MRI: Acquisition of a High-Performance Computing Cluster for Occidental College

点击翻译按钮获取中文摘要

This award is made to Occidental College to purchase a high-performance computer cluster to support faculty research and teaching across the scientific disciplines. Computation has played an increasingly important role in almost all academic disciplines, and Occidental recognizes its centrality to modern research. This cluster will be used by faculty and students to advance research across biology, chemistry, computer science, physics, and economics. The cluster will support on-going research projects that include improving how computers use big data, studying chemical reactivity, and understanding heat transfer in fluids. Other faculty in math, cognitive science, sociology, and the media arts will also use the cluster for interdisciplinary research in numerical simulations and data analysis and visualization, as well as in emerging areas in the digital humanities and interactive media. Additionally, the cluster will prepare undergraduates in 20+ courses for future careers in the sciences. With Occidental's track record of attracting a gifted and diverse student body, the cluster provides an opportunity to engage a broader population in cutting-edge scientific research.This award will provide infrastructure to support the rapidly growing use and need for computational resources across the Occidental College campus. To accommodate the broad range of scientific applications of the co-PIs and the collaborators, the proposed 456-core cluster will consist of a master node, a storage node, and 17 compute nodes. The emphasis on a high number of simultaneous jobs supports the research of faculty while leaving available cores for undergraduate training and classroom use. One of the compute nodes is further equipped with two GPUs, for efficient matrix and vector operations required by many research applications. An uneven distribution of RAM across the compute nodes, ranging from 64GB to 512GB per node, allows differential use of the nodes between memory-intensive and computation-intensive users. The cluster will also include 297TB of storage, 272TB of which are located on dedicated storage nodes, to host the large datasets produced in biology, chemistry, cognitive science, computer science, physics, and economics. In addition to the physical equipment, this proposal requests funding for research software licenses as well as an administrator who will deploy, maintain, and manage the instrument. The administrator will provide technical support and create maintenance documentation and training materials, accelerating the use of the cluster by faculty and students in all areas of the natural and social sciences and the arts and humanities.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.