Towards electronic structure-based ab-initio molecular dynamics simulations with hundreds of millions of atoms

We push the boundaries of electronic structure-based ab-initio molecular dynamics (AIMD) beyond 100 million atoms. This scale is otherwise barely reachable with classical force-field methods or novel neural network and machine learning potentials. We achieve this breakthrough by combining innovation...

Full description

Saved in:
Bibliographic Details
Published in:Parallel computing Vol. 111; p. 102920
Main Authors: Schade, Robert, Kenter, Tobias, Elgabarty, Hossam, Lass, Michael, Schütt, Ole, Lazzaro, Alfio, Pabst, Hans, Mohr, Stephan, Hutter, Jürg, Kühne, Thomas D., Plessl, Christian
Format: Journal Article
Language:English
Published: Elsevier B.V 01.07.2022
Subjects:
Ab-initio molecular dynamics
Approximate computing
High-performance computing
Large-scale linear algebra
Massively-parallel algorithms
Supercomputing
Supercomputing
Massively-parallel algorithms
High-performance computing
Ab-initio molecular dynamics
Approximate computing
Large-scale linear algebra
ISSN:0167-8191, 1872-7336
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We push the boundaries of electronic structure-based ab-initio molecular dynamics (AIMD) beyond 100 million atoms. This scale is otherwise barely reachable with classical force-field methods or novel neural network and machine learning potentials. We achieve this breakthrough by combining innovations in linear-scaling AIMD, efficient and approximate sparse linear algebra, low and mixed-precision floating-point computation on GPUs, and a compensation scheme for the errors introduced by numerical approximations. The core of our work is the non-orthogonalized local submatrix method (NOLSM), which scales very favorably to massively parallel computing systems and translates large sparse matrix operations into highly parallel, dense matrix operations that are ideally suited to hardware accelerators. We demonstrate that the NOLSM method, which is at the center point of each AIMD step, is able to achieve a sustained performance of 324 PFLOP/s in mixed FP16/FP32 precision corresponding to an efficiency of 67.7% when running on 1536 NVIDIA A100 GPUs. •Record-sized electronic structure-based ab-initio molecular dynamics simulations are demonstrated.•Simulations with up to 102 million atoms are conducted using the CP2K simulation package.•Novel linear algebra and approximate computing methods are proposed and implemented.•Massive scalability of our submatrix method is demonstrated on 1536 NVIDIA A100 GPUs.•A sustained performance of 324 PFLOP/s (FP16/FP32) is achieved at an efficiency of 67.7%.
ISSN:0167-8191
1872-7336
DOI:10.1016/j.parco.2022.102920