Cray Ookami Drives Testbed for Computational Studies at Stony Brook
August 20, 2019
A $6.3 million anonymous donation and a $5 million NSF grant promise to accelerate cutting-edge research in the field of high-performance computing at Stony Brook’s Institute for Advanced Computational Science (IACS).
IACS has received a $6.3 million anonymous donation to advance data-driven research that will improve understanding of some of the world’s most pressing challenges, including climate change, machine learning and next generation nuclear energy, among others. Meanwhile, a $5 million grant from the National Science Foundation (NSF) to IACS will enable researchers nationwide to test future supercomputing technologies and advance computational and data-driven research.
Research by Jason Trelewicz, PhD, associate professor of Materials Science and Engineering, seeks to develop materials for next generation nuclear technologies, which are safer and more economical sources of nuclear power. As seen here, Trelewicz prepares to place a surrogate nuclear fuel pellet inside of a 3D X-ray microscope. The large sets of data collected from the microscope are processed by the high performance computing technology at the Institute for Advanced Computational Science.
Advanced computational science accelerates the speed of calculations, allowing researchers to process data and develop conclusions beyond the scope of what was previously possible. Through advanced computation, Stony Brook’s IACS researchers are redefining scientific understanding across multiple disciplines.
Commenting on the anonymous donation, Interim President Michael A. Bernstein said: “Thanks to this exceptionally generous gift, our IACS researchers will redefine the boundaries of advanced computational science.”
“In the age of supercomputing, we have a profound ability to transform scientific understanding across disciplines. I am confident IACS will be at the forefront of these remarkable research endeavors,” he said.
The donation will help to expand research excellence, sustain the IACS with the hiring of additional staff, extend the Institute’s STRIDE program to all students and serve as seed money for engagement initiatives.
Multidisciplinary research will be expanded in pursuit of the following themes:
Machine Learning and Statistical Inference for Scientific Discovery
Modeling, Simulation and Analysis of Complex Systems
High-Performance Computing and Future Computing Technologies
Making Sense of Data
“Whether modeling climate change, the inner workings of the human brain, or nuclear energies, our future is dependent on supercomputing and big data,” said Robert J. Harrison, PhD, professor of applied mathematics and statistics and director of IACS. “Research data has the powerful ability to evolve our perspectives and where new ideas emerge, innovation follows. This generous donation allows us to not only expand our resources but also empower scientific breakthroughs for a more prosperous future.”
The NSF grant is in support of the Ookami system, serving as a testbed for advanced computer technologies, which is expected to signal a new generation of high-speed U.S. supercomputers. Using a Cray ARM-based system, Ookami will deliver remarkably high performance for scientific applications, in part due to its blazing fast memory.
Harrison expects that these advanced technologies will enable researchers to more quickly and effectively conduct computational investigations. The project is led by IACS faculty in partnership with co-PI Matt Jones, PhD at the State University of New York at Buffalo, whose team will lead the capture of detailed operational metrics and provision of extensive analytical capabilities. At Stony Brook, co-PI Barbara Chapman, PhD, Department of Computer Science, studies programming models for parallel computing and scalable machine learning; and co-PI Alan Calder, PhD, Department of Physics and Astronomy, whose research in the field of nuclear astrophysics demands massive calculations to explore bright stellar explosions known as Type Ia supernovae.
“Research in computational science has increasingly defined the next major frontiers for discovery and innovation. This grant gives us the opportunity to imagine, refine, and develop a new generation of supercomputing technology. It is also a testament to the quality of our researchers — colleagues who stand truly at the cutting-edges of advanced computational science,” Bernstein said.
Since 2012, the IACS has led efforts to communicate science, apply high-performance computing to bridge disciplines and enable data-driven research. The institute’s initial funding stemmed from a $10 million anonymous donation, plus matching funds of equal value from the Simons Foundation.