Acquisition of a High Performance Parallel Computing Cluster for the Departments of Biology, Chemistry, Physics, and Mathematics and Computer Science at Clark University

Clark University, Worcester MA

Investigators

Abstract

A grant has been awarded to Clark University under the direction of Dr. David Hibbett (PI) with four co-Principal Investigators (co-PIs), titled "Acquisition of a High Performance Parallel Computing Cluster for the Departments of Biology, Chemistry, Physics, and Mathematics and Computer Science of Clark University". The proposed project involves the development of a shared computational faculty for the natural science programs of Clark University. The instrumentation includes a Beowulf-type clusters, with 32 interconnected "nodes" (individual computers) and associated storage devices, hardware and software. The system will be developed over two years, with one cluster purchased in each of the first two years of the funding period. The proposed system was designed on the basis of an analysis of current and required computing capacity available to the PI and co-PIs. The PI and co-PIs all have active research programs (including graduate and undergraduate students) that require a high performance parallel computing cluster. The proposed instrumentation will be used in diverse research projects, including: 1) Evolutionary studies in fungi based on DNA sequences (Hibbett, Biology Dept.); 2) Studies of liquid-solid phase changes (co-PI H. Gould, Physics Dept.); 3) Algorithmic analyses of geometric relationships between objects (co-PI L. Han, Mathematics and Computer Science Dept.); and 4) Protein modeling and molecular dynamics (co-PIs R. Bruschweiler and S. Huo, Chemistry Dept.). These research projects are relevant to robotics, conservation of biodiversity, medicine, and other applied disciplines. In addition to research, the PI and co-PIs all participate in the Concentration in Computational Science, which provides a research-oriented background in computational applications for Clark University undergraduates majoring in diverse scientific disciplines. The PI and Co-PIs of this proposal will use the proposed instrumentation in a number of undergraduate courses that satisfy the Concentration in Computational Science, including Computer Simulation Laboratory, Molecular Evolution and Systematics, Quantum Chemistry, Biomolecular NMR, and others. Finally, the proposed cluster will be used in annual workshops on parallel computing and its applications that will be offered for area high school teachers. The development of a shared computational facility represents a significant interdepartmental initiative in the sciences at Clark. By pooling resources, the proposed project will develop critical resources for research that would be otherwise inaccessible to the individual Departments. In addition to enabling research, the proposed parallel computing cluster will become the cornerstone of the Concentration in Computational Science, which promotes interdisciplinary training for undergraduates in the sciences. The proposed instrumentation will also play a central role in summer workshops in parallel computing technology for local high school teachers. These workshops represent part of Clark University's ongoing community outreach efforts.