We test the relative performances of two different approaches to the computation of forces for molecular dynamics simulations on graphics processing units. A "vertex-based" approach, where a computing thread is started per particle, is compared to an "edge-based" approach, where a thread is started per each potentially non-zero interaction. We find that the former is more efficient for systems with many simple interactions per particle while the latter is more efficient if the system has more complicated interactions or fewer of them. By comparing computation times on more and less recent graphics processing unit technology, we predict that, if the current trend of increasing the number of processing cores - as opposed to their computing power - remains, the "edge-based" approach will gradually become the most efficient choice in an increasing number of cases.

We test the relative performances of two different approaches to the computation of forces for molecular dynamics simulations on graphics processing units. A "vertex-based" approach, where a computing thread is started per particle, is compared to an "edge-based" approach, where a thread is started per each potentially non-zero interaction. We find that the former is more efficient for systems with many simple interactions per particle while the latter is more efficient if the system has more complicated interactions or fewer of them. By comparing computation times on more and less recent graphics processing unit technology, we predict that, if the current trend of increasing the number of processing cores - as opposed to their computing power - remains, the "edge-based" approach will gradually become the most efficient choice in an increasing number of cases.

A Comparison between Parallelization Approaches in Molecular Dynamics Simulations on GPUs

ROMANO, Flavio
2015-01-01

Abstract

We test the relative performances of two different approaches to the computation of forces for molecular dynamics simulations on graphics processing units. A "vertex-based" approach, where a computing thread is started per particle, is compared to an "edge-based" approach, where a thread is started per each potentially non-zero interaction. We find that the former is more efficient for systems with many simple interactions per particle while the latter is more efficient if the system has more complicated interactions or fewer of them. By comparing computation times on more and less recent graphics processing unit technology, we predict that, if the current trend of increasing the number of processing cores - as opposed to their computing power - remains, the "edge-based" approach will gradually become the most efficient choice in an increasing number of cases.
File in questo prodotto:
File Dimensione Formato  
rovigatti-jcc-2015.pdf

non disponibili

Tipologia: Versione dell'editore
Licenza: Accesso chiuso-personale
Dimensione 532.48 kB
Formato Adobe PDF
532.48 kB Adobe PDF   Visualizza/Apri

I documenti in ARCA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/10278/3673819
Citazioni
  • ???jsp.display-item.citation.pmc??? 32
  • Scopus 82
  • ???jsp.display-item.citation.isi??? 77
social impact