Workload assignment considering NBTI degradation in multicore systems

Jin Sun, Roman L Lysecky, Karthik Shankar, Avinash Kodi, Ahmed Louri, Meiling Wang

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

With continuously shrinking technology, reliability issues such as Negative Bias Temperature Instability (NBTI) has resulted in considerable degradation of device performance, and eventually the short mean-timeto- failure (MTTF) of the whole multicore system. This article proposes a new workload balancing scheme based on device-level fractional NBTI model to balance the workload among active cores while relaxing stressed ones. Starting with NBTI-induced threshold voltage degradation, we define a concept of Capacity Rate (CR) as an indication of one core's ability to accept workload. Capacity rate captures core's performance variability in terms of delay and power metrics under the impact of NBTI aging. The proposed workload balancing framework employs the capacity rates as workload constraints, applies a Dynamic Zoning (DZ) algorithm to group cores into zones to process task flows, and then uses Dynamic Task Scheduling (DTS) to allocate tasks in each zone with balanced workload and minimum communication cost. Experimental results on a 64-core system show that by allowing a small part of the cores to relax over a short time period, the proposed methodology improves multicore system yield (percentage of core failures) by 20%, while extending MTTF by 30% with insignificant degradation in performance (less than 3%).

Original languageEnglish (US)
Article number4
JournalACM Journal on Emerging Technologies in Computing Systems
Volume10
Issue number1
DOIs
StatePublished - Jan 2014

Fingerprint

Degradation
Zoning
Threshold voltage
Aging of materials
Scheduling
Negative bias temperature instability
Communication
Costs

Keywords

  • Dynamic task scheduling
  • Dynamic zoning
  • Multicore systems
  • Negative bias temperature instability capacity rate
  • Workload balancing

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Electrical and Electronic Engineering

Cite this

Workload assignment considering NBTI degradation in multicore systems. / Sun, Jin; Lysecky, Roman L; Shankar, Karthik; Kodi, Avinash; Louri, Ahmed; Wang, Meiling.

In: ACM Journal on Emerging Technologies in Computing Systems, Vol. 10, No. 1, 4, 01.2014.

Research output: Contribution to journalArticle

@article{42c79a4620664e33970cfc0980981b4d,
title = "Workload assignment considering NBTI degradation in multicore systems",
abstract = "With continuously shrinking technology, reliability issues such as Negative Bias Temperature Instability (NBTI) has resulted in considerable degradation of device performance, and eventually the short mean-timeto- failure (MTTF) of the whole multicore system. This article proposes a new workload balancing scheme based on device-level fractional NBTI model to balance the workload among active cores while relaxing stressed ones. Starting with NBTI-induced threshold voltage degradation, we define a concept of Capacity Rate (CR) as an indication of one core's ability to accept workload. Capacity rate captures core's performance variability in terms of delay and power metrics under the impact of NBTI aging. The proposed workload balancing framework employs the capacity rates as workload constraints, applies a Dynamic Zoning (DZ) algorithm to group cores into zones to process task flows, and then uses Dynamic Task Scheduling (DTS) to allocate tasks in each zone with balanced workload and minimum communication cost. Experimental results on a 64-core system show that by allowing a small part of the cores to relax over a short time period, the proposed methodology improves multicore system yield (percentage of core failures) by 20{\%}, while extending MTTF by 30{\%} with insignificant degradation in performance (less than 3{\%}).",
keywords = "Dynamic task scheduling, Dynamic zoning, Multicore systems, Negative bias temperature instability capacity rate, Workload balancing",
author = "Jin Sun and Lysecky, {Roman L} and Karthik Shankar and Avinash Kodi and Ahmed Louri and Meiling Wang",
year = "2014",
month = "1",
doi = "10.1145/2539124",
language = "English (US)",
volume = "10",
journal = "ACM Journal on Emerging Technologies in Computing Systems",
issn = "1550-4832",
publisher = "Association for Computing Machinery (ACM)",
number = "1",

}

TY - JOUR

T1 - Workload assignment considering NBTI degradation in multicore systems

AU - Sun, Jin

AU - Lysecky, Roman L

AU - Shankar, Karthik

AU - Kodi, Avinash

AU - Louri, Ahmed

AU - Wang, Meiling

PY - 2014/1

Y1 - 2014/1

N2 - With continuously shrinking technology, reliability issues such as Negative Bias Temperature Instability (NBTI) has resulted in considerable degradation of device performance, and eventually the short mean-timeto- failure (MTTF) of the whole multicore system. This article proposes a new workload balancing scheme based on device-level fractional NBTI model to balance the workload among active cores while relaxing stressed ones. Starting with NBTI-induced threshold voltage degradation, we define a concept of Capacity Rate (CR) as an indication of one core's ability to accept workload. Capacity rate captures core's performance variability in terms of delay and power metrics under the impact of NBTI aging. The proposed workload balancing framework employs the capacity rates as workload constraints, applies a Dynamic Zoning (DZ) algorithm to group cores into zones to process task flows, and then uses Dynamic Task Scheduling (DTS) to allocate tasks in each zone with balanced workload and minimum communication cost. Experimental results on a 64-core system show that by allowing a small part of the cores to relax over a short time period, the proposed methodology improves multicore system yield (percentage of core failures) by 20%, while extending MTTF by 30% with insignificant degradation in performance (less than 3%).

AB - With continuously shrinking technology, reliability issues such as Negative Bias Temperature Instability (NBTI) has resulted in considerable degradation of device performance, and eventually the short mean-timeto- failure (MTTF) of the whole multicore system. This article proposes a new workload balancing scheme based on device-level fractional NBTI model to balance the workload among active cores while relaxing stressed ones. Starting with NBTI-induced threshold voltage degradation, we define a concept of Capacity Rate (CR) as an indication of one core's ability to accept workload. Capacity rate captures core's performance variability in terms of delay and power metrics under the impact of NBTI aging. The proposed workload balancing framework employs the capacity rates as workload constraints, applies a Dynamic Zoning (DZ) algorithm to group cores into zones to process task flows, and then uses Dynamic Task Scheduling (DTS) to allocate tasks in each zone with balanced workload and minimum communication cost. Experimental results on a 64-core system show that by allowing a small part of the cores to relax over a short time period, the proposed methodology improves multicore system yield (percentage of core failures) by 20%, while extending MTTF by 30% with insignificant degradation in performance (less than 3%).

KW - Dynamic task scheduling

KW - Dynamic zoning

KW - Multicore systems

KW - Negative bias temperature instability capacity rate

KW - Workload balancing

UR - http://www.scopus.com/inward/record.url?scp=84892615172&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84892615172&partnerID=8YFLogxK

U2 - 10.1145/2539124

DO - 10.1145/2539124

M3 - Article

AN - SCOPUS:84892615172

VL - 10

JO - ACM Journal on Emerging Technologies in Computing Systems

JF - ACM Journal on Emerging Technologies in Computing Systems

SN - 1550-4832

IS - 1

M1 - 4

ER -