HPC Cluster
Institute of Crystallography (IC) - Bari
The Institute of Crystallography hosts a high-performance computing cluster designed to provide support to the scientific community for CryoEM, BioSAXS, X-Ray data storage and processing, as well as to enhance theoretical and computational research in Structural Biology and Drug Discovery. Specifically, the system consists of 5 CPU compute nodes and 4 CPU/GPU compute nodes, each with 4 NVIDIA H100 GPUs capable of delivering high computational power for applications such as image processing, artificial intelligence model training, and biological big data processing. Additionally, the system features a Scale-Out NAS storage system capable of supporting the storage of large amounts of data.
HPC Cluster Details
- N. 5 Compute nodes Dell PowerEdge R760: 2x Intel Xeon Gold 6448Y 2.1G, 32C/64T, 16GT/s, 60M Cache; 1024GB RAM 4800MHz DDR5
- N. 4 GPU nodes Dell PowerEdge R760XA: 4x GPU Nvidia H100 94GB NVLINK Bridge
- N. 2 Login/service nodes Dell PowerEdge R760xs
- Storage: Scale-Out NAS 4x Dell PowerScale Isilon A3000 nodes: 1280 TB raw space
Software Stack
- Compilers and libraries: C/C++, Fortran, Python, Perl, Julia, MPI, OpenMP, CUDA
- Molecular Dynamics: Gromacs, NAMD, LAMMPS
- Structural Biology: MX packages, BioSAXS software, Cryo-EM software, AlphaFold2, Imaging
- Computational Chemistry: CP2K, Nwchem, Quantum Espresso, DeepMD
- AI and Deep Learning: Tensorflow, Keras
- Statistics: R
- Drug discovery: Schrödinger Suite
- Docking and Virtual Screening
- Molecular Dynamics Simulations
- Homology Modelling
- QSAR and cheminformatics models