Equipment: MRI Track-I: Acquisition of CyBR: Cyber Infrastructure for Big Data Research Critical for Alaska

This project funds the procurement of modern Graphics Processing Unit (GPU) based compute nodes and sufficient high-performance storage at University of Alaska Fairbanks (UAF) that will support a new range of data and compute-intensive applications to allow new science and engineering discoveries. This system, a Cyberinfrastructure for Big data Research (CyBR), will enable artificial intelligence, machine learning, deep learning, image analysis, statistical modeling, and large-scale simulations at UAF.

The instrument will be a catalyst for increasing high priority, data-driven research in multiple science and engineering disciplines critical to Alaska, its climate and its natural resources. These include Alaska’s distinctive flora and fauna and its sociocultural and economic mobility. At least 20 different research projects will benefit from the instrumentation.

Research areas include Arctic marine biology and implications for Indigenous communities, Alaska coastal modeling, boreal wildfire management, Alaska earthquake detection, Alaska mining safety, wildlife and ecology, Arctic Ocean modeling and space science. NSF Alaska EPSCoR recognizes many of these as high-priority research areas. Other research areas include, glaciology, volcanology, water treatment, critical mineral extraction and discovery of low carbon construction material in Alaska.

All of these will potentially support new funding opportunities. This project will also create training opportunities for first-generation college students, female students, and those from marginalized communities, including Alaska Native communities. CyBR will aid in diversifying the data analysis and computationally intensive workforce, which is invaluable not only in Alaska but also in the nation where data analysis jobs are expanding at an unprecedented rate.

The Cyberinfrastructure for Big data Research high-performance computing instrument will consist of 10 compute nodes, each with AMD CPUs, NVIDIA GPUs, at least 256GB RAM, and NVMe SSDs. This hardware will be accompanied by a large Lustre storage array to accommodate big data and an HDR Infiniband switch for high-speed, low-latency data transfer between nodes and the storage so that computational performance is not bottlenecked by network bandwidth.

The research team will collaborate with the UAF Geophysical Institute’s Research Computing Systems group to purchase, install and maintain the instrument. It will be located in the Butrovich Data Center, where it will be connected to UAF’s existing Chinook HPC cluster. Integrating the new CyBR instrument with the existing HPC cluster will reduce the time to make this new instrument available to researchers, reduce the infrastructure needed to operate it and reduce the management overhead to maintain it.

This project is jointly funded by the Major Research Instrumentation (MRI) program, the Established Program to Stimulate Competitive Research (EPSCoR), the Information Technology Research (ITR) program, and the Office of Advanced Cyberinfrastructure (OAC).

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.