Breaking down the parallel performance of GROMACS, a high-performance molecular dynamics software

DOI10.1007/978-3-031-30442-2_25arXiv2208.13658OpenAlexW4367355863MaRDI QIDQ6135469FDOQ6135469

Authors: Måns I. Andersson, Natarajan Arul Murugan, Artur Podobas, S. Markidis

Publication date: 25 August 2023

Published in: Parallel Processing and Applied Mathematics (Search for Journal in Brave)

Abstract: GROMACS is one of the most widely used HPC software packages using the Molecular Dynamics (MD) simulation technique. In this work, we quantify GROMACS parallel performance using different configurations, HPC systems, and FFT libraries (FFTW, Intel MKL FFT, and FFT PACK). We break down the cost of each GROMACS computational phase and identify non-scalable stages, such as MPI communication during the 3D FFT computation when using a large number of processes. We show that the Particle-Mesh Ewald phase and the 3D FFT calculation significantly impact the GROMACS performance. Finally, we discuss performance opportunities with a particular interest in developing GROMACS for the FFT calculations.

Full work available at URL: https://arxiv.org/abs/2208.13658

zbMATH Keywords

fast Fourier transform molecular dynamics Particle-Mesh Ewald calculations

Mathematics Subject Classification ID

Parallel numerical computation (65Y05) Research exposition (monographs, survey articles) pertaining to numerical analysis (65-02) Numerical methods for discrete and fast Fourier transforms (65T50) Research exposition (monographs, survey articles) pertaining to fluid mechanics (76-02)

This page was built for publication: Breaking down the parallel performance of GROMACS, a high-performance molecular dynamics software

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6135469)