CC* Compute: Shared Computing Infrastructure for Large-scale Science Problems

点击翻译按钮获取中文摘要

Advances in scientific instrumentation over the past decade have led to phenomenal growth in both the size and complexity of data across a wide variety of scientific domains. Researchers at the University of Connecticut are expanding the capabilities of the existing shared high-performance computer facility to meet this challenge by the addition of a 28-node cluster, each node with 40 Intel cores and 192 GB of memory, and 1000 TB of shared storage.This balance of resources has been chosen to accommodate applications in the areas of particle physics, astrophysics, geophysics, and statistics in use by UConn researchers, with a view to meet a rising need for data-intensive computing across all of science and engineering at the University. Furthermore, the cluster will be configured as a shared scientific computing resource on the Open Science Grid (OSG). OSG is a consortium of universities and national research facilities who are using grid technology to aggregate their individual computer clusters into a single unified national compute infrastructure for science. In joining this cluster to the OSG, UConn researchers will see a much higher throughput than they would see running on local resources alone, granting them faster turn-around and access to bigger data than could be stored and processed locally. The cluster will enable the introduction of a new big data component within an existing course on scientific computing at the graduate level. It will also be used to produce visualizations of astronomical observations that will be used in K-12 outreach.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.