Scaling an optimistic parallel simulation of large-scale interconnection networks

Nilesh Choudhury; Yogesh Mehta; Terry L. Wilmarth; Eric J. Bohm; Laxmikant V Kale

doi:10.1109/WSC.2005.1574299

Scaling an optimistic parallel simulation of large-scale interconnection networks

Nilesh Choudhury, Yogesh Mehta, Terry L. Wilmarth, Eric J. Bohm, Laxmikant V Kale

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Parallel computers today are designed with larger number of processors than ever before, connected by large scale Interconnection Networks. Communication is the key to achieving high performance on such machines, making the study of Interconnection Networks important. Parallel simulations of Interconnection Networks present a unique problem characterized by fine-grained computation and strong dependence among events. The absence of large lookaheads makes it unsuitable to use a conservative simulation. Using an optimistic Parallel Discrete Event Simulation allows us to extract reasonable parallelism from this simulation. In this paper we present BigNetSim, an Interconnection Network simulator. We analyze its performance and present techniques related to enhancing performance and scaling it to a large number of processors on different artificial traffic patterns and real application logs. Inspite of the overheads of a parallel optimistic simulation, we have achieved a breakeven with sequential simulation at four processors and demonstrate perfect scaling to 128 processors.

Original language	English (US)
Title of host publication	Proceedings of the 2005 Winter Simulation Conference
Pages	591-600
Number of pages	10
DOIs	https://doi.org/10.1109/WSC.2005.1574299
State	Published - Dec 1 2005
Event	2005 Winter Simulation Conference - Orlando, FL, United States Duration: Dec 4 2005 → Dec 7 2005

Publication series

Name	Proceedings - Winter Simulation Conference
Volume	2005
ISSN (Print)	0891-7736

Other

Other	2005 Winter Simulation Conference
Country	United States
City	Orlando, FL
Period	12/4/05 → 12/7/05

ASJC Scopus subject areas

Software
Modeling and Simulation
Computer Science Applications

Access to Document

10.1109/WSC.2005.1574299

Fingerprint Dive into the research topics of 'Scaling an optimistic parallel simulation of large-scale interconnection networks'. Together they form a unique fingerprint.

Cite this

Choudhury, N., Mehta, Y., Wilmarth, T. L., Bohm, E. J., & Kale, L. V. (2005). Scaling an optimistic parallel simulation of large-scale interconnection networks. In Proceedings of the 2005 Winter Simulation Conference (pp. 591-600). [1574299] (Proceedings - Winter Simulation Conference; Vol. 2005). https://doi.org/10.1109/WSC.2005.1574299

Scaling an optimistic parallel simulation of large-scale interconnection networks. / Choudhury, Nilesh; Mehta, Yogesh; Wilmarth, Terry L.; Bohm, Eric J.; Kale, Laxmikant V.

Proceedings of the 2005 Winter Simulation Conference. 2005. p. 591-600 1574299 (Proceedings - Winter Simulation Conference; Vol. 2005).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Choudhury, N, Mehta, Y, Wilmarth, TL, Bohm, EJ & Kale, LV 2005, Scaling an optimistic parallel simulation of large-scale interconnection networks. in Proceedings of the 2005 Winter Simulation Conference., 1574299, Proceedings - Winter Simulation Conference, vol. 2005, pp. 591-600, 2005 Winter Simulation Conference, Orlando, FL, United States, 12/4/05. https://doi.org/10.1109/WSC.2005.1574299

@inproceedings{61851c1ffad24067b4a8b4ea1b7f2a89,

title = "Scaling an optimistic parallel simulation of large-scale interconnection networks",

abstract = "Parallel computers today are designed with larger number of processors than ever before, connected by large scale Interconnection Networks. Communication is the key to achieving high performance on such machines, making the study of Interconnection Networks important. Parallel simulations of Interconnection Networks present a unique problem characterized by fine-grained computation and strong dependence among events. The absence of large lookaheads makes it unsuitable to use a conservative simulation. Using an optimistic Parallel Discrete Event Simulation allows us to extract reasonable parallelism from this simulation. In this paper we present BigNetSim, an Interconnection Network simulator. We analyze its performance and present techniques related to enhancing performance and scaling it to a large number of processors on different artificial traffic patterns and real application logs. Inspite of the overheads of a parallel optimistic simulation, we have achieved a breakeven with sequential simulation at four processors and demonstrate perfect scaling to 128 processors.",

author = "Nilesh Choudhury and Yogesh Mehta and Wilmarth, {Terry L.} and Bohm, {Eric J.} and Kale, {Laxmikant V}",

year = "2005",

month = dec,

day = "1",

doi = "10.1109/WSC.2005.1574299",

language = "English (US)",

isbn = "0780395204",

series = "Proceedings - Winter Simulation Conference",

pages = "591--600",

booktitle = "Proceedings of the 2005 Winter Simulation Conference",

note = "2005 Winter Simulation Conference ; Conference date: 04-12-2005 Through 07-12-2005",

}

TY - GEN

T1 - Scaling an optimistic parallel simulation of large-scale interconnection networks

AU - Choudhury, Nilesh

AU - Mehta, Yogesh

AU - Wilmarth, Terry L.

AU - Bohm, Eric J.

AU - Kale, Laxmikant V

PY - 2005/12/1

Y1 - 2005/12/1

N2 - Parallel computers today are designed with larger number of processors than ever before, connected by large scale Interconnection Networks. Communication is the key to achieving high performance on such machines, making the study of Interconnection Networks important. Parallel simulations of Interconnection Networks present a unique problem characterized by fine-grained computation and strong dependence among events. The absence of large lookaheads makes it unsuitable to use a conservative simulation. Using an optimistic Parallel Discrete Event Simulation allows us to extract reasonable parallelism from this simulation. In this paper we present BigNetSim, an Interconnection Network simulator. We analyze its performance and present techniques related to enhancing performance and scaling it to a large number of processors on different artificial traffic patterns and real application logs. Inspite of the overheads of a parallel optimistic simulation, we have achieved a breakeven with sequential simulation at four processors and demonstrate perfect scaling to 128 processors.

AB - Parallel computers today are designed with larger number of processors than ever before, connected by large scale Interconnection Networks. Communication is the key to achieving high performance on such machines, making the study of Interconnection Networks important. Parallel simulations of Interconnection Networks present a unique problem characterized by fine-grained computation and strong dependence among events. The absence of large lookaheads makes it unsuitable to use a conservative simulation. Using an optimistic Parallel Discrete Event Simulation allows us to extract reasonable parallelism from this simulation. In this paper we present BigNetSim, an Interconnection Network simulator. We analyze its performance and present techniques related to enhancing performance and scaling it to a large number of processors on different artificial traffic patterns and real application logs. Inspite of the overheads of a parallel optimistic simulation, we have achieved a breakeven with sequential simulation at four processors and demonstrate perfect scaling to 128 processors.

UR - http://www.scopus.com/inward/record.url?scp=33846684644&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33846684644&partnerID=8YFLogxK

U2 - 10.1109/WSC.2005.1574299

DO - 10.1109/WSC.2005.1574299

M3 - Conference contribution

AN - SCOPUS:33846684644

SN - 0780395204

SN - 9780780395206

T3 - Proceedings - Winter Simulation Conference

SP - 591

EP - 600

BT - Proceedings of the 2005 Winter Simulation Conference

T2 - 2005 Winter Simulation Conference

Y2 - 4 December 2005 through 7 December 2005

ER -