Semi-Empirical Shadow Molecular Dynamics: A PyTorch implementation

TL;DR

This paper introduces a GPU-accelerated PyTorch implementation of semi-empirical shadow molecular dynamics using XL-BOMD, enabling efficient large-scale simulations of complex chemical systems with charge instabilities.

Contribution

The paper presents a novel PyTorch-based implementation of XL-BOMD with advanced features like finite electronic temperatures and Krylov subspace approximation, optimized for GPU hardware.

Findings

Simulation of 840 carbon atoms in 4 seconds per step on GPU

Enhanced ability to study charge unstable systems with low energy gaps

Modular architecture supports large-scale semiempirical quantum simulations

Abstract

Extended Lagrangian Born-Oppenheimer molecular dynamics (XL-BOMD) in its most recent shadow potential energy version has been implemented in the semiempirical PyTorch-based software PySeQM. The implementation includes finite electronic temperatures, canonical density matrix perturbation theory, and an adaptive Krylov Subspace Approximation for the integration of the electronic equations of motion within the XL-BOMB approach (KSA-XL-BOMD). The PyTorch implementation leverages the use of GPU and machine learning hardware accelerators for the simulations. The new XL-BOMD formulation allows studying more challenging chemical systems with charge instabilities and low electronic energy gaps. Current public release of PySeQM continues our development of modular architecture for large-scale simulations employing semiempirical quantum mechanical treatment. Applied to molecular dynamics simulation of 840 carbon atoms, one integration time step executes in 4 seconds on a single Nvidia RTX A6000 GPU.