MRI: Acquisition of an HPC System for Data-Driven Discovery in Computational Astrophysics, Biology, Chemistry, and Materials Science

点击翻译按钮获取中文摘要

The project funds the purchase of a high-performance computing and storage system at the Georgia Institute of Technology. This computing instrument will support data-driven research in astrophysics, biosciences, computational chemistry, materials and manufacturing, and computational science. These projects contribute to national initiatives in big data, strategic computing, materials genome, and manufacturing partnership; and NSF supported observatories such as the gravitational wave observatory and the South Pole neutrino observatory. The system also serves as a springboard for developments of codes, software prototyping, and scalability studies prior to using national supercomputers. Advances made in computational methods and scientific software are disseminated in the form of open-source codes and data analysis portals. Over 33 faculty, 54 research scientists/postdocs, 195 graduate students, and 56 undergraduate students will immediately benefit from the instrument. In addition, the system provides training opportunity at all levels from undergraduate students to early career researchers, in important interdisciplinary areas of national need. A fifth of the system capacity is utilized to enable research activities of regional partners, researchers from minority serving institutions, and other users nationally through XSEDE participation. The project involves undergraduate student participation from historically black colleges from Atlanta metropolitan area. Public outreach efforts are planned through videos of public interest and local events such as the Atlanta Science Festival.The cluster will combine regular compute nodes with others configured to emphasize one of the following: big memory, big local storage, solid state storage, Graphics Processing Units (GPU), and ARM processors. In doing so, the system can be employed by a diversity of projects. In astrophysics, the instrument bolsters data-driven research including detection of gravitational waves, astrophysical neutrinos, and gamma rays. It does it by leveraging data from leading astroparticle observatories and contributing to their mission. It also leads to improved insights into formation of supermassive black holes and large-scale structure of the universe. The computing system also aids the development of parallel software in computational genomics, systems biology, and health analytics. Important applications in assembly and network analysis of plant genomes, and environmental metagenomics are pursued. The instrument also enables next generation algorithms and software for computational chemistry and expands the boundaries of molecular simulation. The system enables advances in density function theory, enhances studies of crystal defects and nanostructures, and injects novel use of machine learning techniques in computational chemistry. It also fosters the development of data science methodologies to identify building blocks of materials at multiple scales, thus significantly reducing the development and deployments cycles for new materials.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.