Studierende	Beschäftigte	Einrichtungen	stud. Gruppen

For very computationally and memory intensive calculations exceeding the locally available resources, additional computer systems are available within the NHR Alliance, in particular at the former HLRN centers in Berlin (NHR@ZIB) and Göttingen (NHR@Göttingen). We would be happy to advise you on the decision of suitable hardware systems and support you in applying for an account.

NHR-System (ZIB), Foto: ITMZ | Universität Rostock.

NHR@ZIB: System configuration

• Intel CPU partition
  • 1270 nodes (121920 cores)
    • each equipped with 2 Intel Cascade Lake CPUs
      • 48 cores per CPU
    • 1236 nodes with 384GB RAM
    • 32 nodes with 768GB RAM
    • 2 nodes with 1.5TB RAM
• Nvidia GPU cluster
  • 42 GPU nodes each equipped with
    • 4 Nvidia A100 (80GB) cards
    • 2 Intel Xeon 8360y CPUs

NHR-System at GWDG

NHR@Göttingen: System configuration

• Intel CPU partition
  • 974 nodes (93504 cores)
    • each equipped with 2 Intel Cascade Lake CPUs
      • 48 cores per CPU
    • 956 nodes with 384GB RAM
    • 16 nodes with 768GB RAM
    • 2 nodes with 1,5TB RAM
  • 448 nodes (17920 cores)
    • each equipped with 2 Intel Skylake CPUs
      • 20 cores per CPU
    • 432 nodes with 192GB RAM
    • 16 nodes with 768GB RAM
• Nvidia GPU cluster
  • 33 GPU nodes with
    • 4 Nvidia A100 (40GB) cards
    • 2 AMD Epyc 7513 CPUs

Further information:

• For NHR@ZIB and NHR@Göttingen:
  •  Applying for an account
  •  User documentation
•  Workshops and courses within the NHR Alliance
•  NHR news

Contact persons at Kiel University

If you are interested in using the NHR systems, the following employees of the Computing Centre will provide you with further information:

• Dr. Karsten Balzer, balzer@rz.uni-kiel.de, Telefon: 880-4659
• Dr. Simone Knief, knief@rz.uni-kiel.de, Telefon: 880-2766