Anzeige
Erschienen in:
2012 | OriginalPaper | Buchkapitel
Free Energy Monte Carlo Simulations on a Distributed Network
verfasst von : Luke Czapla, Alexey Siretskiy, John Grime, Malek O. Khan
Erschienen in: Applied Parallel and Scientific Computing
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
While the use of enhanced sampling techniques and parallel computing to determine potentials of mean force is in widespread use in modern Molecular Dynamics and Monte Carlo simulation studies, there have been few methods that efficiently combine heterogeneous computer resources of varying quality and speeds in realizing a single simulation result on a distributed network. Here, we apply an algorithm based on the Monte Carlo method of Wang and Landau within a client-server framework, in which individual computing nodes report a histogram of regions of phase space visited and corresponding updates to a centralized server at regular intervals entirely asynchronously. The server combines the data and reports the sum to all nodes so that the overall free energy determination scales linearly with the total amount of resources allocated. We discuss our development of this technique and present results for molecular simulations of DNA.