Scientific Computing Research Environments for the Mathematical Sciences

College Of William And Mary, Williamsburg VA

Investigators

Abstract

Under the SCREMS program, the Department of Mathematics at the College of William & Mary will purchase equipment for the creation of a computing network dedicated to the support of research in the mathematical sciences. The award will provide funds for high-bandwidth networking equipment, high-performance computers for numerical calculations, and a file server. This equipment, together with desktop machines providing platforms for smaller computations, will provide a numerical/file server cluster for the faculty investigators and students working on a number of research projects. This equipment will be used for computationally intensive research projects spanning a wide range of applied mathematics. These include numerical methods for engineering design, numerical mathematics, computational biochemistry, and operations research. The projects specifically included in the grant are: (1) Efficient approaches to the optimization of systems governed by partial differential equations, (2) distance parameterized methods for molecular conformation, (3) a parallel Hessenberg QR eigenvalue algorithm, and (4) the calculation of the exact distribution of the time to traverse a stochastic activity network and the elimination of resampling error in bootstrapping. PI Leemis and two former doctoral students have written APPL (A Probability Programming Language). This Maple-based language is used to compute probabilities and perform operations on random variables. The size of many problems that can be solved by APPL is limited by computing speed. The speed offered by the hardware obtained as part of the SCREMS grant is crucial to solving problems of the magnitude encountered in practice.