Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations

A. G. Landge; J. A. Levine; A. Bhatele; K. E. Isaacs; T. Gamblin; M. Schulz; S. H. Langer; Peer-Timo Bremer; V. Pascucci

doi:10.1109/TVCG.2012.286

Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations

A. G. Landge, J. A. Levine, A. Bhatele, K. E. Isaacs, T. Gamblin, M. Schulz, S. H. Langer, Peer-Timo Bremer, V. Pascucci

Research output: Contribution to journal › Article › peer-review

51 Scopus citations

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.

Original language	English (US)
Pages (from-to)	2467-2476
Number of pages	10
Journal	IEEE Transactions on Visualization and Computer Graphics
Volume	18
Issue number	12
DOIs	https://doi.org/10.1109/TVCG.2012.286
State	Published - Dec 2012
Externally published	Yes

Bibliographical note

KAUST Repository Item: Exported on 2020-10-01
Acknowledged KAUST grant number(s): KUS-C1-016-04
Acknowledgements: 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).
This publication acknowledges KAUST support, but has no KAUST affiliated authors.

Access to Document

10.1109/TVCG.2012.286

Cite this

@article{eeee820365c64878835e6999b0a5a271,

title = "Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations",

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. {\textcopyright} 1995-2012 IEEE.",

author = "Landge, {A. G.} and Levine, {J. A.} and A. Bhatele and Isaacs, {K. E.} and T. Gamblin and M. Schulz and Langer, {S. H.} and Peer-Timo Bremer and V. Pascucci",

note = "KAUST Repository Item: Exported on 2020-10-01 Acknowledged KAUST grant number(s): KUS-C1-016-04 Acknowledgements: 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). This publication acknowledges KAUST support, but has no KAUST affiliated authors.",

year = "2012",

month = dec,

doi = "10.1109/TVCG.2012.286",

language = "English (US)",

volume = "18",

pages = "2467--2476",

journal = "IEEE Transactions on Visualization and Computer Graphics",

issn = "1077-2626",

publisher = "IEEE Computer Society",

number = "12",

}

TY - JOUR

T1 - Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations

AU - Landge, A. G.

AU - Levine, J. A.

AU - Bhatele, A.

AU - Isaacs, K. E.

AU - Gamblin, T.

AU - Schulz, M.

AU - Langer, S. H.

AU - Bremer, Peer-Timo

AU - Pascucci, V.

N1 - KAUST Repository Item: Exported on 2020-10-01 Acknowledged KAUST grant number(s): KUS-C1-016-04 Acknowledgements: 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). This publication acknowledges KAUST support, but has no KAUST affiliated authors.

PY - 2012/12

Y1 - 2012/12

N2 - 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.

AB - 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.

UR - http://hdl.handle.net/10754/600175

UR - http://ieeexplore.ieee.org/document/6327252/

UR - http://www.scopus.com/inward/record.url?scp=84867660665&partnerID=8YFLogxK

U2 - 10.1109/TVCG.2012.286

DO - 10.1109/TVCG.2012.286

M3 - Article

C2 - 26357155

SN - 1077-2626

VL - 18

SP - 2467

EP - 2476

JO - IEEE Transactions on Visualization and Computer Graphics

JF - IEEE Transactions on Visualization and Computer Graphics

IS - 12

ER -

Visualizing Network Traffic to Understand the Performance of Massively Parallel Simulations

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this