A Soft Error Tolerant Network-on-Chip Router Pipeline for Multi-Core Systems

Pavan Poluri, Ahmed Louri

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Network-on-Chip (NoC) paradigm is rapidly evolving into an efficient interconnection network to handle the strict communication requirements between the increasing number of cores on a single chip. Diminishing transistor size is making the NoC increasingly vulnerable to both hard faults and soft errors. This paper concentrates on soft errors in NoCs. A soft error in an NoC router results in significant consequences such as data corruption, packet retransmission and deadlock among others. To this end, we propose Soft Error T olerant NoC Router (STNR) architecture, that is capable of detecting and recovering from soft errors occurring in different control stages of the routing pipeline. STNR exploits the use of idle cycles inherent in NoC packet routing pipeline to perform time redundant executions necessary for soft error tolerance. In doing so, STNR is able to detect and correct all single transient faults in the control stages of the pipeline. Simulation results using PARSEC and SPLASH-2 benchmarks show that STNR is able to accomplish such high level of soft error protection with a minimal impact on latency (an increase of 1.7 and 1.6 percent respectively). Additionally, STNR incurs an area overhead of 7 percent and power overhead of 13 percent as compared to the baseline unprotected router.

Original languageEnglish (US)
Article number6912980
Pages (from-to)107-110
Number of pages4
JournalIEEE Computer Architecture Letters
Volume14
Issue number2
DOIs
StatePublished - Jul 1 2015

Fingerprint

Routers
Pipelines
Network-on-chip
Transistors
Communication

Keywords

  • Network-on-chip
  • performance
  • reliability
  • soft error

ASJC Scopus subject areas

  • Hardware and Architecture

Cite this

A Soft Error Tolerant Network-on-Chip Router Pipeline for Multi-Core Systems. / Poluri, Pavan; Louri, Ahmed.

In: IEEE Computer Architecture Letters, Vol. 14, No. 2, 6912980, 01.07.2015, p. 107-110.

Research output: Contribution to journalArticle

@article{9cdf0a94195846a596601580cfd2af94,
title = "A Soft Error Tolerant Network-on-Chip Router Pipeline for Multi-Core Systems",
abstract = "Network-on-Chip (NoC) paradigm is rapidly evolving into an efficient interconnection network to handle the strict communication requirements between the increasing number of cores on a single chip. Diminishing transistor size is making the NoC increasingly vulnerable to both hard faults and soft errors. This paper concentrates on soft errors in NoCs. A soft error in an NoC router results in significant consequences such as data corruption, packet retransmission and deadlock among others. To this end, we propose Soft Error T olerant NoC Router (STNR) architecture, that is capable of detecting and recovering from soft errors occurring in different control stages of the routing pipeline. STNR exploits the use of idle cycles inherent in NoC packet routing pipeline to perform time redundant executions necessary for soft error tolerance. In doing so, STNR is able to detect and correct all single transient faults in the control stages of the pipeline. Simulation results using PARSEC and SPLASH-2 benchmarks show that STNR is able to accomplish such high level of soft error protection with a minimal impact on latency (an increase of 1.7 and 1.6 percent respectively). Additionally, STNR incurs an area overhead of 7 percent and power overhead of 13 percent as compared to the baseline unprotected router.",
keywords = "Network-on-chip, performance, reliability, soft error",
author = "Pavan Poluri and Ahmed Louri",
year = "2015",
month = "7",
day = "1",
doi = "10.1109/LCA.2014.2360686",
language = "English (US)",
volume = "14",
pages = "107--110",
journal = "IEEE Computer Architecture Letters",
issn = "1556-6056",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - A Soft Error Tolerant Network-on-Chip Router Pipeline for Multi-Core Systems

AU - Poluri, Pavan

AU - Louri, Ahmed

PY - 2015/7/1

Y1 - 2015/7/1

N2 - Network-on-Chip (NoC) paradigm is rapidly evolving into an efficient interconnection network to handle the strict communication requirements between the increasing number of cores on a single chip. Diminishing transistor size is making the NoC increasingly vulnerable to both hard faults and soft errors. This paper concentrates on soft errors in NoCs. A soft error in an NoC router results in significant consequences such as data corruption, packet retransmission and deadlock among others. To this end, we propose Soft Error T olerant NoC Router (STNR) architecture, that is capable of detecting and recovering from soft errors occurring in different control stages of the routing pipeline. STNR exploits the use of idle cycles inherent in NoC packet routing pipeline to perform time redundant executions necessary for soft error tolerance. In doing so, STNR is able to detect and correct all single transient faults in the control stages of the pipeline. Simulation results using PARSEC and SPLASH-2 benchmarks show that STNR is able to accomplish such high level of soft error protection with a minimal impact on latency (an increase of 1.7 and 1.6 percent respectively). Additionally, STNR incurs an area overhead of 7 percent and power overhead of 13 percent as compared to the baseline unprotected router.

AB - Network-on-Chip (NoC) paradigm is rapidly evolving into an efficient interconnection network to handle the strict communication requirements between the increasing number of cores on a single chip. Diminishing transistor size is making the NoC increasingly vulnerable to both hard faults and soft errors. This paper concentrates on soft errors in NoCs. A soft error in an NoC router results in significant consequences such as data corruption, packet retransmission and deadlock among others. To this end, we propose Soft Error T olerant NoC Router (STNR) architecture, that is capable of detecting and recovering from soft errors occurring in different control stages of the routing pipeline. STNR exploits the use of idle cycles inherent in NoC packet routing pipeline to perform time redundant executions necessary for soft error tolerance. In doing so, STNR is able to detect and correct all single transient faults in the control stages of the pipeline. Simulation results using PARSEC and SPLASH-2 benchmarks show that STNR is able to accomplish such high level of soft error protection with a minimal impact on latency (an increase of 1.7 and 1.6 percent respectively). Additionally, STNR incurs an area overhead of 7 percent and power overhead of 13 percent as compared to the baseline unprotected router.

KW - Network-on-chip

KW - performance

KW - reliability

KW - soft error

UR - http://www.scopus.com/inward/record.url?scp=84961710656&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961710656&partnerID=8YFLogxK

U2 - 10.1109/LCA.2014.2360686

DO - 10.1109/LCA.2014.2360686

M3 - Article

VL - 14

SP - 107

EP - 110

JO - IEEE Computer Architecture Letters

JF - IEEE Computer Architecture Letters

SN - 1556-6056

IS - 2

M1 - 6912980

ER -