Artificial intelligence and scientific computing play a critical role in helping researchers address some of the world’s most complex challenges — from solving environmental challenges and modeling the spread of infectious diseases to developing solutions that combat climate change.
With a new five-year, $15 million award from the U.S. Department of Energy’s Los Alamos National Laboratory, the University of Michigan will partner with the New Mexico-based laboratory to develop advanced computing technologies — including artificial intelligence and increasingly sophisticated modeling techniques — to address critical challenges.
For example, efficiently characterizing the behavior of turbulent plasma is key to unlocking the potential of nuclear fusion as a limitless source of clean energy, and to understanding fundamental phenomena that describe the evolution of stars.
“Even though there have been significant advances in computational algorithms and computer hardware, much of the potential of modeling and simulations remains to be realized in many applications that are of broad interest to society,” said Karthik Duraisamy, principal investigator on the project, a professor of aerospace engineering and director of the Michigan Institute for Computational Discovery and Engineering.
“In many cases, this is primarily due to a lack of alignment between the complex underlying physical processes, the computational algorithms to describe them and the computer codes and hardware used to execute the algorithms.
“Through this award, we have a unique opportunity to address this challenge by harnessing the breadth and depth of expertise from across U-M and LANL. Together, we will create a vertically integrated solution to help unlock a spectrum of unmet needs in scientific discovery and engineering design.”
The vertically integrated solution involves tightly coupling extreme-scale algorithm development and domain knowledge with elements of AI and computer architecture to generate significant efficiencies, scales and throughput so that researchers can address the root causes of, and potential solutions for, important global challenges.
While the research is foundational in nature and focuses on transformational advances in scientific computing, the team will initially apply these methods to explore the fundamental physics of fusion and its potential impact on clean energy and basic processes in stellar evolution.
As part of their ongoing collaboration, U-M also will provide LANL researchers with access to its suite of custom generative AI tools, as developed by the university’s Information & Technology Services team.
“We created these AI services to foster a spirit of innovation in our community and beyond,” said Ravi Pendse, vice president for information technology and chief information officer. “We look forward to growing that spirit exponentially through our ongoing work with Los Alamos National Laboratory.”
The collaboration builds upon President Santa J. Ono’s universitywide strategy to amplify research and scholarship, which explicitly calls for enhanced partnerships with national laboratories. LANL is one of 17 national laboratories overseen by the Energy Department, and each one possesses unique facilities and instruments to facilitate transdisciplinary research, with a strong emphasis on translating basic science to innovation.
In addition to its latest partnership with LANL, U-M also maintains strong collaborative research agreements with five other DoE national laboratories, including Idaho National Laboratory, Argonne National Laboratory and Fermilab in Illinois, Oak Ridge National Laboratory in Tennessee, and Pacific Northwest National Laboratory in Washington state.
“To strengthen and sustain our place as an exceptional public research university, and to establish and build U-M’s leadership in emerging research areas that are of high value to the nation, we are striving to enhance our partnerships with the Department of Energy’s national laboratories,” said Arthur Lupia, interim vice president for research and innovation.
“This particular partnership is especially exciting because it presents significant opportunities for a stronger collaboration between U-M and Los Alamos National Laboratory that spans a vast range of areas, including artificial intelligence, sustainability and clean energy, genomics, bioinformatics and quantum sciences.”
The following U-M faculty members are co-principal investigators on this research project: Reetuparna Das, associate professor of electrical engineering and computer science; Alex Gorodetsky, assistant professor of aerospace engineering; Brendan Kochunas, assistant professor of nuclear engineering and radiological sciences; and Scott Mahlke, professor of electrical engineering and computer science. The team also includes faculty and students from the College of Engineering and LSA.
