NDSU will acquire, deploy, and operate a new supercomputing cluster through a National Science Foundation (NSF) Major Research Instrumentation (MRI) grant. The state-of-the-art system, called “Thunder Prime,” will target research and educational activities that require processing massive amounts of data and will serve as an educational environment for the latest computational techniques and systems design.
The NSF award totals $884,596. With an additional $379,113 in cost sharing from other sources, the total value of the project will be $1,263,709. The project is expected to take three years to fully implement with initial operations targeted for June 2021.
Thunder Prime will form the updated core of NDSU’s Center for Computationally Assisted Science and Technology (CCAST). An NDSU research unit, CCAST is the foremost hub in the region for scientific and engineering endeavors and supports NDSU research and education activities while promoting technology-based economic development across areas of national and state priority including energy, transportation, precision agriculture, and biomedicine.
The new cluster will significantly improve NDSU’s capabilities to conduct leading-edge research by substantially expanding and updating the current advanced research computing resources at CCAST. Thunder Prime will have a total peak performance of at least 238 teraFLOPS (TFLOPS). For comparison, a one TFLOP computer system can do one trillion floating-point operations per second, which is the equivalent of one calculation every second for 31,688.77 years.
"This award will refresh the foundation of CCAST's infrastructure and provide NDSU a solid base upon which to build for the next five years," said Dane Skow, CCAST executive director.
While a multi-disciplinary team of 19 NDSU researchers have been involved in securing funding for Thunder Prime, the system will serve the immediate needs of many other researchers and students across nearly all colleges at NDSU and researchers from the tribal colleges in the state of North Dakota as well as at the primarily undergraduate institutions and Master’s colleges/universities within the North Dakota University System. The new cluster will also help foster collaborations outside of North Dakota which hold the promise to make significant research contributions to while further enhancing NDSU’s research portfolio.
“The new facility will significantly improve the research that has a computational component, including bioinformatics, cheminformatics, artificial intelligence, quantum mechanical and molecular dynamics simulations,” said principal investigator Bakhtiyor Rasulev. “It will put NDSU on a new level, making it well-recognized among researchers nationally and internationally.”
The principal investigators for the MRI grant are Bakhtiyor Rasulev, assistant professor of coatings and polymeric materials; Dane Skow, executive director of CCAST; Dinesh R. Katti, professor of civil and environmental engineering; Svetlana V. Kilina, associate professor of chemistry and biochemistry; and Khang Hoang, CCAST research facilitator and adjunct professor of physics. All are from NDSU.
Additional NDSU researchers involved in the project include Kalpana Katti, professor of civil and environmental engineering; Dmitri Kilin, assistant professor of chemistry and biochemistry; Ghasideh Pourhashem, assistant professor of coatings and polymeric materials; Simone Ludwig, professor of computer science; Andrei Kryjevski, assistant professor of physics; Wenjie Xia, assistant professor of civil and environmental engineering; Xin “Rex” Sun, assistant professor of agricultural and biosystem engineering; Rick Jansen, assistant professor of public health; Yildirim B. Suzen, associate professor of mechanical engineering; Shaobin Zhong, professor of plant pathology; Nonoy Bandillo, assistant professor of plant sciences; Trung “Tim” Le, assistant professor of industrial and manufacturing engineering; Trung B. Le, assistant professor of civil and environmental engineering; and Limin Zhang, associate professor of accounting and information systems.