TU Darmstadt / ULB / TUbiblio

Crystal Structure Determination by Global Optimization in Configuration Space: A Case Study for Distributed Computing on the NRW-grid

Roth, G. ; Bischof, Christian ; Eifert, Th. (2011)
Crystal Structure Determination by Global Optimization in Configuration Space: A Case Study for Distributed Computing on the NRW-grid.
In: International Journal of Computational Science and Engineering, 6 (3)
doi: 10.1504/IJCSE.2011.042020
Artikel, Bibliographie

Kurzbeschreibung (Abstract)

The problem of ab-initio crystal structure determination from (incomplete) powder X-ray diffraction data is formulated as a global optimisation problem. The problem is of complexity class P, which suggests that the problem itself is practically infeasible for any but trivial structures. Yet, we show in this report that the inherent symmetry of the problem along with suitable heuristics and with massive computational power, allows the determination of structures with up to 20 atoms in the asymmetric unit from scratch. We give an overview of the methods employed and show numerical examples obtained by solving a model problem on the NRW-grid, which harnesses the computational power of several hundred computers at several universities in the state of Northrhine-Westfalia.

Typ des Eintrags: Artikel
Erschienen: 2011
Autor(en): Roth, G. ; Bischof, Christian ; Eifert, Th.
Art des Eintrags: Bibliographie
Titel: Crystal Structure Determination by Global Optimization in Configuration Space: A Case Study for Distributed Computing on the NRW-grid
Sprache: Englisch
Publikationsjahr: 1 Januar 2011
Verlag: Inderscience Publishing
Titel der Zeitschrift, Zeitung oder Schriftenreihe: International Journal of Computational Science and Engineering
Jahrgang/Volume einer Zeitschrift: 6
(Heft-)Nummer: 3
DOI: 10.1504/IJCSE.2011.042020
Zugehörige Links:
Kurzbeschreibung (Abstract):

The problem of ab-initio crystal structure determination from (incomplete) powder X-ray diffraction data is formulated as a global optimisation problem. The problem is of complexity class P, which suggests that the problem itself is practically infeasible for any but trivial structures. Yet, we show in this report that the inherent symmetry of the problem along with suitable heuristics and with massive computational power, allows the determination of structures with up to 20 atoms in the asymmetric unit from scratch. We give an overview of the methods employed and show numerical examples obtained by solving a model problem on the NRW-grid, which harnesses the computational power of several hundred computers at several universities in the state of Northrhine-Westfalia.

Fachbereich(e)/-gebiet(e): 20 Fachbereich Informatik
20 Fachbereich Informatik > Scientific Computing
Hinterlegungsdatum: 10 Jul 2012 11:59
Letzte Änderung: 11 Jan 2021 10:55
PPN:
Export:
Suche nach Titel in: TUfind oder in Google
Frage zum Eintrag Frage zum Eintrag

Optionen (nur für Redakteure)
Redaktionelle Details anzeigen Redaktionelle Details anzeigen