Montana State awarded grant to boost high-performance computing capacity

Montana State University, the state's largest research entity and a designated Carnegie R1 institution, has received a $500,000 grant from the National Science Foundation to significantly expand its computational infrastructure for research.

The grant was awarded to MSU's Research Cyberinfrastructure Core Facility, which manages the advanced information technology systems that support large research objectives like quantum science, large data analysis, machine learning and artificial intelligence. After the new equipment is installed and operational, MSU will have the most powerful high-performance computing capability in the state, according to Coltran Hophan-Nichols, the RCI facility's director.

The upgrade will roughly double current computation capacity and increase computing performance for some workloads tenfold, he said.

"These are large growth areas for the university, and emerging areas of research are particularly demanding," Hophan-Nichols said. "There is existing demand now, and we expect demands and workloads to continue increasing in the future."

During this first week of the spring semester, the new equipment acquired under the grant will be installed and operational, Hophan-Nichols said. Users will notice the change immediately – for example, some work that currently requires two weeks of computer run time will be completed in one to two days, he said. The upgraded system also will generate higher-fidelity models than those that can currently be produced.

"Right now, many researchers have to limit the scope of what they are analyzing, but with these new resources, that won't be the case nearly as often," Hophan-Nichols said.

The enhancements are slated for RCI's Tempest supercomputing system. The system's existing central processing units, or CPUs, will be augmented by new processors, which are twice as fast. Each new CPU-based system will be equipped with 1.5 terabytes of memory.

The GPUs, or graphics processing units, can handle multiple tasks and complex calculations. Their current hardware will be augmented by NVIDIA H100 GPUs that are 10 times more powerful than Tempest's current GPUs.

Hophan-Nichols said the upgrade is necessary to support the university's expanding research portfolio and to train students to use computing tools in research and other practical applications, skills they may need in their careers after college.

For example, engineering students will use the system as they learn parallel processing - running two or more processors simultaneously to handle separate parts of computation tasks. Earth sciences students will use the system in lab settings to learn how to manage and analyze geographic information systems data. Chemistry students will use it to practice computational chemistry, which employs computer simulation to help solve complex chemical problems and even predict the interactions and properties of molecules outside a physical lab.

"There are a lot of different applications for this – emerging quantum science, machine learning, data processing, precision agriculture, biology and genomics, to name a few," Hophan-Nichols said. Before submitting the NSF grant application, he added, his team collaborated with researchers across campus to understand their needs.

"We looked at current and emerging areas where we had a gap and looked at the best equipment to fill that need," he said.

Hophan-Nichols said the NSF technology grant is the first IT infrastructure grant MSU has received in eight years and is by far the largest cyberinfrastructure grant in the university's history.

Ryan Knutson, vice president of information technology, said the grant is a major boost to the MSU's mission of providing computational infrastructure to its researchers.

"This cutting-edge computing infrastructure will help to accelerate research," he said.