RUI: A High Performance Computing Environment at Smith College

Smith College, Northampton MA

Investigators

Abstract

Mathematics is changing due to the advent of high-powered symbolic computation. This grant will fund a high performance computational cluster to serve the research needs of a variety of faculty at Smith College. A project in symbolic computation and algebra will use high powered computation to prove conjectures about exceptional Weyl Groups which require significant computation as well as working toward developing a new specialized computer algebra package for Weyl Groups and symmetric spaces. A researcher in computational geometry will use the cluster to search for special configurations of protein backbone polygonal chains, and to generate random geometric objects, in both R2 and R3. A graph theorist will use parallel computing to determine algebraic, geometric, and combinatorial parameters of graphs in order to find examples that show that particular bounds are sharp. A project in dynamical systems and mathematical biology involves intricate image processing for data collecting, computationally costly exploration of high dimensional dynamical systems, as well as optimization of the fit of these models to large data sets. A statistician will use high powered computation to develop likelihood methods for incomplete data in regression models. A project in computational geometry, rigidity and motion simulation will take advantage of parallel processing in an interdisciplinary project with bio-physicists and computational chemists for experimenting with and developing new tools for molecular simulations. This grant will provide computing power for research projects that span a broad spectrum of mathematics including graph theory, statistics, computational geometry, and symbolic computation. They are both theoretical and applied. Several of the projects involve interdisciplinary work with biologist, physicists, chemists and computer scientists. The possibility of using high-powered computer systems allows us to ask different questions and devise different techniques, thus inherently changing the course of our research. Improving the computational capabilities at Smith College creates a better research environment for the faculty. This is a College known for its ability to engage undergraduate women in research and to send them to graduate school in the mathematical sciences. Thus this grant will enable projects that are sophisticated, highly computational and yet accessible for student work. This will enable Smith undergraduate students (all women), and students at the new Center for Women in Mathematics at Smith to participate in a significant way in a serious research projects in collaboration with Smith faculty.