Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations

Handle URI:
http://hdl.handle.net/10754/600175
Title:
Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations
Authors:
Landge, A. G.; Levine, J. A.; Bhatele, A.; Isaacs, K. E.; Gamblin, T.; Schulz, M.; Langer, S. H.; Bremer, Peer-Timo; Pascucci, V.
Abstract:
The performance of massively parallel applications is often heavily impacted by the cost of communication among compute nodes. However, determining how to best use the network is a formidable task, made challenging by the ever increasing size and complexity of modern supercomputers. This paper applies visualization techniques to aid parallel application developers in understanding the network activity by enabling a detailed exploration of the flow of packets through the hardware interconnect. In order to visualize this large and complex data, we employ two linked views of the hardware network. The first is a 2D view, that represents the network structure as one of several simplified planar projections. This view is designed to allow a user to easily identify trends and patterns in the network traffic. The second is a 3D view that augments the 2D view by preserving the physical network topology and providing a context that is familiar to the application developers. Using the massively parallel multi-physics code pF3D as a case study, we demonstrate that our tool provides valuable insight that we use to explain and optimize pF3D-s performance on an IBM Blue Gene/P system. © 1995-2012 IEEE.
Citation:
Landge AG, Levine JA, Bhatele A, Isaacs KE, Gamblin T, et al. (2012) Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations. IEEE Transactions on Visualization and Computer Graphics 18: 2467–2476. Available: http://dx.doi.org/10.1109/TVCG.2012.286.
Publisher:
Institute of Electrical and Electronics Engineers (IEEE)
Journal:
IEEE Transactions on Visualization and Computer Graphics
KAUST Grant Number:
KUS-C1-016-04
Issue Date:
Dec-2012
DOI:
10.1109/TVCG.2012.286
PubMed ID:
26357155
Type:
Article
ISSN:
1077-2626
Sponsors:
This work is supported in part by NSF awards IIS-1045032, OCI-0904631, OCI-0906379 and CCF-0702817, and by a KAUST awardKUS-C1-016-04. This work was also performed under the auspicesof the U.S. Department of Energy by the University of Utah undercontracts DE-SC0001922, DE-AC52-07NA27344 and DE-FC02-06ER25781, and by Lawrence Livermore National Laboratory undercontract DE-AC52-07NA27344 (LLNL-CONF-543359).
Appears in Collections:
Publications Acknowledging KAUST Support

Full metadata record

DC FieldValue Language
dc.contributor.authorLandge, A. G.en
dc.contributor.authorLevine, J. A.en
dc.contributor.authorBhatele, A.en
dc.contributor.authorIsaacs, K. E.en
dc.contributor.authorGamblin, T.en
dc.contributor.authorSchulz, M.en
dc.contributor.authorLanger, S. H.en
dc.contributor.authorBremer, Peer-Timoen
dc.contributor.authorPascucci, V.en
dc.date.accessioned2016-02-28T06:44:25Zen
dc.date.available2016-02-28T06:44:25Zen
dc.date.issued2012-12en
dc.identifier.citationLandge AG, Levine JA, Bhatele A, Isaacs KE, Gamblin T, et al. (2012) Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations. IEEE Transactions on Visualization and Computer Graphics 18: 2467–2476. Available: http://dx.doi.org/10.1109/TVCG.2012.286.en
dc.identifier.issn1077-2626en
dc.identifier.pmid26357155en
dc.identifier.doi10.1109/TVCG.2012.286en
dc.identifier.urihttp://hdl.handle.net/10754/600175en
dc.description.abstractThe performance of massively parallel applications is often heavily impacted by the cost of communication among compute nodes. However, determining how to best use the network is a formidable task, made challenging by the ever increasing size and complexity of modern supercomputers. This paper applies visualization techniques to aid parallel application developers in understanding the network activity by enabling a detailed exploration of the flow of packets through the hardware interconnect. In order to visualize this large and complex data, we employ two linked views of the hardware network. The first is a 2D view, that represents the network structure as one of several simplified planar projections. This view is designed to allow a user to easily identify trends and patterns in the network traffic. The second is a 3D view that augments the 2D view by preserving the physical network topology and providing a context that is familiar to the application developers. Using the massively parallel multi-physics code pF3D as a case study, we demonstrate that our tool provides valuable insight that we use to explain and optimize pF3D-s performance on an IBM Blue Gene/P system. © 1995-2012 IEEE.en
dc.description.sponsorshipThis work is supported in part by NSF awards IIS-1045032, OCI-0904631, OCI-0906379 and CCF-0702817, and by a KAUST awardKUS-C1-016-04. This work was also performed under the auspicesof the U.S. Department of Energy by the University of Utah undercontracts DE-SC0001922, DE-AC52-07NA27344 and DE-FC02-06ER25781, and by Lawrence Livermore National Laboratory undercontract DE-AC52-07NA27344 (LLNL-CONF-543359).en
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)en
dc.subjectnetwork traffic visualizationen
dc.subjectPerformance analysisen
dc.subjectprojected graph layoutsen
dc.titleVisualizing Network Traffic to Understand the Performance of Massively Parallel Simulationsen
dc.typeArticleen
dc.identifier.journalIEEE Transactions on Visualization and Computer Graphicsen
dc.contributor.institutionUniversity of Utah, Salt Lake City, United Statesen
dc.contributor.institutionUC Davis, Davis, United Statesen
dc.contributor.institutionLawrence Livermore National Laboratory, Livermore, United Statesen
kaust.grant.numberKUS-C1-016-04en

Related articles on PubMed

All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.