Optimizing the Performance for Nonbonded Force Calculation of Chemical Simulation on CBEA

Guo Fu Feng; Ming Wang; Ming Chen; Tao Chi

doi:10.4028/www.scientific.net/AMM.80-81.322

Paper Titles

Linear Location of Acoustic Emission Source Based on LS-SVR and NGA
p.302

Research and Development of Design Technologies for Functionally Gradient Ceramic Materials
p.307

Heat Transfer Model for Solidification Process of NH₄Cl Solution
p.313

Analysis of Vibration and Sound Propagation in a Fluid-Filled Pipe
p.317

Optimizing the Performance for Nonbonded Force Calculation of Chemical Simulation on CBEA
p.322

Preparation and Electrochemical Properties of VO₂(B) Nanobelts
p.327

Effect of Nitric Acid Modification on LiMn₂O₄ Prepared by Solution Combustion Synthesis
p.332

Simulation Research on Stress of Polymeric Patterns during Micro Hot Embossing
p.339

Research on the Combination Degree between Nylon and Insert in the Pipe Wrench
p.346

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 80-81Optimizing the Performance for Nonbonded Force...

Optimizing the Performance for Nonbonded Force Calculation of Chemical Simulation on CBEA

Abstract:

The molecular dynamics tool usually employs a neighbor list that contains the particle pairs for which non-bonded forces would be calculated. Though accessing to the array of atoms’ position and force does not exhibit a cache friendly pattern on a computer during the nonbonded force calculation, the neighbor list could be referred as a directive for the memory access. Designing specific data prefetching policy according to the neighbour list which was generated before the calculation can reduce the latency of memory access and improve the molecular dynamics simulation process. In this paper, an optimization memory access method referred by particle neighbor list based on CBEA was proposed. The experimental result shows that the proposed method reduces the latency of memory access and improves performance of the molecular-dynamics-based chemical simulation greatly.

You might also be interested in these eBooks

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volumes 80-81)

Pages:

322-326

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.80-81.322

Citation:

Cite this paper

Online since:

July 2011

Authors:

Guo Fu Feng, Ming Wang, Ming Chen, Tao Chi

Keywords:

Cell Broadband Engine, Chemical Simulation, Nonbonded Force, Software Cache

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

References

[1] Lindhal, E., Hess, B., van der Spoel, D., Gromacs 3. 0: A package for molecular simulation and trajectory analysis., Journal of Molecular Modeling. 7: 306-317, (2001).

DOI: 10.1007/s008940100045

Google Scholar

[2] J. A. Kahle, M. N. Day, H. P. Hofstee, C. R. Johns, T. R. Maeurer, and D. Shippy. Introduction to the cell multiprocessor. IBM Systems Journal, 49(4), (2005).

DOI: 10.1147/rd.494.0589

Google Scholar

[3] Meredith, Jeremy S. Alam, Sadaf R. Vetter, Jeffrey S. Analysis of a Computational Biology Simulation Technique on Emerging Processing Architectures, HiCOMB 2007, Sixth IEEE International Workshop on High Performance Computational Biology.

DOI: 10.1109/ipdps.2007.370444

Google Scholar

[4] Levon J . OProfile [ EB/ OL ] . http: /oprofile. source- forge. net/news.

Google Scholar

[5] S. Olivier, J. Prins, J. Derby, K. Vu, Porting the GROMACS Molecular Dynamics Code to the Cell Processor, Proc. of 8th IEEE Intl. Workshop on Parallel and Distributed Scientific and Engineering Computing (PDSEC-07) Long Beach, CA, March (2007).

DOI: 10.1109/ipdps.2007.370560

Google Scholar

[6] H. J. C. Berendsen, D. van der Spoel, and R. van Drunen. GROMACS - a message-passing parallel molecular dynamics implementation. Computer Phys. Commun., 91: 43-56, (1995).

DOI: 10.1016/0010-4655(95)00042-e

Google Scholar