MRI: Acquisition of High Performance Scientific Computing Cluster at Trinity University

点击翻译按钮获取中文摘要

This instrument supports large-scale, high-performance computing (HPC) at Trinity University, which enables a broad range of scientific research efforts spanning mathematics, computer science, chemistry, biology, and physics. Many ongoing projects at Trinity relate to questions of 'big data' analysis, including such diverse areas as weak interactions in host-guest chemistry, multi-scale simulations of protein interactions, models of fluid dynamics related to cellular motility in membranes, analysis of high-throughput genomic sequencing data, multi-agent modeling of security systems, and simulations of particle collisions and gravity within the rings of Saturn. Each of these research areas requires considerable computational power to analyze extremely large datasets, whether generated through traditional laboratory experiments or through robust computational simulations. The HPC cluster provides a shared computational resource to all researchers at Trinity, and includes both traditional central processing units (CPUs) and newer general-purpose graphics processor unit (GPU) systems optimized for large quantities of floating-point numeric operations. The cluster operates on the Linux operating system, with a hybrid queuing/priority system so that computational jobs submitted by any researcher are quickly and efficiently executed. This resource enables Trinity University, an undergraduate institution, to more effectively train and educate students on tangible aspects of large-scale computational projects and research endeavors, such as the value of test-driven development, efficient algorithm design, memory management, and data visualization, while still tackling the challenges of open research questions on a modern scale.