HPRA: A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips

Elena Kakoulli, Vassos Soteriou Soteriou, Theocharis Theocharides

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The inherent spatio-temporal unevenness of traffic flows in Networks-on-Chips (NoCs) can cause unforeseen, and in cases, severe forms of congestion, known as hotspots. Hotspots reduce the NoC's effective throughput, where in the worst case scenario, the entire network can be brought to an unrecoverable halt as a hotspot(s) spreads across the topology. To alleviate this problematic phenomenon several adaptive routing algorithms employ online load-balancing functions, aiming to reduce the possibility of hotspots arising. Most, however, work passively, merely distributing traffic as evenly as possible among alternative network paths, and they cannot guarantee the absence of network congestion as their reactive capability in reducing hotspot formation(s) is limited. In this paper we present a new pro-active Hotspot-Preventive Routing Algorithm (HPRA) which uses the advance knowledge gained from network-embedded Artificial Neural Network-based (ANN) hotspot predictors to guide packet routing across the network in an effort to mitigate any unforeseen near-future occurrences of hotspots. These ANNs are trained offline and during multicore operation they gather online buffer utilization data to predict about-to-be-formed hotspots, promptly informing the HPRA routing algorithm to take appropriate action in preventing hotspot formation(s). Evaluation results across two synthetic traffic patterns, and traffic benchmarks gathered from a chip multiprocessor architecture, show that HPRA can reduce network latency and improve network throughput up to 81% when compared against several existing state-of-the-art congestion-aware routing functions. Hardware synthesis results demonstrate the efficacy of the HPRA mechanism.

Original languageEnglish (US)
Title of host publication2012 IEEE 30th International Conference on Computer Design, ICCD 2012
Pages249-255
Number of pages7
DOIs
StatePublished - Dec 1 2012
Event2012 IEEE 30th International Conference on Computer Design, ICCD 2012 - Montreal, QC, Canada
Duration: Sep 30 2012Oct 3 2012

Other

Other2012 IEEE 30th International Conference on Computer Design, ICCD 2012
CountryCanada
CityMontreal, QC
Period9/30/1210/3/12

Fingerprint

Routing algorithms
Throughput
Adaptive algorithms
Resource allocation
Network-on-chip
Topology
Neural networks
Hardware

ASJC Scopus subject areas

  • Hardware and Architecture
  • Electrical and Electronic Engineering

Cite this

Kakoulli, E., Soteriou, V. S., & Theocharides, T. (2012). HPRA: A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips. In 2012 IEEE 30th International Conference on Computer Design, ICCD 2012 (pp. 249-255). [6378648] https://doi.org/10.1109/ICCD.2012.6378648

HPRA : A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips. / Kakoulli, Elena; Soteriou, Vassos Soteriou; Theocharides, Theocharis.

2012 IEEE 30th International Conference on Computer Design, ICCD 2012. 2012. p. 249-255 6378648.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kakoulli, E, Soteriou, VS & Theocharides, T 2012, HPRA: A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips. in 2012 IEEE 30th International Conference on Computer Design, ICCD 2012., 6378648, pp. 249-255, 2012 IEEE 30th International Conference on Computer Design, ICCD 2012, Montreal, QC, Canada, 9/30/12. https://doi.org/10.1109/ICCD.2012.6378648
Kakoulli E, Soteriou VS, Theocharides T. HPRA: A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips. In 2012 IEEE 30th International Conference on Computer Design, ICCD 2012. 2012. p. 249-255. 6378648 https://doi.org/10.1109/ICCD.2012.6378648
Kakoulli, Elena ; Soteriou, Vassos Soteriou ; Theocharides, Theocharis. / HPRA : A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips. 2012 IEEE 30th International Conference on Computer Design, ICCD 2012. 2012. pp. 249-255
@inproceedings{26eb6f8467c94448b805f679a15ac3f0,
title = "HPRA: A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips",
abstract = "The inherent spatio-temporal unevenness of traffic flows in Networks-on-Chips (NoCs) can cause unforeseen, and in cases, severe forms of congestion, known as hotspots. Hotspots reduce the NoC's effective throughput, where in the worst case scenario, the entire network can be brought to an unrecoverable halt as a hotspot(s) spreads across the topology. To alleviate this problematic phenomenon several adaptive routing algorithms employ online load-balancing functions, aiming to reduce the possibility of hotspots arising. Most, however, work passively, merely distributing traffic as evenly as possible among alternative network paths, and they cannot guarantee the absence of network congestion as their reactive capability in reducing hotspot formation(s) is limited. In this paper we present a new pro-active Hotspot-Preventive Routing Algorithm (HPRA) which uses the advance knowledge gained from network-embedded Artificial Neural Network-based (ANN) hotspot predictors to guide packet routing across the network in an effort to mitigate any unforeseen near-future occurrences of hotspots. These ANNs are trained offline and during multicore operation they gather online buffer utilization data to predict about-to-be-formed hotspots, promptly informing the HPRA routing algorithm to take appropriate action in preventing hotspot formation(s). Evaluation results across two synthetic traffic patterns, and traffic benchmarks gathered from a chip multiprocessor architecture, show that HPRA can reduce network latency and improve network throughput up to 81{\%} when compared against several existing state-of-the-art congestion-aware routing functions. Hardware synthesis results demonstrate the efficacy of the HPRA mechanism.",
author = "Elena Kakoulli and Soteriou, {Vassos Soteriou} and Theocharis Theocharides",
year = "2012",
month = "12",
day = "1",
doi = "10.1109/ICCD.2012.6378648",
language = "English (US)",
isbn = "9781467330503",
pages = "249--255",
booktitle = "2012 IEEE 30th International Conference on Computer Design, ICCD 2012",

}

TY - GEN

T1 - HPRA

T2 - A pro-active hotspot-preventive high-performance routing algorithm for Networks-on-Chips

AU - Kakoulli, Elena

AU - Soteriou, Vassos Soteriou

AU - Theocharides, Theocharis

PY - 2012/12/1

Y1 - 2012/12/1

N2 - The inherent spatio-temporal unevenness of traffic flows in Networks-on-Chips (NoCs) can cause unforeseen, and in cases, severe forms of congestion, known as hotspots. Hotspots reduce the NoC's effective throughput, where in the worst case scenario, the entire network can be brought to an unrecoverable halt as a hotspot(s) spreads across the topology. To alleviate this problematic phenomenon several adaptive routing algorithms employ online load-balancing functions, aiming to reduce the possibility of hotspots arising. Most, however, work passively, merely distributing traffic as evenly as possible among alternative network paths, and they cannot guarantee the absence of network congestion as their reactive capability in reducing hotspot formation(s) is limited. In this paper we present a new pro-active Hotspot-Preventive Routing Algorithm (HPRA) which uses the advance knowledge gained from network-embedded Artificial Neural Network-based (ANN) hotspot predictors to guide packet routing across the network in an effort to mitigate any unforeseen near-future occurrences of hotspots. These ANNs are trained offline and during multicore operation they gather online buffer utilization data to predict about-to-be-formed hotspots, promptly informing the HPRA routing algorithm to take appropriate action in preventing hotspot formation(s). Evaluation results across two synthetic traffic patterns, and traffic benchmarks gathered from a chip multiprocessor architecture, show that HPRA can reduce network latency and improve network throughput up to 81% when compared against several existing state-of-the-art congestion-aware routing functions. Hardware synthesis results demonstrate the efficacy of the HPRA mechanism.

AB - The inherent spatio-temporal unevenness of traffic flows in Networks-on-Chips (NoCs) can cause unforeseen, and in cases, severe forms of congestion, known as hotspots. Hotspots reduce the NoC's effective throughput, where in the worst case scenario, the entire network can be brought to an unrecoverable halt as a hotspot(s) spreads across the topology. To alleviate this problematic phenomenon several adaptive routing algorithms employ online load-balancing functions, aiming to reduce the possibility of hotspots arising. Most, however, work passively, merely distributing traffic as evenly as possible among alternative network paths, and they cannot guarantee the absence of network congestion as their reactive capability in reducing hotspot formation(s) is limited. In this paper we present a new pro-active Hotspot-Preventive Routing Algorithm (HPRA) which uses the advance knowledge gained from network-embedded Artificial Neural Network-based (ANN) hotspot predictors to guide packet routing across the network in an effort to mitigate any unforeseen near-future occurrences of hotspots. These ANNs are trained offline and during multicore operation they gather online buffer utilization data to predict about-to-be-formed hotspots, promptly informing the HPRA routing algorithm to take appropriate action in preventing hotspot formation(s). Evaluation results across two synthetic traffic patterns, and traffic benchmarks gathered from a chip multiprocessor architecture, show that HPRA can reduce network latency and improve network throughput up to 81% when compared against several existing state-of-the-art congestion-aware routing functions. Hardware synthesis results demonstrate the efficacy of the HPRA mechanism.

UR - http://www.scopus.com/inward/record.url?scp=84872081537&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872081537&partnerID=8YFLogxK

U2 - 10.1109/ICCD.2012.6378648

DO - 10.1109/ICCD.2012.6378648

M3 - Conference contribution

SN - 9781467330503

SP - 249

EP - 255

BT - 2012 IEEE 30th International Conference on Computer Design, ICCD 2012

ER -