A Monte Carlo approach for processor allocation in molecular dynamics simulations

Author

Ballestrero, P. ; Ruggiero, C.

Author_Institution

DIST, Genoa Univ., Italy

fYear

1994

fDate

1994

Firstpage

742

Abstract

A Monte Carlo algorithm for optimal processor allocation for Molecular Dynamics simulations of proteins on Single Instruction Multiple Data flow computer is presented. The algorithm starts from a conventional data structure based on space decomposition algorithm and produces a modified structure based on the atom decomposition approach, optimised in term of computation and communication time. The authors´ approach is well suited for Molecular Dynamics simulations on massively parallel computers based on square or cubic communication networks

Keywords

Monte Carlo methods; biology computing; digital simulation; molecular biophysics; molecular dynamics method; proteins; Monte Carlo algorithm; atom decomposition approach; communication time; cubic communication networks; massively parallel computers; molecular dynamics simulations; optimal processor allocation; protein dynamics; single instruction multiple data flow computer; space decomposition algorithm; square communication networks; Communication networks; Computational modeling; Computer aided instruction; Computer networks; Computer simulation; Concurrent computing; Data flow computing; Data structures; Monte Carlo methods; Proteins;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1994. Engineering Advances: New Opportunities for Biomedical Engineers. Proceedings of the 16th Annual International Conference of the IEEE

Conference_Location

Baltimore, MD

Print_ISBN

0-7803-2050-6

Type

conf

DOI

10.1109/IEMBS.1994.415270

Filename

415270