MRI: Acquisition of a High-Performance Computing Cluster for Interdisciplinary Research and Teaching

Dordt University, Incorporated, Sioux Center IA

Investigators

Abstract

The acquisition of a high-performance computing (HPC) cluster for use by faculty and undergraduate students at rurally located Dordt University will significantly impact several new and ongoing, multidisciplinary research projects in both STEM and non-STEM fields. This cluster will generate opportunities for Dordt faculty and other regional college and high school faculty to integrate high-performance computing into their research, training activities, and classrooms. Research projects covering a wide range of academic disciplines will be supported by the new HPC cluster, including computational biomedicine, medicinal chemistry, statistical genetics, planetary chemistry, and structural integrity under seismic loading. The project will also expand opportunities for HPC use within other fields such as social sciences statistical research, digital humanities, institutional research, and IT. The research endeavors of this team of undergraduate research mentors will be substantially catalyzed by the HPC cluster, leading to a diverse and interdisciplinary set of high-impact research findings, providing cutting-edge, high performance computing research experiences for undergraduate students, and acting as a model for cross disciplinary computational resource sharing at Primarily Undergraduate Institutions. Other faculty in computer science, data science, statistics, engineering, and chemistry will also utilize the cluster in upper-level courses as research training opportunities for students. This Major Research Instrumentation award will provide infrastructure that will enable faculty to further develop and sustain research programs that already produce skillfully trained undergraduate researchers, generate peer reviewed publications, and attract external support. To accommodate the multidisciplinary range of the research applications of HPC users, this project will provide a high performance computing cluster consisting of a head node and 18 compute nodes in three standardized configurations: eight regular compute nodes, eight high RAM nodes, and two Graphics Processing Unit (GPU)-based nodes for applications, such as large molecular modeling systems, that will harness the massively parallel vector processing abilities of GPUs. Over the three years of the project, approximately fifty graduate school-bound undergraduate students (over half female; at least 10% minority) will receive intensive research training on the proposed HPC cluster by the primary faculty users. Furthermore, at least 1500 additional undergraduate, middle and high school students (many from rural, first generation college, or lower socio-economic strata), and faculty will use the cluster as part of STEM courses and outreach activities, including high school teacher training and faculty development workshops, as well as research/education programs for middle and high school students. The HPC resources will also be integral to a new initiative at Dordt University focused on recruitment and retention of academically talented students with financial barriers to education. This award will bring additional high-performance computing capability to a region of the country which currently has limited access to HPC resources, but has seen recent rapid growth in STEM. The HPC cluster will reach a diverse user community and maximize student impact, inspiring more students to pursue STEM careers and use computational tools across disciplines. 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.