- Hosting Legal Entity
- Vilnius University
- Coordinating Country
- Type Of RI
- Life Cycle Status
- Operational since 2011
- Sauletekio 9, Vilnius University/Center of Applications of Informational Technologies Vilnius University/Mathematics and Informatics Faculty, Vilnius, PO: 10222 (Lithuania), Lithuania(55.3499984741211, 23.75)
- RI Keywords
- HPC,Grid computing,Parallel computing,Material sciences,Quantum chemistry,Hydrometereology
- RI Category
- Distributed Computing Facilities;Biomedical Imaging Facilities;Data Archives, Data Repositories and Collections;Software Service Facilities
- Scientific Domain
- Information Science and Technology;Physics, Astronomy, Astrophysics and Mathematics;Chemistry and Material Sciences;Biological and Medical Sciences
Mission and Objective
National Grid Infrastructure is a part of the research infrastructure for Physical and Technological Sciences of Lithuania which is owned by Vilnius University and funded by European Structural Funds.LitGrid-HPC is based on an existing HPC infrastructure at Vilnius University and on the network of distributed and parallel computing and e-services (LitGrid). It enables comprehensive scientific data maintenance. The RI is meant to create, maintain and develop an e-infrastructure (grid, cloud computing, high-performance computing, virtual repositories, related data sets), to serve the academic environment, public sector, business needs, and requests from foreign partners. It also provides calculation resources, modeling services, virtual repositories, data sets storage and usage and provide an “on demand” supply: computing, data repositories, related e-services. LitGrid-HPC deploys technological innovations or use new technologies in HPC, grid and cloud computing, especially via participation in the European Grid Initiative (EGI) and HPC activities. Supercomputing resources are open to any user according to the rules of public access centers of Vilnius University.
OpenMP, MPI and GPU computations in linux environment. Quantum chemistry and material science computations by the following software: Gaussian09, gamess-US, Qchem, Molcas, Molpro, Amber12, Vasp, Matlab, according to existing licenses.
Five distributed computing clusters from 192 to 960 computing cores each and from 2 to 4 GB RAM per core (total 3000 computing cores). distributed memory cluster with GPU ( 3 computing nodes with total 6 GPU cards).shared memory node with 120 computing cores and 2,5 TB RAM.