During 2015 — 2016 we have established a modern set of computing facilities for comprehensive daily work with challenging computing tasks.
By now our facilities include access to central Skoltech IBM computing cluster, MSU Lomonosov supercomputer and we also have own mini-cluster shared between CEE and CDISE CREIs. Mini-cluster consists of 14 computing nodes with 16 CPU cores and 128 Gb memory at each node.
First 7 computing nodes belong exclusively to CEE CREI and other 7 to group of prof. A. Shapeev (CDISE). Each node in the system has shown 0.5 TFlops performance under LinPack test. Cluster is configured under CentOS operating system with SLURM scheduler and XFS file-system. Current state of the system and its power consumption basements allow to enlarge it twice, up to 28 computing nodes without serious investments to infrastructure.
At CEE-owned part of the cluster we have optimized installations of important computing packages (VASP, Gaussian, LAMMPS) with implementations of standard submission scripts. Sergey Matveev provides full administrative support of the cluster at software level and gets a support at hardware level from the IT-department.
Also, we have established a basic data storage server allowing members of our team to update/synchronize their local data-bases on different platforms daily in a safe way.
Please, contact with Sergey Matveev (s.matveev@skoltech.ru) if you need more information about our computing facilities.