Advances in Patch-Based Adaptive Mesh Refinement Scalability
Abstract
Patch-based structured adaptive mesh refinement (SAMR) is widely used for high-resolution simu- lations. Combined with modern supercomputers, it could provide simulations of unprecedented size and resolution. A persistent challenge for this com- bination has been managing dynamically adaptive meshes on more and more MPI tasks. The dis- tributed mesh management scheme in SAMRAI has made some progress SAMR scalability, but early al- gorithms still had trouble scaling past the regime of 105 MPI tasks. This work provides two critical SAMR regridding algorithms, which are integrated into that scheme to ensure efficiency of the whole. The clustering algorithm is an extension of the tile- clustering approach, making it more flexible and efficient in both clustering and parallelism. The partitioner is a new algorithm designed to prevent the network congestion experienced by its prede- cessor. We evaluated performance using weak- and strong-scaling benchmarks designed to be difficult for dynamic adaptivity. Results show good scaling on up to 1.5M cores and 2M MPI tasks. Detailed timing diagnostics suggest scaling would continue well past that.
- Authors:
-
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Publication Date:
- Research Org.:
- Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1241965
- Report Number(s):
- LLNL-JRNL-668377
Journal ID: ISSN 0743-7315
- Grant/Contract Number:
- AC52-07NA27344
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Journal of Parallel and Distributed Computing
- Additional Journal Information:
- Journal Volume: 89; Journal ID: ISSN 0743-7315
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 97 MATHEMATICS AND COMPUTING
Citation Formats
Gunney, Brian T.N., and Anderson, Robert W. Advances in Patch-Based Adaptive Mesh Refinement Scalability. United States: N. p., 2015.
Web. doi:10.1016/j.jpdc.2015.11.005.
Gunney, Brian T.N., & Anderson, Robert W. Advances in Patch-Based Adaptive Mesh Refinement Scalability. United States. https://doi.org/10.1016/j.jpdc.2015.11.005
Gunney, Brian T.N., and Anderson, Robert W. Fri .
"Advances in Patch-Based Adaptive Mesh Refinement Scalability". United States. https://doi.org/10.1016/j.jpdc.2015.11.005. https://www.osti.gov/servlets/purl/1241965.
@article{osti_1241965,
title = {Advances in Patch-Based Adaptive Mesh Refinement Scalability},
author = {Gunney, Brian T.N. and Anderson, Robert W.},
abstractNote = {Patch-based structured adaptive mesh refinement (SAMR) is widely used for high-resolution simu- lations. Combined with modern supercomputers, it could provide simulations of unprecedented size and resolution. A persistent challenge for this com- bination has been managing dynamically adaptive meshes on more and more MPI tasks. The dis- tributed mesh management scheme in SAMRAI has made some progress SAMR scalability, but early al- gorithms still had trouble scaling past the regime of 105 MPI tasks. This work provides two critical SAMR regridding algorithms, which are integrated into that scheme to ensure efficiency of the whole. The clustering algorithm is an extension of the tile- clustering approach, making it more flexible and efficient in both clustering and parallelism. The partitioner is a new algorithm designed to prevent the network congestion experienced by its prede- cessor. We evaluated performance using weak- and strong-scaling benchmarks designed to be difficult for dynamic adaptivity. Results show good scaling on up to 1.5M cores and 2M MPI tasks. Detailed timing diagnostics suggest scaling would continue well past that.},
doi = {10.1016/j.jpdc.2015.11.005},
journal = {Journal of Parallel and Distributed Computing},
number = ,
volume = 89,
place = {United States},
year = {Fri Dec 18 00:00:00 EST 2015},
month = {Fri Dec 18 00:00:00 EST 2015}
}
Web of Science
Works referenced in this record:
An arbitrary Lagrangian–Eulerian method with adaptive mesh refinement for the solution of the Euler equations
journal, September 2004
- Anderson, R. W.; Elliott, N. S.; Pember, R. B.
- Journal of Computational Physics, Vol. 199, Issue 2
Local adaptive mesh refinement for shock hydrodynamics
journal, May 1989
- Berger, M. J.; Colella, P.
- Journal of Computational Physics, Vol. 82, Issue 1
Adaptive mesh refinement for hyperbolic partial differential equations
journal, March 1984
- Berger, Marsha J.; Oliger, Joseph
- Journal of Computational Physics, Vol. 53, Issue 3
An algorithm for point clustering and grid generation
journal, January 1991
- Berger, M.; Rigoutsos, I.
- IEEE Transactions on Systems, Man, and Cybernetics, Vol. 21, Issue 5
PT-Scotch: A tool for efficient parallel graph ordering
journal, July 2008
- Chevalier, C.; Pellegrini, F.
- Parallel Computing, Vol. 34, Issue 6-8
Multidimensional upwind methods for hyperbolic conservation laws
journal, March 1990
- Colella, Phillip
- Journal of Computational Physics, Vol. 87, Issue 1
Efficient solution algorithms for the Riemann problem for real gases
journal, June 1985
- Colella, Phillip; Glaz, Harland M.
- Journal of Computational Physics, Vol. 59, Issue 2
Zoltan data management services for parallel dynamic applications
journal, January 2002
- Devine, K.; Boman, E.; Heaphy, R.
- Computing in Science & Engineering, Vol. 4, Issue 2
Parallel clustering algorithms for structured AMR
journal, November 2006
- Gunney, Brian T. N.; Wissink, Andrew M.; Hysom, David A.
- Journal of Parallel and Distributed Computing, Vol. 66, Issue 11
Managing application complexity in the SAMRAI object-oriented framework
journal, January 2002
- Hornung, Richard D.; Kohn, Scott R.
- Concurrency and Computation: Practice and Experience, Vol. 14, Issue 5
A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs
journal, January 1998
- Karypis, George; Kumar, Vipin
- SIAM Journal on Scientific Computing, Vol. 20, Issue 1
Scalable parallel regridding algorithms for block-structured adaptive mesh refinement
journal, March 2011
- Luitjens, J.; Berzins, M.
- Concurrency and Computation: Practice and Experience, Vol. 23, Issue 13
Parallel space-filling curve generation through sorting
journal, January 2007
- Luitjens, J.; Berzins, M.; Henderson, T.
- Concurrency and Computation: Practice and Experience, Vol. 19, Issue 10
Adaptive Mesh Refinement for Wave Propagation in Nonlinear Solids
journal, July 1995
- Trangenstein, John A.
- SIAM Journal on Scientific Computing, Vol. 16, Issue 4
Enhancing scalability of parallel structured AMR calculations
conference, January 2003
- Wissink, Andrew M.; Hysom, David; Hornung, Richard D.
- Proceedings of the 17th annual international conference on Supercomputing - ICS '03
Works referencing / citing this record:
Full Eulerian deformable solid‐fluid interaction scheme based on building‐cube method for large‐scale parallel computing
journal, October 2018
- Nishiguchi, Koji; Bale, Rahul; Okazawa, Shigenobu
- International Journal for Numerical Methods in Engineering, Vol. 117, Issue 2
Extension of the subgrid-scale gradient model for compressible magnetohydrodynamics turbulent instabilities
journal, October 2019
- Viganò, Daniele; Aguilera-Miret, Ricard; Palenzuela, Carlos
- Physics of Fluids, Vol. 31, Issue 10
A Simflowny-based finite-difference code for high-performance computing in numerical relativity
journal, August 2018
- Palenzuela, Carlos; Miñano, Borja; Viganò, Daniele
- Classical and Quantum Gravity, Vol. 35, Issue 18
Gravitational waves from dark boson star binary mergers
journal, November 2018
- Bezares, Miguel; Palenzuela, Carlos
- Classical and Quantum Gravity, Vol. 35, Issue 23
Gradient subgrid-scale model for relativistic MHD large-eddy simulations
journal, March 2020
- Carrasco, Federico; Viganò, Daniele; Palenzuela, Carlos
- Physical Review D, Vol. 101, Issue 6
Extension of the sub-grid-scale gradient model for compressible magnetohydrodynamics turbulent instabilities
text, January 2019
- Viganò, Daniele; Aguilera-Miret, Ricard; Palenzuela, Carlos
- arXiv
Gradient sub-grid-scale model for relativistic MHD Large Eddy Simulations
text, January 2019
- Carrasco, Federico; Viganò, Daniele; Palenzuela, Carlos
- arXiv