Deep Potential Molecular Dynamics: a scalable model with the accuracy of quantum mechanics

TL;DR

Deep Potential Molecular Dynamics (DeePMD) employs a symmetry-preserving deep neural network trained on ab initio data to enable scalable, accurate molecular simulations comparable to quantum mechanics.

Contribution

The paper introduces DeePMD, a neural network-based scheme that accurately models many-body interatomic forces with linear scaling, preserving physical symmetries without ad hoc assumptions.

Findings

DeePMD achieves quantum-mechanics-level accuracy in simulations.

The method scales linearly with system size.

Results are indistinguishable from ab initio data.

Abstract

We introduce a scheme for molecular simulations, the Deep Potential Molecular Dynamics (DeePMD) method, based on a many-body potential and interatomic forces generated by a carefully crafted deep neural network trained with ab initio data. The neural network model preserves all the natural symmetries in the problem. It is "first principle-based" in the sense that there are no ad hoc components aside from the network model. We show that the proposed scheme provides an efficient and accurate protocol in a variety of systems, including bulk materials and molecules. In all these cases, DeePMD gives results that are essentially indistinguishable from the original data, at a cost that scales linearly with system size.