Designing an Exascale Interconnect using Multi-objective OptimizationCitation formats

  • External authors:
  • Jose Pascual Saiz
  • Joshua Lant
  • Andrew Attwood
  • Caroline Concatto
  • Javier Navaridas
  • Mikel Luján

Standard

Designing an Exascale Interconnect using Multi-objective Optimization. / Pascual Saiz, Jose; Lant, Joshua; Attwood, Andrew; Concatto, Caroline; Navaridas, Javier; Luján, Mikel; Goodacre, John.

IEEE Congress on Evolutionary Computation 2017. IEEE, 2017.

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Harvard

Pascual Saiz, J, Lant, J, Attwood, A, Concatto, C, Navaridas, J, Luján, M & Goodacre, J 2017, Designing an Exascale Interconnect using Multi-objective Optimization. in IEEE Congress on Evolutionary Computation 2017. IEEE, 2017 IEEE Congress on Evolutionary Computation, San Sebastian, Spain, 5/06/17. https://doi.org/10.1109/CEC.2017.7969572

APA

Pascual Saiz, J., Lant, J., Attwood, A., Concatto, C., Navaridas, J., Luján, M., & Goodacre, J. (2017). Designing an Exascale Interconnect using Multi-objective Optimization. In IEEE Congress on Evolutionary Computation 2017 IEEE. https://doi.org/10.1109/CEC.2017.7969572

Vancouver

Pascual Saiz J, Lant J, Attwood A, Concatto C, Navaridas J, Luján M et al. Designing an Exascale Interconnect using Multi-objective Optimization. In IEEE Congress on Evolutionary Computation 2017. IEEE. 2017 https://doi.org/10.1109/CEC.2017.7969572

Author

Pascual Saiz, Jose ; Lant, Joshua ; Attwood, Andrew ; Concatto, Caroline ; Navaridas, Javier ; Luján, Mikel ; Goodacre, John. / Designing an Exascale Interconnect using Multi-objective Optimization. IEEE Congress on Evolutionary Computation 2017. IEEE, 2017.

Bibtex

@inproceedings{2419a3a993f643a0ad2df4d61b1a281d,
title = "Designing an Exascale Interconnect using Multi-objective Optimization",
abstract = "Exascale performance will be delivered by systems composed of millions of interconnected computing cores. The way these computing elements are connected with each other (network topology) has a strong impact on many performance characteristics. In this work we propose a multi-objective optimizationbased framework to explore possible network topologies to be implemented in the EU-funded ExaNeSt project. The modular design of this system{\textquoteright}s interconnect provides great flexibility to design topologies optimized for specific performance targets such as communications locality, fault tolerance or energyconsumption. The generation procedure of the topologies is formulated as a three-objective optimization problem (minimizing some topological characteristics) where solutions are searched using evolutionary techniques. The analysis of the results, carried out using simulation, shows that the topologies meet the required performance objectives. In addition, a comparison with a wellknown topology reveals that the generated solutions can provide better topological characteristics and also higher performancefor parallel applications.",
author = "{Pascual Saiz}, Jose and Joshua Lant and Andrew Attwood and Caroline Concatto and Javier Navaridas and Mikel Luj{\'a}n and John Goodacre",
year = "2017",
doi = "10.1109/CEC.2017.7969572",
language = "English",
booktitle = "IEEE Congress on Evolutionary Computation 2017",
publisher = "IEEE",
address = "United States",
note = "2017 IEEE Congress on Evolutionary Computation, CEC 2017 ; Conference date: 05-06-2017 Through 08-06-2017",
url = "http://www.cec2017.org/",

}

RIS

TY - GEN

T1 - Designing an Exascale Interconnect using Multi-objective Optimization

AU - Pascual Saiz, Jose

AU - Lant, Joshua

AU - Attwood, Andrew

AU - Concatto, Caroline

AU - Navaridas, Javier

AU - Luján, Mikel

AU - Goodacre, John

PY - 2017

Y1 - 2017

N2 - Exascale performance will be delivered by systems composed of millions of interconnected computing cores. The way these computing elements are connected with each other (network topology) has a strong impact on many performance characteristics. In this work we propose a multi-objective optimizationbased framework to explore possible network topologies to be implemented in the EU-funded ExaNeSt project. The modular design of this system’s interconnect provides great flexibility to design topologies optimized for specific performance targets such as communications locality, fault tolerance or energyconsumption. The generation procedure of the topologies is formulated as a three-objective optimization problem (minimizing some topological characteristics) where solutions are searched using evolutionary techniques. The analysis of the results, carried out using simulation, shows that the topologies meet the required performance objectives. In addition, a comparison with a wellknown topology reveals that the generated solutions can provide better topological characteristics and also higher performancefor parallel applications.

AB - Exascale performance will be delivered by systems composed of millions of interconnected computing cores. The way these computing elements are connected with each other (network topology) has a strong impact on many performance characteristics. In this work we propose a multi-objective optimizationbased framework to explore possible network topologies to be implemented in the EU-funded ExaNeSt project. The modular design of this system’s interconnect provides great flexibility to design topologies optimized for specific performance targets such as communications locality, fault tolerance or energyconsumption. The generation procedure of the topologies is formulated as a three-objective optimization problem (minimizing some topological characteristics) where solutions are searched using evolutionary techniques. The analysis of the results, carried out using simulation, shows that the topologies meet the required performance objectives. In addition, a comparison with a wellknown topology reveals that the generated solutions can provide better topological characteristics and also higher performancefor parallel applications.

U2 - 10.1109/CEC.2017.7969572

DO - 10.1109/CEC.2017.7969572

M3 - Conference contribution

BT - IEEE Congress on Evolutionary Computation 2017

PB - IEEE

T2 - 2017 IEEE Congress on Evolutionary Computation

Y2 - 5 June 2017 through 8 June 2017

ER -