB of RAM,
• 42 nodes have 1 TB of RAM,
• 2 nodes have 2 TB of RAM
  • approx 250 TB total

HPC documentation

sciCORE uses SLURM workload manager https://slurm.schedmd.com/overview.html

Examples of documentation on simimlar infrastructure.

• NIH Biowulf https://hpc.nih.gov
• For the sciCORE cluster: https://wiki.biozentrum.unibas.ch/display/scicore/sciCORE+user+guide
• For using SLURM: https://wiki.biozentrum.unibas.ch/display/scicore/SLURM+user+guide
• EPFL SCITAS: https://www.epfl.ch/research/facilities/scitas/documentation/

Acknowledgements

• sciCORE: “Calculations were performed at sciCORE (http://scicore.unibas.ch/) scientific computing core facility at University of Basel.”
• sciCORE/SIB: “Calculations were performed at sciCORE (http://scicore.unibas.ch/) scientific computing core facility at University of Basel, with support by the SIB Swiss Institute of Bioinformatics.”

Data stream

Read about where the data is generated, how it comes fro BioMedIT and how the responsibility of management is controlled here: Data stream