Local search to improve coordinate-based task mapping
Abstract
We present a local search strategy to improve the coordinate-based mapping of a parallel job’s tasks to the MPI ranks of its parallel allocation in order to reduce network congestion and the job’s communication time. The goal is to reduce the number of network hops between communicating pairs of ranks. Our target is applications with a nearest-neighbor stencil communication pattern running on mesh systems with non-contiguous processor allocation, such as Cray XE and XK Systems. Utilizing the miniGhost mini-app, which models the shock physics application CTH, we demonstrate that our strategy reduces application running time while also reducing the runtime variability. Furthermore, we further show that mapping quality can vary based on the selected allocation algorithm, even between allocation algorithms of similar apparent quality.
- Authors:
-
- Washington Univ., St. Louis, MO (United States)
- Knox College, Galesburg, IL (United States)
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Publication Date:
- Research Org.:
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Sponsoring Org.:
- USDOE National Nuclear Security Administration (NNSA)
- OSTI Identifier:
- 1238587
- Alternate Identifier(s):
- OSTI ID: 1251777
- Report Number(s):
- SAND-2015-8243J
Journal ID: ISSN 0167-8191; PII: S0167819115001441
- Grant/Contract Number:
- AC04-94AL85000; 899808
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Parallel Computing
- Additional Journal Information:
- Journal Volume: 51; Journal Issue: C; Journal ID: ISSN 0167-8191
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 97 MATHEMATICS AND COMPUTING; task mapping; stencil communication pattern; non-contiguous allocation; local search
Citation Formats
Balzuweit, Evan, Bunde, David P., Leung, Vitus J., Finley, Austin, and Lee, Alan C. S. Local search to improve coordinate-based task mapping. United States: N. p., 2015.
Web. doi:10.1016/j.parco.2015.10.012.
Balzuweit, Evan, Bunde, David P., Leung, Vitus J., Finley, Austin, & Lee, Alan C. S. Local search to improve coordinate-based task mapping. United States. https://doi.org/10.1016/j.parco.2015.10.012
Balzuweit, Evan, Bunde, David P., Leung, Vitus J., Finley, Austin, and Lee, Alan C. S. Sat .
"Local search to improve coordinate-based task mapping". United States. https://doi.org/10.1016/j.parco.2015.10.012. https://www.osti.gov/servlets/purl/1238587.
@article{osti_1238587,
title = {Local search to improve coordinate-based task mapping},
author = {Balzuweit, Evan and Bunde, David P. and Leung, Vitus J. and Finley, Austin and Lee, Alan C. S.},
abstractNote = {We present a local search strategy to improve the coordinate-based mapping of a parallel job’s tasks to the MPI ranks of its parallel allocation in order to reduce network congestion and the job’s communication time. The goal is to reduce the number of network hops between communicating pairs of ranks. Our target is applications with a nearest-neighbor stencil communication pattern running on mesh systems with non-contiguous processor allocation, such as Cray XE and XK Systems. Utilizing the miniGhost mini-app, which models the shock physics application CTH, we demonstrate that our strategy reduces application running time while also reducing the runtime variability. Furthermore, we further show that mapping quality can vary based on the selected allocation algorithm, even between allocation algorithms of similar apparent quality.},
doi = {10.1016/j.parco.2015.10.012},
journal = {Parallel Computing},
number = C,
volume = 51,
place = {United States},
year = {Sat Oct 31 00:00:00 EDT 2015},
month = {Sat Oct 31 00:00:00 EDT 2015}
}
Web of Science
Works referenced in this record:
Reducing the bulk in the bulk Synchronous Parallel Model
journal, December 2013
- Barrett, R. F.; Vaughan, C. T.; Hammond, S. D.
- Parallel Processing Letters, Vol. 23, Issue 04
Benefits of Topology Aware Mapping for Mesh Interconnects
journal, December 2008
- Bhatelé, Abhinav; Kalé, Laxmikant V.
- Parallel Processing Letters, Vol. 18, Issue 04
Noncontiguous processor allocation algorithms for mesh-connected multicomputers
journal, July 1997
- Lo, V.; Windisch, K. J.
- IEEE Transactions on Parallel and Distributed Systems, Vol. 8, Issue 7
Communication-Aware Processor Allocation for Supercomputers: Finding Point Sets of Small Average Distance
journal, September 2007
- Bender, Michael A.; Bunde, David P.; Demaine, Erik D.
- Algorithmica, Vol. 50, Issue 2
Compact location problems
journal, July 1997
- Krumke, S. O.; Marathe, M. V.; Noltemeier, H.
- Theoretical Computer Science, Vol. 181, Issue 2
Navigating an Evolutionary Fast Path to Exascale
conference, November 2012
- Barrett, R. F.; Hammond, S. D.; Vaughan, C. T.
- 2012 SC Companion: High Performance Computing, Networking, Storage and Analysis (SCC), 2012 SC Companion: High Performance Computing, Networking Storage and Analysis
PReMAS: Simulator for Resource Management
conference, September 2014
- Bunde, David P.; Leung, Vitus J.
- 2014 43nd International Conference on Parallel Processing Workshops (ICCPW), 2014 43rd International Conference on Parallel Processing Workshops
Resource allocation and utilization in the Blue Gene/L supercomputer
journal, March 2005
- Aridor, Y.; Domany, T.; Goldshmidt, O.
- IBM Journal of Research and Development, Vol. 49, Issue 2.3
Parallel static and dynamic multi-constraint graph partitioning
journal, January 2002
- Schloegel, Kirk; Karypis, George; Kumar, Vipin
- Concurrency and Computation: Practice and Experience, Vol. 14, Issue 3
On the Mapping Problem
journal, March 1981
- Bokhari,
- IEEE Transactions on Computers, Vol. C-30, Issue 3
Genetic algorithm based heuristics for the mapping problem
journal, January 1995
- Chockalingam, T.; Arunkumar, S.
- Computers & Operations Research, Vol. 22, Issue 1
Heuristic technique for processor and link assignment in multicomputers
journal, March 1991
- Bollinger, S. W.; Midkiff, S. F.
- IEEE Transactions on Computers, Vol. 40, Issue 3
Works referencing / citing this record:
On the effects of allocation strategies for exascale computing systems with distributed storage and unified interconnects
journal, July 2018
- Pascual, Jose A.; Lant, Joshua; Concatto, Caroline
- Concurrency and Computation: Practice and Experience, Vol. 31, Issue 21