Suscripción a Biblioteca: Guest
Portal Digitalde Biblioteca Digital eLibros Revistas Referencias y Libros de Ponencias Colecciones
International Journal for Multiscale Computational Engineering
Factor de Impacto: 1.016 Factor de Impacto de 5 años: 1.194 SJR: 0.554 SNIP: 0.68 CiteScore™: 1.18

ISSN Imprimir: 1543-1649
ISSN En Línea: 1940-4352

International Journal for Multiscale Computational Engineering

DOI: 10.1615/IntJMultCompEng.v5.i1.40
pages 27-38

Efficient Parallel Execution of Event-Driven Electromagnetic Hybrid Models

Kalyan Perumalla
Oak Ridge National Laboratory, Oak Ridge, TN, USA
Homa Karimabadi
SciberQuest Inc, Solana Beach, CA, USA
Richard Fujimoto
Georgia Institute of Technology, Atlanta, GA, USA

SINOPSIS

New discrete-event formulations of physics simulation models are emerging that can outperform traditional time-stepped models, especially in simulations containing multiple timescales. Detailed simulation of the Earth's magnetosphere, for example, requires execution of submodels that operate at timescales that differ by orders of magnitude. In contrast to time-stepped simulation, which requires tightly coupled updates to almost the entire system state at regular time intervals, the new approaches that use discrete event simulation (DES) modeling help evolve the states of submodels on relatively independent timescales. However, in contrast to the relative ease of parallelization of time-stepped codes, the parallelization of DES-based models raises challenges with respect to their scalability and performance. One of the key challenges is to improve the computation granularity to offset synchronization and communication overheads within and across processors. Our previous work on parallelization was limited in scalability and run-time performance due to such challenges. Here, we report on optimizations we performed on DES-based plasma simulation models to improve parallel execution performance. The mapping of the model to simulation processes is optimized via aggregation techniques, and the parallel run-time engine is optimized for communication and memory efficiency. The net result is the capability to simulate hybrid particle-in-cell models with over two billion ion particles using 512 processors on supercomputing platforms.


Articles with similar content:

Wavelet-Based Spatial Scaling of Coupled Reaction-Diffusion Fields
International Journal for Multiscale Computational Engineering, Vol.6, 2008, issue 4
Srdjan Simunovic, Sudib Kumar Mishra, Krishna Muralidharan, George Frantziskonis, Sreekanth Pannala, Pierre A. Deymier
Multilevel Parallel Programming for Multiscale Modeling of Composite Materials
International Journal for Multiscale Computational Engineering, Vol.2, 2004, issue 3
Somnath Ghosh, Paul M. Eder, James E. Giuliani
A MULTIGRID MULTILEVEL MONTE CARLO METHOD USING HIGH-ORDER FINITE-VOLUME SCHEME FOR LOGNORMAL DIFFUSION PROBLEMS
International Journal for Uncertainty Quantification, Vol.7, 2017, issue 1
Cornelis W. Oosterlee, Prashant Kumar, Richard P. Dwight
A MULTIZONE ADAPTIVE PARALLEL SCHEME FOR SOLIDIFICATION IN A RECTANGULAR CAVITY
International Heat Transfer Conference 11, Vol.20, 1998, issue
W. K. Chui, James Glimm, Yang Qiangsheng , F. Tangerman , V. Prasad, R. K. Sahoo, H. Zhang
PRIOR AND POSTERIOR ROBUST STOCHASTIC PREDICTIONS FOR DYNAMICAL SYSTEMS USING PROBABILITY LOGIC
International Journal for Uncertainty Quantification, Vol.3, 2013, issue 4
Alexandros Taflanidis, James L. Beck