Efficiency of classical molecular dynamics algorithms on supercomputing hardware

Authors Grigory S. Smirnov, Vladimir V. Stegailov

Publication date 21 October 2016

Published in Matematicheskoe Modelirovanie (Search for Journal in Brave)

Full work available at URL http://mathnet.ru/eng/mm3732

algorithms for computer arithmetic algorithms for specific classes of architectures complexity and performance of numerical algorithms LAMMPS quantum algorithms and complexity

Mathematics Subject Classification ID

Numerical algorithms for computer arithmetic, etc. (65Y04) Numerical algorithms for specific classes of architectures (65Y10) Complexity and performance of numerical algorithms (65Y20) Quantum algorithms and complexity in the theory of computing (68Q12)

Review text

A new notion of the metric ``time of calculation -- peak performance based on practical grounds is introduced as a clear criterion of the calculation efficiency of some algorithms and corresponding software. The known program LINPACK fails to satisfy necessary requirements. Here, the metric introduced is applied for comparison of different program software (modern and old-fashioned) by use the test program software LAMMPS widely applied in quantum and molecular approaches to modeling processes. It is demonstrated that the metric can be efficiently used for definitive comparison of different combinations of central processors, accelerators and interconnects.

Recommendations

Cited in

(1)

Integrating state of the art compute, communication, and autotuning strategies to multiply the performance of ab initio molecular dynamics on massively parallel multi-core supercomputers

Describes a project that uses

Uses Software

LAMMPS

This page was built for publication: Efficiency of classical molecular dynamics algorithms on supercomputing hardware

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2827943)