HPC@HU – Cluster

The Computer- und Medienservice at HU Berlin provides a powerful high-performance computing platform called HPC@HU. This infrastructure supports computationally intensive tasks in the areas of simulation, data mining, machine learning, artificial intelligence, and other application areas that work with large datasets and complex models. Researchers and lecturers at the university can use HPC@HU to perform demanding calculations and analyses.

HPC@HU – Cluster

• 35 servers
  • 2 x 64 cores per server
  • 30 x 1TB RAM
  • 5 x 4TB RAM
  • 2 x 7,6TB local SSD
• Lustre Cluster with 1,6PB
• 100 GBit/s – Ethernet
• OpenStack & Azimuth
  • Different operating systems and platforms (Kubernetes, Slurm, Workstation, …)
• Upcoming: Resource management with Blazar

Further information: