Abstract:Neural-network quantum molecular dynamics (NNQMD) simulations based on machine learning are revolutionizing atomistic simulations of materials by providing quantum-mechanical accuracy but orders-of-magnitude faster, illustrated by ACM Gordon Bell prize (2020) and finalist (2021). State-of-the-art (SOTA) NNQMD model founded on group theory featuring rotational equivariance and local descriptors has provided much higher accuracy and speed than those models, thus named Allegro (meaning fast). On massively parallel supercomputers, however, it suffers a fidelity-scaling problem, where growing number of unphysical predictions of interatomic forces prohibits simulations involving larger numbers of atoms for longer times. Here, we solve this problem by combining the Allegro model with sharpness aware minimization (SAM) for enhancing the robustness of model through improved smoothness of the loss landscape. The resulting Allegro-Legato (meaning fast and "smooth") model was shown to elongate the time-to-failure $t_\textrm{failure}$, without sacrificing computational speed or accuracy. Specifically, Allegro-Legato exhibits much weaker dependence of timei-to-failure on the problem size, $t_{\textrm{failure}} \propto N^{-0.14}$ ($N$ is the number of atoms) compared to the SOTA Allegro model $\left(t_{\textrm{failure}} \propto N^{-0.29}\right)$, i.e., systematically delayed time-to-failure, thus allowing much larger and longer NNQMD simulations without failure. The model also exhibits excellent computational scalability and GPU acceleration on the Polaris supercomputer at Argonne Leadership Computing Facility. Such scalable, accurate, fast and robust NNQMD models will likely find broad applications in NNQMD simulations on emerging exaflop/s computers, with a specific example of accounting for nuclear quantum effects in the dynamics of ammonia.

Fully Convolutional Generative Machine Learning Method for Accelerating Non-Equilibrium Greens Function Simulations

Derivative of Electron Density in Non-Equilibrium Green'S Function Technique and Its Application to Boost Performance of Convergence

Predicting interacting Green's functions with neural networks

ELEQTRONeX: A GPU-Accelerated Exascale Framework for Non-Equilibrium Quantum Transport in Nanomaterials

A novel machine learning workflow to optimize cooling devices grounded in solid-state physics

Machine Learning for Performance Enhancement of Molecular Dynamics Simulations

Global Field Heterogeneous Graph Neural Networks for Accelerating Quantum Transport Calculation

Accelerating Nonequilibrium Green functions simulations with embedding selfenergies

Simulation and Modeling of Novel Electronic Device Architectures with NESS (Nano-Electronic Simulation Software): A Modular Nano TCAD Simulation Framework

AD-NEGF: An End-to-End Differentiable Quantum Transport Simulator for Sensitivity Analysis and Inverse Problems

Allegro-Legato: Scalable, Fast, and Robust Neural-Network Quantum Molecular Dynamics via Sharpness-Aware Minimization

Combining linear-scaling quantum transport and machine-learning molecular dynamics to study thermal and electronic transports in complex materials

Deep Learning Accelerated Quantum Transport Simulations in Nanoelectronics: From Break Junctions to Field-Effect Transistors

Molecular Simulations with a Pretrained Neural Network and Universal Pairwise Force Fields

Full Geometric Simulation of Miniaturized GaN Double-Heterojunction High Electron Mobility Transistors by a Multiscale Approach Coupling Quantum and Semi-Classical Transport

Non-Equilibrium Green Functions in Electronic Device Modeling

Neuroevolution machine learning potentials: Combining high accuracy and low cost in atomistic simulations and application to heat transport

LeapFrog: Getting the Jump on Multi-Scale Materials Simulations Using Machine Learning

Simulate Time-integrated Coarse-grained Molecular Dynamics with Geometric Machine Learning

NNP/MM: Accelerating molecular dynamics simulations with machine learning potentials and molecular mechanic

Enhancing Open Quantum Dynamics Simulations Using Neural Network-Based Non-Markovian Stochastic Schrödinger Equation Method