Exploiting algorithmic noise tolerance for scalable on-chip voltage regulation

Longfei Wang, S. Karen Khatamifard, Ulya Karpuzcu, Selcuk Kose

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

With the advent of on-chip digital low-dropout (DLDO) regulators, distributed on-chip voltage regulation has become increasingly promising. Environmental and operating conditions have been demonstrated to degrade DLDO performance, which directly affects execution accuracy. The area overhead (OH) needed to compensate aging-induced voltage noise degradation can be significant. Accordingly, in this paper, the algorithmic noise tolerance of certain processor components is exploited as an area-quality control knob to trade the program output quality for area OH. Furthermore, efficient and lightweight techniques utilizing a unidirectional shift register and reduced clock pulsewidth triggering are proposed to realize a novel aging-aware (AA) DLDO to achieve a better area and quality tradeoff. Owing to the large number and distributed nature of voltage regulators, with the proposed design, both the number of regulators utilized in the system and the size of each local regulator are scalable to satisfy the needs of different applications and processor components with varying algorithmic noise tolerance. It is demonstrated through simulation of an IBM POWER8 like processor that the proposed AA design can achieve up to, respectively, 43.2% and 3 × transient and steady-state performance improvement. Additionally, more than 10% area OH saving can be achieved over a 5-year period.

Original languageEnglish (US)
Article number8528543
Pages (from-to)229-242
Number of pages14
JournalIEEE Transactions on Very Large Scale Integration (VLSI) Systems
Volume27
Issue number1
DOIs
StatePublished - Jan 2019

Fingerprint

Voltage control
Aging of materials
Knobs
Voltage regulators
Shift registers
Quality control
Clocks
Degradation
Electric potential

Keywords

  • Algorithmic noise tolerance
  • digital low-dropout (DLDO) regulator
  • reliability
  • scalable on-chip voltage regulation
  • voltage noise

Cite this

Exploiting algorithmic noise tolerance for scalable on-chip voltage regulation. / Wang, Longfei; Khatamifard, S. Karen; Karpuzcu, Ulya; Kose, Selcuk.

In: IEEE Transactions on Very Large Scale Integration (VLSI) Systems, Vol. 27, No. 1, 8528543, 01.2019, p. 229-242.

Research output: Contribution to journalArticle

@article{3a519664675f47fcb2c057b64569caf4,
title = "Exploiting algorithmic noise tolerance for scalable on-chip voltage regulation",
abstract = "With the advent of on-chip digital low-dropout (DLDO) regulators, distributed on-chip voltage regulation has become increasingly promising. Environmental and operating conditions have been demonstrated to degrade DLDO performance, which directly affects execution accuracy. The area overhead (OH) needed to compensate aging-induced voltage noise degradation can be significant. Accordingly, in this paper, the algorithmic noise tolerance of certain processor components is exploited as an area-quality control knob to trade the program output quality for area OH. Furthermore, efficient and lightweight techniques utilizing a unidirectional shift register and reduced clock pulsewidth triggering are proposed to realize a novel aging-aware (AA) DLDO to achieve a better area and quality tradeoff. Owing to the large number and distributed nature of voltage regulators, with the proposed design, both the number of regulators utilized in the system and the size of each local regulator are scalable to satisfy the needs of different applications and processor components with varying algorithmic noise tolerance. It is demonstrated through simulation of an IBM POWER8 like processor that the proposed AA design can achieve up to, respectively, 43.2{\%} and 3 × transient and steady-state performance improvement. Additionally, more than 10{\%} area OH saving can be achieved over a 5-year period.",
keywords = "Algorithmic noise tolerance, digital low-dropout (DLDO) regulator, reliability, scalable on-chip voltage regulation, voltage noise",
author = "Longfei Wang and Khatamifard, {S. Karen} and Ulya Karpuzcu and Selcuk Kose",
year = "2019",
month = "1",
doi = "10.1109/TVLSI.2018.2871381",
language = "English (US)",
volume = "27",
pages = "229--242",
journal = "IEEE Transactions on Very Large Scale Integration (VLSI) Systems",
issn = "1063-8210",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - Exploiting algorithmic noise tolerance for scalable on-chip voltage regulation

AU - Wang, Longfei

AU - Khatamifard, S. Karen

AU - Karpuzcu, Ulya

AU - Kose, Selcuk

PY - 2019/1

Y1 - 2019/1

N2 - With the advent of on-chip digital low-dropout (DLDO) regulators, distributed on-chip voltage regulation has become increasingly promising. Environmental and operating conditions have been demonstrated to degrade DLDO performance, which directly affects execution accuracy. The area overhead (OH) needed to compensate aging-induced voltage noise degradation can be significant. Accordingly, in this paper, the algorithmic noise tolerance of certain processor components is exploited as an area-quality control knob to trade the program output quality for area OH. Furthermore, efficient and lightweight techniques utilizing a unidirectional shift register and reduced clock pulsewidth triggering are proposed to realize a novel aging-aware (AA) DLDO to achieve a better area and quality tradeoff. Owing to the large number and distributed nature of voltage regulators, with the proposed design, both the number of regulators utilized in the system and the size of each local regulator are scalable to satisfy the needs of different applications and processor components with varying algorithmic noise tolerance. It is demonstrated through simulation of an IBM POWER8 like processor that the proposed AA design can achieve up to, respectively, 43.2% and 3 × transient and steady-state performance improvement. Additionally, more than 10% area OH saving can be achieved over a 5-year period.

AB - With the advent of on-chip digital low-dropout (DLDO) regulators, distributed on-chip voltage regulation has become increasingly promising. Environmental and operating conditions have been demonstrated to degrade DLDO performance, which directly affects execution accuracy. The area overhead (OH) needed to compensate aging-induced voltage noise degradation can be significant. Accordingly, in this paper, the algorithmic noise tolerance of certain processor components is exploited as an area-quality control knob to trade the program output quality for area OH. Furthermore, efficient and lightweight techniques utilizing a unidirectional shift register and reduced clock pulsewidth triggering are proposed to realize a novel aging-aware (AA) DLDO to achieve a better area and quality tradeoff. Owing to the large number and distributed nature of voltage regulators, with the proposed design, both the number of regulators utilized in the system and the size of each local regulator are scalable to satisfy the needs of different applications and processor components with varying algorithmic noise tolerance. It is demonstrated through simulation of an IBM POWER8 like processor that the proposed AA design can achieve up to, respectively, 43.2% and 3 × transient and steady-state performance improvement. Additionally, more than 10% area OH saving can be achieved over a 5-year period.

KW - Algorithmic noise tolerance

KW - digital low-dropout (DLDO) regulator

KW - reliability

KW - scalable on-chip voltage regulation

KW - voltage noise

UR - http://www.scopus.com/inward/record.url?scp=85056322281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056322281&partnerID=8YFLogxK

U2 - 10.1109/TVLSI.2018.2871381

DO - 10.1109/TVLSI.2018.2871381

M3 - Article

AN - SCOPUS:85056322281

VL - 27

SP - 229

EP - 242

JO - IEEE Transactions on Very Large Scale Integration (VLSI) Systems

JF - IEEE Transactions on Very Large Scale Integration (VLSI) Systems

SN - 1063-8210

IS - 1

M1 - 8528543

ER -