Accelerating Quantum Light-Matter Dynamics on Graphics Processing Units
ORAL
Abstract
We have developed a simulation method to study how electrons interact with an external laser pulse by solving the Maxwell equation for electrons and time-dependent density functional theory equations for electrons. We present an efficient way to perform electronic time-propagation as well as to reformulate the compute-intensive nonlocal correction into matrix operations, which resulted in a 644-fold speedup on Nvidia A100 GPU over AMD EPYC 7543 CPU of the Polaris computer at Argonne Leadership Computing Facility. In addition, the resulting DC-MESH (divide-&conquer Maxwell-Ehrenfest-surface hopping) code exhibited a weak-scaling parallel efficiency of 96.73% on 256 nodes (or 1,024 GPUs) of Polaris for 5,120-atom PbTiO3 material. This enables the study of light-induced topological switching for future ultrafast and ultralow-power ferroelectric topotronics applications.
*This work was supported by NSF grant OAC-2118061. The scalable code development was supported by the Aurora ESP program. An award for computer time was provided by the U.S. DOE Innovative and Novel Computational Impact on Theory and Experiment (INCITE) Program.
–
Publication: T. M. Razakh, T. Linker, Y. Luo, R. K. Kalia, K. Nomura, P. Vashishta, and A. Nakano, Proceedings of the International Workshop on Parallel and Distributed Scientific and Engineering Computing, PDSEC, pp. 1057-1066 (IEEE, San Francisco, CA, 2024).
Presenters
-
Taufeq Mohammed Razakh
- University of Southern California