UW Wins $4M to Install State-of-the-Art Research Computing System

The University of Wyoming’s ability to use artificial intelligence (AI) and high-performance computing to solve problems in areas including the environment, energy, agriculture and public health will take a major step forward through a nearly $4 million grant to acquire state-of-the-art computing infrastructure from the National Science Foundation (NSF).

The three-year, $3.9 million award includes $3.25 million for UW to acquire a specialized high-performance computing testbed composed of 24 nodes of NVIDIA Grace Hopper Superchips with 400 terabytes of data storage, cutting-edge technology that currently is not available in the Rocky Mountain region. An additional $340,000 will be used for system support and student training.

The award for the “Acquisition of Advanced Infrastructure to Accelerate Impact of AI Through Applications and Innovation for Wyoming (AI4WY)” project was announced this week as part of NSF’s Major Research Infrastructure Track 2 Program. UW will have access to 75 percent of the new computing system’s capacity, with external partners Colorado State University (CSU) receiving a 15 percent allocation and the Rocky Mountain Advanced Computing Consortium -- composed of 33 institutions in the West, including UW -- receiving 10 percent.

“This award from a highly competitive NSF program is a huge step forward for UW, as it will significantly advance the application and innovation of artificial intelligence and computational science across the university and the Rocky Mountain region,” UW President Ed Seidel says. “AI4WY will amplify and build on our recent investments to create our new School of Computing; our existing Advanced Research Computing Center (ARCC); and our partnership with the National Center for Atmospheric Research-Wyoming Supercomputing Center in Cheyenne.”

The project’s principal investigator is Andrew Kirby, an associate research scientist in UW’s School of Computing. Co-principal investigators are Gabrielle Allen, director of the School of Computing; Suresh Muknahallipatna, a UW professor of electrical engineering and computer science, and faculty director for UW computing resources; Michael Killean, deputy director of ARCC; and Michael Kirby, director of CSU’s Data Science Research Institute. Multiple UW academic departments are involved, ranging from the departments of Mechanical Engineering and Atmospheric Science to the departments of Anthropology and History.

“Every scientific field and industry have been upended by the emergence of AI, and this award will allow UW researchers to leverage this amazing computing instrument to find new approaches to tackle Wyoming’s challenges in a range of fields, including ecological systems, climate, public health, tourism, archaeology, digital humanities and energy from wind, nuclear and petroleum resources,” UW’s Kirby says. “It also will allow us to accelerate the research and training collaborations that are underway to focus on interstate research problems in our region.”

The new high-performance computing system will address two significant computing challenges: data movement and energy consumption. It will allow for the realization of what is called “digital twins” -- a virtual representation of a system characterized by real-time monitoring, simulation and prediction; and adaptive feedback for design and development.

The computing hardware needed for these coupled big-data and modeling algorithms requires state-of-the-art, tightly coupled central processing and graphics processing units with abundant onboard memory, as these problems require high-speed data transmission between processing units. The NVIDIA Grace Hopper Superchip is the only commercially available hardware fitting these specifications.

AI4WY specifically highlights four broad areas of research important to Wyoming and the region that present computationally challenging and data-intensive problems: environment, agriculture, energy and society.

“Overall, this platform will provide the foundational cyberinfrastructure for users in all fields to develop and explore transformative AI solutions; it will provide the scaffolding for new collaborations with partners at CSU and across the Rocky Mountain region; and it provides a bridge to national exascale AI resources,” Allen says. “This will be a tremendous asset for Wyoming and its university, and the School of Computing is excited to be part of building associated AI research and learning capacity, including the new master’s degree in AI that is now available to UW students.”