Breaking Down the Parallel Performance of GROMACS, a High-Performance Molecular Dynamics Software

TL;DR

This paper analyzes the parallel performance of GROMACS, a widely used molecular dynamics software, by examining computational phases, configurations, and FFT libraries to identify bottlenecks and opportunities for optimization.

Contribution

It provides a detailed breakdown of GROMACS's computational phases and identifies non-scalable stages, offering insights for performance improvements, especially in FFT calculations.

Findings

MPI communication during 3D FFT limits scalability

Particle-Mesh Ewald and 3D FFT are performance bottlenecks

Performance varies with different FFT libraries

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.