CAREER: Machine Learning Methods for Accelerating Large-Scale Quantum Mechanical Simulations

Modern scientific progress in areas such as drug discovery and energy depends on the ability to accurately predict how matter behaves at the atomic level. This project aims to bridge the gap between small-scale theoretical models and the large-scale simulations required to design real-world technologies. By developing new machine learning tools that model physical systems with much greater efficiency, this research aims to enable discoveries at the length and timescales necessary for advanced engineering while maintaining high physical accuracy. The research focuses on overcoming the computational trade-off between the precision of quantum mechanical methods and the speed required for large-scale modeling. This will be achieved by developing new machine learning frameworks, guided by empirical scaling laws, that integrate physical principles to predict atomic interactions across diverse environments, allowing for the exploration of larger systems and longer timescales than currently possible. These technical goals are integrated with a mission to develop an interdisciplinary workforce through new training programs and a cross-disciplinary curriculum, ensuring that these tools and skills benefit a broad range of scientific fields. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2544469 | Program: 01002627DB NSF RESEARCH & RELATED ACTIVIT,01002930DB NSF RESEARCH & RELATED ACTIVIT,01003031DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Aditi Krishnapriyan | Institution: University of California-Berkeley, BERKELEY, CA | Award Amount: $426,864 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2544469 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2544469.html