A Performance Model of Software-Based Deadlock Recovery Routing Algorithm in Hypercubes

A. Khonsari, H. Sarbazi-Azad, M. Ould-Khaoua

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Recent studies have revealed that deadlocks are generally infrequent in the network. Thus the hardware resources, e.g. virtual channels, dedicated for deadlock avoidance are not utilised most of the time. This consideration has motivated the development of novel adaptive routing algorithms with deadlock recovery. This paper describes a new analytical model to predict message latency in hypercubes with a true fully adaptive routing algorithm with progressive deadlock recovery. One of the main features of the proposed model is the use of results from queueing systems with impatient customers to capture the effects of the timeout mechanism used in this routing algorithm for deadlock detection. The validity of the model is demonstrated by comparing analytical results with those obtained through simulation experiments.

Original languageEnglish
Pages (from-to)153-168
Number of pages16
JournalParallel Processing Letters
Volume15
Issue number1-2
DOIs
Publication statusPublished - Mar 2005

Fingerprint

Deadlock
Performance Model
Routing algorithms
Routing Algorithm
Hypercube
Recovery
Adaptive algorithms
Adaptive Routing
Software
Adaptive Algorithm
Deadlock Avoidance
Virtual Channel
Analytical models
Queueing System
Analytical Model
Simulation Experiment
Hardware
Latency
Customers
Predict

Keywords

  • M/G/1 queues
  • Performance modelling
  • Wormhole switching

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

A Performance Model of Software-Based Deadlock Recovery Routing Algorithm in Hypercubes. / Khonsari, A.; Sarbazi-Azad, H.; Ould-Khaoua, M.

In: Parallel Processing Letters, Vol. 15, No. 1-2, 03.2005, p. 153-168.

Research output: Contribution to journalArticle

@article{62b1e09117794d5c96e9e69cc84ef714,
title = "A Performance Model of Software-Based Deadlock Recovery Routing Algorithm in Hypercubes",
abstract = "Recent studies have revealed that deadlocks are generally infrequent in the network. Thus the hardware resources, e.g. virtual channels, dedicated for deadlock avoidance are not utilised most of the time. This consideration has motivated the development of novel adaptive routing algorithms with deadlock recovery. This paper describes a new analytical model to predict message latency in hypercubes with a true fully adaptive routing algorithm with progressive deadlock recovery. One of the main features of the proposed model is the use of results from queueing systems with impatient customers to capture the effects of the timeout mechanism used in this routing algorithm for deadlock detection. The validity of the model is demonstrated by comparing analytical results with those obtained through simulation experiments.",
keywords = "M/G/1 queues, Performance modelling, Wormhole switching",
author = "A. Khonsari and H. Sarbazi-Azad and M. Ould-Khaoua",
year = "2005",
month = "3",
doi = "10.1142/S012962640500212X",
language = "English",
volume = "15",
pages = "153--168",
journal = "Parallel Processing Letters",
issn = "0129-6264",
publisher = "World Scientific Publishing Co. Pte Ltd",
number = "1-2",

}

TY - JOUR

T1 - A Performance Model of Software-Based Deadlock Recovery Routing Algorithm in Hypercubes

AU - Khonsari, A.

AU - Sarbazi-Azad, H.

AU - Ould-Khaoua, M.

PY - 2005/3

Y1 - 2005/3

N2 - Recent studies have revealed that deadlocks are generally infrequent in the network. Thus the hardware resources, e.g. virtual channels, dedicated for deadlock avoidance are not utilised most of the time. This consideration has motivated the development of novel adaptive routing algorithms with deadlock recovery. This paper describes a new analytical model to predict message latency in hypercubes with a true fully adaptive routing algorithm with progressive deadlock recovery. One of the main features of the proposed model is the use of results from queueing systems with impatient customers to capture the effects of the timeout mechanism used in this routing algorithm for deadlock detection. The validity of the model is demonstrated by comparing analytical results with those obtained through simulation experiments.

AB - Recent studies have revealed that deadlocks are generally infrequent in the network. Thus the hardware resources, e.g. virtual channels, dedicated for deadlock avoidance are not utilised most of the time. This consideration has motivated the development of novel adaptive routing algorithms with deadlock recovery. This paper describes a new analytical model to predict message latency in hypercubes with a true fully adaptive routing algorithm with progressive deadlock recovery. One of the main features of the proposed model is the use of results from queueing systems with impatient customers to capture the effects of the timeout mechanism used in this routing algorithm for deadlock detection. The validity of the model is demonstrated by comparing analytical results with those obtained through simulation experiments.

KW - M/G/1 queues

KW - Performance modelling

KW - Wormhole switching

UR - http://www.scopus.com/inward/record.url?scp=23844472661&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23844472661&partnerID=8YFLogxK

U2 - 10.1142/S012962640500212X

DO - 10.1142/S012962640500212X

M3 - Article

VL - 15

SP - 153

EP - 168

JO - Parallel Processing Letters

JF - Parallel Processing Letters

SN - 0129-6264

IS - 1-2

ER -