TorchSim: An efficient atomistic simulation engine in PyTorch

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cohen, Orion, Riebesell, Janosh, Goodall, Rhys, Kolluru, Adeesh, Falletta, Stefano, Krause, Joseph, Colindres, Jorge, Ceder, Gerbrand, Gangan, Abhijeet S.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916889786056704
author Cohen, Orion
Riebesell, Janosh
Goodall, Rhys
Kolluru, Adeesh
Falletta, Stefano
Krause, Joseph
Colindres, Jorge
Ceder, Gerbrand
Gangan, Abhijeet S.
author_facet Cohen, Orion
Riebesell, Janosh
Goodall, Rhys
Kolluru, Adeesh
Falletta, Stefano
Krause, Joseph
Colindres, Jorge
Ceder, Gerbrand
Gangan, Abhijeet S.
contents We introduce TorchSim, an open-source atomistic simulation engine tailored for the Machine Learned Interatomic Potential (MLIP) era. By rewriting core atomistic simulation primitives in PyTorch, TorchSim can achieve orders of magnitude acceleration for popular MLIPs. Unlike existing molecular dynamics packages, which simulate one system at a time, TorchSim performs batched simulations that efficiently utilize modern GPUs by evolving multiple systems concurrently. TorchSim supports molecular dynamics integrators, structural relaxation optimizers, both machine-learned and classical interatomic potentials (such as Lennard-Jones, Morse, soft-sphere), batching with automatic memory management, differentiable simulation, and integration with popular materials informatics tools.
format Preprint
id arxiv_https___arxiv_org_abs_2508_06628
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TorchSim: An efficient atomistic simulation engine in PyTorch
Cohen, Orion
Riebesell, Janosh
Goodall, Rhys
Kolluru, Adeesh
Falletta, Stefano
Krause, Joseph
Colindres, Jorge
Ceder, Gerbrand
Gangan, Abhijeet S.
Computational Physics
Materials Science
We introduce TorchSim, an open-source atomistic simulation engine tailored for the Machine Learned Interatomic Potential (MLIP) era. By rewriting core atomistic simulation primitives in PyTorch, TorchSim can achieve orders of magnitude acceleration for popular MLIPs. Unlike existing molecular dynamics packages, which simulate one system at a time, TorchSim performs batched simulations that efficiently utilize modern GPUs by evolving multiple systems concurrently. TorchSim supports molecular dynamics integrators, structural relaxation optimizers, both machine-learned and classical interatomic potentials (such as Lennard-Jones, Morse, soft-sphere), batching with automatic memory management, differentiable simulation, and integration with popular materials informatics tools.
title TorchSim: An efficient atomistic simulation engine in PyTorch
topic Computational Physics
Materials Science
url https://arxiv.org/abs/2508.06628