Micron-scale heterogeneous catalysis with Bayesian force fields from first principles and active learning

TL;DR

This paper presents a Bayesian force field trained via active learning that enables quantum-mechanically accurate reactive MD simulations of H₂/Pt(111) at an unprecedented scale, leveraging GPU acceleration and uncertainty quantification.

Contribution

It introduces a scalable Bayesian force field trained with active learning for large-scale reactive MD, combining first-principles accuracy with high-performance computing.

Findings

Achieved reactive MD simulations with over 0.5 trillion atoms.

Enabled uncertainty quantification for atomic environments.

Demonstrated efficient GPU-accelerated simulations on Summit supercomputer.

Abstract

Quantum-mechanically accurate reactive molecular dynamics (MD) at the scale of billions of atoms has been achieved for the heterogeneous catalytic system of H$_2$/Pt(111) using the FLARE Bayesian force field. This achievement provides accelerated time-to-solution from first principles, with Bayesian active learning enabling efficient and autonomous training of the machine learning model. The resulting model is then deployed in LAMMPS on GPUs using the Kokkos performance portability library. The Bayesian force field provides quantitative uncertainty of predictions on every atomic environment, critical for detecting configurations in large reactive simulations that are outside of the training set. Scaling benchmarks were performed using real-application MD of the H$_2$/Pt(111) heterogeneous catalysis on the Summit supercomputer, with simulations reaching 0.5 trillion atoms on 4556 GPU nodes.