Parallel discrete molecular dynamics simulation with speculation and in-order commitment

Authors:
Md. Ashfaquzzaman Khan;Martin C. Herbordt
Affiliations:
Computer Architecture and Automated Design Laboratory, Department of Electrical and Computer Engineering, Boston University, Boston, MA 02215, United States;Computer Architecture and Automated Design Laboratory, Department of Electrical and Computer Engineering, Boston University, Boston, MA 02215, United States
Venue:
Journal of Computational Physics
Year:
2011

Citing 12
Cited 0

Parallel discrete event simulation

Communications of the ACM - Special issue on simulation
How to simulate billiards and similar systems

Journal of Computational Physics
Efficient algorithms for many-body hard particle molecular dynamics

Journal of Computational Physics
Analysis of an efficient algorithm for the hard-sphere problem

ACM Transactions on Modeling and Computer Simulation (TOMACS)
Billiards and related systems on the bulk-synchronous parallel model

Proceedings of the eleventh workshop on Parallel and distributed simulation
Algorithms for particle-field simulations with collisions

Journal of Computational Physics
Parallel Computer Architecture: A Hardware/Software Approach

Parallel Computer Architecture: A Hardware/Software Approach
Computer Architecture, Fourth Edition: A Quantitative Approach

Computer Architecture, Fourth Edition: A Quantitative Approach
A Complexity O(1) priority queue for event driven molecular dynamics simulations

Journal of Computational Physics
An Efficient O(1) Priority Queue for Large FPGA-Based Discrete Event Simulations of Molecular Dynamics

FCCM '08 Proceedings of the 2008 16th International Symposium on Field-Programmable Custom Computing Machines
Parallel Discrete Event Simulation of Molecular Dynamics Through Event-Based Decomposition

ASAP '09 Proceedings of the 2009 20th IEEE International Conference on Application-specific Systems, Architectures and Processors
The Art of Molecular Dynamics Simulation

The Art of Molecular Dynamics Simulation

Quantified Score

Hi-index	31.45

Visualization

Abstract

Discrete molecular dynamics simulation (DMD) uses simplified and discretized models enabling simulations to advance by event rather than by timestep. DMD is an instance of discrete event simulation and so is difficult to scale: even in this multi-core era, all reported DMD codes are serial. In this paper we discuss the inherent difficulties of scaling DMD and present our method of parallelizing DMD through event-based decomposition. Our method is microarchitecture inspired: speculative processing of events exposes parallelism, while in-order commitment ensures correctness. We analyze the potential of this parallelization method for shared-memory multiprocessors. Achieving scalability required extensive experimentation with scheduling and synchronization methods to mitigate serialization. The speed-up achieved for a variety of system sizes and complexities is nearly 6x on an 8-core and over 9x on a 12-core processor. We present and verify analytical models that account for the achieved performance as a function of available concurrency and architectural limitations.