skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Zonal methods for the parallel execution of range-limited N-body simulations

Abstract

Particle simulations in fields ranging from biochemistry to astrophysics require the evaluation of interactions between all pairs of particles separated by less than some fixed interaction radius. The applicability of such simulations is often limited by the time required for calculation, but the use of massive parallelism to accelerate these computations is typically limited by inter-processor communication requirements. Recently, Snir [M. Snir, A note on N-body computations with cutoffs, Theor. Comput. Syst. 37 (2004) 295-318] and Shaw [D.E. Shaw, A fast, scalable method for the parallel evaluation of distance-limited pairwise particle interactions, J. Comput. Chem. 26 (2005) 1318-1328] independently introduced two distinct methods that offer asymptotic reductions in the amount of data transferred between processors. In the present paper, we show that these schemes represent special cases of a more general class of methods, and introduce several new algorithms in this class that offer practical advantages over all previously described methods for a wide range of problem parameters. We also show that several of these algorithms approach an approximate lower bound on inter-processor data transfer.

Authors:
 [1];  [1];  [2]
  1. D. E. Shaw Research, LLC, New York, NY 10036 (United States)
  2. D. E. Shaw Research, LLC, New York, NY 10036 (United States) and Center for Computational Biology and Bioinformatics, Columbia University, New York, NY 10032 (United States). E-mail: david@deshaw.com
Publication Date:
OSTI Identifier:
20991555
Resource Type:
Journal Article
Resource Relation:
Journal Name: Journal of Computational Physics; Journal Volume: 221; Journal Issue: 1; Other Information: DOI: 10.1016/j.jcp.2006.06.014; PII: S0021-9991(06)00291-9; Copyright (c) 2006 Elsevier Science B.V., Amsterdam, The Netherlands, All rights reserved; Country of input: International Atomic Energy Agency (IAEA)
Country of Publication:
United States
Language:
English
Subject:
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS; ALGORITHMS; ASTROPHYSICS; BIOCHEMISTRY; EVALUATION; MANY-BODY PROBLEM; MATHEMATICAL MODELS; MOLECULAR DYNAMICS METHOD; PARTICLE INTERACTIONS; SIMULATION

Citation Formats

Bowers, Kevin J., Dror, Ron O., and Shaw, David E. Zonal methods for the parallel execution of range-limited N-body simulations. United States: N. p., 2007. Web. doi:10.1016/j.jcp.2006.06.014.
Bowers, Kevin J., Dror, Ron O., & Shaw, David E. Zonal methods for the parallel execution of range-limited N-body simulations. United States. doi:10.1016/j.jcp.2006.06.014.
Bowers, Kevin J., Dror, Ron O., and Shaw, David E. Sat . "Zonal methods for the parallel execution of range-limited N-body simulations". United States. doi:10.1016/j.jcp.2006.06.014.
@article{osti_20991555,
title = {Zonal methods for the parallel execution of range-limited N-body simulations},
author = {Bowers, Kevin J. and Dror, Ron O. and Shaw, David E.},
abstractNote = {Particle simulations in fields ranging from biochemistry to astrophysics require the evaluation of interactions between all pairs of particles separated by less than some fixed interaction radius. The applicability of such simulations is often limited by the time required for calculation, but the use of massive parallelism to accelerate these computations is typically limited by inter-processor communication requirements. Recently, Snir [M. Snir, A note on N-body computations with cutoffs, Theor. Comput. Syst. 37 (2004) 295-318] and Shaw [D.E. Shaw, A fast, scalable method for the parallel evaluation of distance-limited pairwise particle interactions, J. Comput. Chem. 26 (2005) 1318-1328] independently introduced two distinct methods that offer asymptotic reductions in the amount of data transferred between processors. In the present paper, we show that these schemes represent special cases of a more general class of methods, and introduce several new algorithms in this class that offer practical advantages over all previously described methods for a wide range of problem parameters. We also show that several of these algorithms approach an approximate lower bound on inter-processor data transfer.},
doi = {10.1016/j.jcp.2006.06.014},
journal = {Journal of Computational Physics},
number = 1,
volume = 221,
place = {United States},
year = {Sat Jan 20 00:00:00 EST 2007},
month = {Sat Jan 20 00:00:00 EST 2007}
}
  • A number of complex physical problems can be approached through N-body simulation, from fluid flow at high Reynolds number to gravitational astrophysics and molecular dynamics. In all these applications, direct summation is prohibitively expensive for large N and thus hierarchical methods are employed for fast summation. This work introduces new algorithms, based on k-means clustering, for partitioning parallel hierarchical N-body interactions. We demonstrate that the number of particle-cluster interactions and the order at which they are performed are directly affected by partition geometry. Weighted k-means partitions minimize the sum of clusters' second moments and create well-localized domains, and thus reducemore » the computational cost of N-body approximations by enabling the use of lower-order approximations and fewer cells. We also introduce compatible techniques for dynamic load balancing, including adaptive scaling of cluster volumes and adaptive redistribution of cluster centroids. We demonstrate the performance of these algorithms by constructing a parallel treecode for vortex particle simulations, based on the serial variable-order Cartesian code developed by Lindsay and Krasny [Journal of Computational Physics 172 (2) (2001) 879-907]. The method is applied to vortex simulations of a transverse jet. Results show outstanding parallel efficiencies even at high concurrencies, with velocity evaluation errors maintained at or below their serial values; on a realistic distribution of 1.2 million vortex particles, we observe a parallel efficiency of 98% on 1024 processors. Excellent load balance is achieved even in the face of several obstacles, such as an irregular, time-evolving particle distribution containing a range of length scales and the continual introduction of new vortex particles throughout the domain. Moreover, results suggest that k-means yields a more efficient partition of the domain than a global oct-tree.« less
  • In this paper, the influence of the parallel nonlinearity on zonal flows and heat transport in global particle-in-cell ion-temperature-gradient simulations is studied. Although this term is in theory orders of magnitude smaller than the others, several authors [L. Villard, P. Angelino, A. Bottino et al., Plasma Phys. Contr. Fusion 46, B51 (2004); L. Villard, S. J. Allfrey, A. Bottino et al., Nucl. Fusion 44, 172 (2004); J. C. Kniep, J. N. G. Leboeuf, and V. C. Decyck, Comput. Phys. Commun. 164, 98 (2004); J. Candy, R. E. Waltz, S. E. Parker et al., Phys. Plasmas 13, 074501 (2006)] found differentmore » results on its role. The study is performed using the global gyrokinetic particle-in-cell codes TORB (theta-pinch) [R. Hatzky, T. M. Tran, A. Koenies et al., Phys. Plasmas 9, 898 (2002)] and ORB5 (tokamak geometry) [S. Jolliet, A. Bottino, P. Angelino et al., Comput. Phys. Commun. 177, 409 (2007)]. In particular, it is demonstrated that the parallel nonlinearity, while important for energy conservation, affects the zonal electric field only if the simulation is noise dominated. When a proper convergence is reached, the influence of parallel nonlinearity on the zonal electric field, if any, is shown to be small for both the cases of decaying and driven turbulence.« less
  • The evolution of {bold E{times}B}-drifting plasma clouds is investigated with the aid of a computational technique denoted here as {open_quotes}dielectric-in-cell.{close_quotes} Many of the familiar phenomena associated with clouds of collisionless plasma are seen and explained and less-well-known phenomena associated with convection patterns, with the stripping of cloud material, and with the evolution of plasma clouds composed of differing ion species are investigated. The effects of spatially uniform diffusion are studied with the dielectric-in-cell technique and with another computational technique denoted as {open_quotes}{ital N}-body dielectric;{close_quotes} the suppression of convection, the suppression of structure growth, the increase in material stripping, and themore » evolution of cloud anisotropy are examined. {copyright} {ital 1998 American Institute of Physics.}« less
  • A novel execution model for rule application in active databases is developed and applied to the problem of updating derived data in a database represented using a semantic, object-based database model. The execution model is based on the use of `limited ambiguity rules` (LARs), which permit disjunction in rule actions. The execution model essentially performs a breadth-first exploration of alternative extensions of a user-requested update. Given an object-based database scheme, both integrity constraints and specifications of derived classes and attributes are compiled into a family of limited ambiguity rules. A theoretical analysis shows that the approach is sound: the executionmore » model returns all valid `completions` of a user-requested update, or terminates with an appropriate error notification. The complexity of the approach in connection with derived data update is considered. 42 refs., 10 figs., 3 tabs.« less