GPUMD: A package for constructing accurate machine-learned potentials and performing highly efficient atomistic simulations

• Published in: The Journal of chemical physics Vol. 157; no. 11; pp. 114801 - 114826
• Main Authors: Fan, Zheyong, Wang, Yanzhou, Ying, Penghua, Song, Keke, Wang, Junjie, Wang, Yong, Zeng, Zezhu, Xu, Ke, Lindgren, Eric, Rahm, J. Magnus, Gabourie, Alexander J., Liu, Jiahui, Dong, Haikuan, Wu, Jianyang, Chen, Yue, Zhong, Zheng, Sun, Jian, Erhart, Paul, Su, Yanjing, Ala-Nissila, Tapio
• Format: Journal Article
• Language: English
• Published: Melville American Institute of Physics 21.09.2022
• Subjects: Basis functions; Chebyshev approximation; Graphics processing units; Model accuracy; Simulation; Workflow

We present our latest advancements of machine-learned potentials (MLPs) based on the neuroevolution potential (NEP) framework introduced in Fan et al. [Phys. Rev. B 104, 104309 (2021)] and their implementation in the open-source package gpumd. We increase the accuracy of NEP models both by improving...