Equipment Catalog

Our computing nodes are part of SDCC’s Institutional Cluster (IC). It has infinitiBand EDR connectivity and GPFS distributed storage, 200TB owned by CFN. We have shared access to 216 GPU-loaded IC nodes. Each node is equipped with two 2.10 GHz 18-core Intel processors and 256 GB of RAM, and two Nvidia K80 or P100 GPUs. CFN owns 70 such nodes, or about 7.4 million CPU hours per cycle. CFN users also have exclusive access to 110 gen04 nodes (two 2.40 GHz hex-core Intel processors and 48 GBy of RAM), providing in total 3.5 million additional CPU hours per cycle. We have diverse software tools that include deep learning frameworks, density functional theory packages for solids, surface and quantum chemistry, molecular-dynamics packages, and specialized tools for electronic excitations and optical properties.

Software for exploring properties of nanomaterials

• Deep learning frameworks (Keras, TensorFlow and Pytorch)
• Solid-state techniques based on density functional theory. (VASP, Quantum Espresso, WIEN2k, CP2K)
• Excited state techniques based on many-body perturbation theory. (Quantum Espresso, VASP, WIEN2k, local)
• Tools to simulate X-ray core level spectroscopy (FEFF, XSpectra, OCEAN and VASP)
• Quantum chemistry techniques including DFT based approaches as well as MP2, CASSCF, and coupled-cluster methods. (Gaussian 09 and 16, Q-Chem 4.2 and 5.0, ORCA, and local NWChem)
•  Classical molecular dynamics simulations with empirical potentials. (CHARMM, LAMMPS, Gromacs, local hybrid Brownian dynamics with Monte Carlo)