High Performance Computing Cluster

The High Performance Computing (HPC) Cluster HYDRA is used for processing time intense tasks like molecular dynamics (MD) simulations, training of machine learning models, generation and virtual screening of large molecule databases, quantum chemical calculations, .…

The cluster has been built solely from commodity-hardware components and gets continuously updated with newer, more powerful components to stay ready for future demands dictated by our research in cheminformatics and computer-aided drug design.

Key Facts:

• 8 Nodes each with 16 CPU-cores, 1 GPU and 64 GB RAM, 4 TB hard disk
• 8 Nodes each with 64 CPU-cores, 1 GPU and 256 GB RAM, 4 TB hard disk 640 CPU-Cores and 16 GPUs in total
• Nodes are connected via a 10 GbE-switch to the cluster-head and outside network
• Operating system: CentOS 7 Linux
• Job and computing resource management performed by the Sun Grid Engine (SGE)
• Located in Room 2Z135 (maintained by the ZID; equipped with racks, air condition, UPS and 10GbE-connectivity)
• Available MD-software: CHARMM, GROMACS, AMBER16 (all GPU-accellerated)