Node-Performance Cluster Chemistry

• Subject Area: Physical Chemistry
• Term: Funded in 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 443548934

The Faculty of Chemistry and Food Chemistry at TU Dresden currently does not possess a modern high performance computing system. The Centre for information services and high performance computing (ZIH), a central institution of TU Dresden, offers a powerful high performance computer in form of the Taurus cluster, which is optimized for large-scale computations, and which already is intensively employed by the faculty.The present HPC system is unsuitable for a very large part of the calculations carried out in the field of chemistry, in particular for intensive input/output operations and for poorly parallelizing applications. In cooperation with ZIH Dresden, a conception for a meaningful extension of the present system has been developed. The cluster subject of this application is optimized for node level performance and high I/O performance, is directly connected to the existing HPC cluster and its hard- and software will be maintained in collaboration with ZIH Dresden. The joint operation with the HPC cluster of ZIH Dresden allows a wide software portfolio, file service, backup, network and technicial support at both administrator and user level. The acquisition shall secure the research of the working groups of the Faculty of Chemistry and Food Chemistry in the field of computational and theoretical chemistry for about 5 years, with focus on high throughput and high availability. A new teaching concept with a course program (including computer laboratories) in theoretical and computational chemistry supports the use of quantum chemical methods in the experimental work groups, and we expect qualified undergraduate and graduate students of the entire faculty to utilize this cluster.

• DFG Programme: Major Research Instrumentation
• Major Instrumentation: Node-Performance Cluster Chemie
• Instrumentation Group: 7030 Dedizierte, dezentrale Rechenanlagen, Prozeßrechner
• Applicant Institution: Technische Universität Dresden
• Leader: Professor Dr. Thomas Heine