CC* Compute: A Balanced Cluster For Science and Engineering in the Great Lakes Region

点击翻译按钮获取中文摘要

The University of Wisconsin-Milwaukee (UWM) is building and deploying a modern, high-performance, high-throughput, and compute-on-demand cluster that balances the growing but distinct research needs for computing, storage, and memory in the diverse areas of astronomy/astrophysics, atmospheric sciences, biomedical sciences, engineering, freshwater sciences, physics, and other fields. This cluster enables UWM researchers to push the computational frontiers of their research, create cutting-edge computational tools, and scale their computations to run on national resources such as XSEDE. The system is also integrated into the Open Science Grid to enable full utilization of idle computing resources to enhance local, regional, and national scientific computational capabilities. Finally, the cluster enhances undergraduate and graduate education and research by providing superior access to large-scale computing resources for student research and computationally intensive classes.The cluster has a total system capacity of 1,568 cores, 9 TB of RAM, 504 TB of storage, and 27K GPU cores.The cluster is comprised of a login/storage server, eight compute nodes, two high-memory nodes, two GPU nodes, a 100 Gbps InfiniBand interconnect, and a 10 Gbps Ethernet network. The nodes are locally accessible to UWM researchers, available to external researchers through the Open Science Grid, and secured with the best practices in cybersecurity and monitoring. The cluster achieves a balance between large CPU counts, GPU cores, large memory, and storage to provide a flexible and responsive resource for a diverse group of researchers to perform their complex computational analyses.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.