Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration

Sai P. Chenna; Nalini Kumar; Leonardo Borges; Michael Steyer; Philippe Thierry; Maria Garzaran

doi:10.1007/978-3-031-40843-4_27

Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration

Sai P. Chenna, Nalini Kumar, Leonardo Borges, Michael Steyer, Philippe Thierry, Maria Garzaran

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

Abstract

The last decade has seen high-performance computing (HPC) systems become denser and denser. Higher node and rack density has led to development of multi-level networks - at socket, node, ‘pod’, rack, and between nodes. As sockets become more complex with integrated or co-packaged heterogeneous architectures, this network complexity is going to increase. In this paper, we extend Structural Simulation Toolkit (SST) to model these multi-level networks designs. We demonstrate this newly introduced capability by modeling a combination of a few different network topologies at different levels of the system and simulating the performance of collectives and some popular HPC communication patterns.

Original language	English (US)
Title of host publication	High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers
Editors	Amanda Bienz, Michèle Weiland, Marc Baboulin, Carola Kruse
Publisher	Springer
Pages	366-377
Number of pages	12
ISBN (Print)	9783031408427
DOIs	https://doi.org/10.1007/978-3-031-40843-4_27
State	Published - 2023
Externally published	Yes
Event	38th International Conference on High Performance Computing, ISC High Performance 2023 - Hamburg, Germany Duration: May 21 2023 → May 25 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	13999 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	38th International Conference on High Performance Computing, ISC High Performance 2023
Country/Territory	Germany
City	Hamburg
Period	5/21/23 → 5/25/23

Keywords

coarse-grained simulation
design space exploration
hierarchical network modeling
multi-level network modeling

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Online availability

10.1007/978-3-031-40843-4_27

Library availability

Discover UIUC Full Text

Cite this

Chenna, S. P., Kumar, N., Borges, L., Steyer, M., Thierry, P., & Garzaran, M. (2023). Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration. In A. Bienz, M. Weiland, M. Baboulin, & C. Kruse (Eds.), High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers (pp. 366-377). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13999 LNCS). Springer. https://doi.org/10.1007/978-3-031-40843-4_27

Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration. / Chenna, Sai P.; Kumar, Nalini; Borges, Leonardo et al.
High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers. ed. / Amanda Bienz; Michèle Weiland; Marc Baboulin; Carola Kruse. Springer, 2023. p. 366-377 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 13999 LNCS).

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

Chenna, SP, Kumar, N, Borges, L, Steyer, M, Thierry, P & Garzaran, M 2023, Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration. in A Bienz, M Weiland, M Baboulin & C Kruse (eds), High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 13999 LNCS, Springer, pp. 366-377, 38th International Conference on High Performance Computing, ISC High Performance 2023, Hamburg, Germany, 5/21/23. https://doi.org/10.1007/978-3-031-40843-4_27

Chenna SP, Kumar N, Borges L, Steyer M, Thierry P, Garzaran M. Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration. In Bienz A, Weiland M, Baboulin M, Kruse C, editors, High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers. Springer. 2023. p. 366-377. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-40843-4_27

Chenna, Sai P. ; Kumar, Nalini ; Borges, Leonardo et al. / Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration. High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers. editor / Amanda Bienz ; Michèle Weiland ; Marc Baboulin ; Carola Kruse. Springer, 2023. pp. 366-377 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{155da2bdc1a947dbb76a202de24a2ef0,

title = "Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration",

abstract = "The last decade has seen high-performance computing (HPC) systems become denser and denser. Higher node and rack density has led to development of multi-level networks - at socket, node, {\textquoteleft}pod{\textquoteright}, rack, and between nodes. As sockets become more complex with integrated or co-packaged heterogeneous architectures, this network complexity is going to increase. In this paper, we extend Structural Simulation Toolkit (SST) to model these multi-level networks designs. We demonstrate this newly introduced capability by modeling a combination of a few different network topologies at different levels of the system and simulating the performance of collectives and some popular HPC communication patterns.",

keywords = "coarse-grained simulation, design space exploration, hierarchical network modeling, multi-level network modeling",

author = "Chenna, {Sai P.} and Nalini Kumar and Leonardo Borges and Michael Steyer and Philippe Thierry and Maria Garzaran",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 38th International Conference on High Performance Computing, ISC High Performance 2023 ; Conference date: 21-05-2023 Through 25-05-2023",

year = "2023",

doi = "10.1007/978-3-031-40843-4_27",

language = "English (US)",

isbn = "9783031408427",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "366--377",

editor = "Amanda Bienz and Mich{\`e}le Weiland and Marc Baboulin and Carola Kruse",

booktitle = "High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers",

address = "Germany",

}

TY - GEN

T1 - Enabling Multi-level Network Modeling in Structural Simulation Toolkit for Next-Generation HPC Network Design Space Exploration

AU - Chenna, Sai P.

AU - Kumar, Nalini

AU - Borges, Leonardo

AU - Steyer, Michael

AU - Thierry, Philippe

AU - Garzaran, Maria

PY - 2023

Y1 - 2023

N2 - The last decade has seen high-performance computing (HPC) systems become denser and denser. Higher node and rack density has led to development of multi-level networks - at socket, node, ‘pod’, rack, and between nodes. As sockets become more complex with integrated or co-packaged heterogeneous architectures, this network complexity is going to increase. In this paper, we extend Structural Simulation Toolkit (SST) to model these multi-level networks designs. We demonstrate this newly introduced capability by modeling a combination of a few different network topologies at different levels of the system and simulating the performance of collectives and some popular HPC communication patterns.

AB - The last decade has seen high-performance computing (HPC) systems become denser and denser. Higher node and rack density has led to development of multi-level networks - at socket, node, ‘pod’, rack, and between nodes. As sockets become more complex with integrated or co-packaged heterogeneous architectures, this network complexity is going to increase. In this paper, we extend Structural Simulation Toolkit (SST) to model these multi-level networks designs. We demonstrate this newly introduced capability by modeling a combination of a few different network topologies at different levels of the system and simulating the performance of collectives and some popular HPC communication patterns.

KW - coarse-grained simulation

KW - design space exploration

KW - hierarchical network modeling

KW - multi-level network modeling

UR - http://www.scopus.com/inward/record.url?scp=85171368660&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85171368660&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-40843-4_27

DO - 10.1007/978-3-031-40843-4_27

M3 - Conference contribution

AN - SCOPUS:85171368660

SN - 9783031408427

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 366

EP - 377

BT - High Performance Computing - ISC High Performance 2023 International Workshops, Revised Selected Papers

A2 - Bienz, Amanda

A2 - Weiland, Michèle

A2 - Baboulin, Marc

A2 - Kruse, Carola

PB - Springer

T2 - 38th International Conference on High Performance Computing, ISC High Performance 2023

Y2 - 21 May 2023 through 25 May 2023

ER -