Selective block buffering TLB system for embedded processors

J. H. Lee, C. Weems, Shin-Dug Kim

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

The authors present a translation lookaside buffer (TLB) system with low power consumption for embedded processors. The proposed TLB is constructed as multiple banks, each with an associated block buffer and a corresponding comparator. Either the block buffer or the main bank is selectively accessed on the basis of two bits in the tag buffer. Dynamic power savings are achieved by reducing the number of entries accessed in parallel, as a result of using the tag buffer as a filtering mechanism. The performance overhead of the proposed TLB is negligible compared with other hierarchical TLB structures. For example, the two-cycle overhead of the proposed TLB is only ∼1%, as compared with 5% overhead for a filter (micro)-TLB and 14% overhead for a banked-TLB with block buffering. The authors show that the average hit ratios of the block buffers and the main banks of the proposed TLB are 94% and 6%, respectively. Dynamic power is reduced by ∼93% with respect to a fully associative TLB, 87% with respect to a filter-TLB and 60% relative to a banked-TLB with block buffering. Therefore, significant power savings are achieved with only a small performance degradation.

Original languageEnglish
Pages (from-to)507-516
Number of pages10
JournalIEE Proceedings: Computers and Digital Techniques
Volume152
Issue number4
DOIs
Publication statusPublished - 2005 Jul 1

Fingerprint

Embedded Processor
Buffer
Electric power utilization
Degradation
Power Saving
Filter

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Hardware and Architecture
  • Computational Theory and Mathematics

Cite this

@article{78cbfd5ff3e945aa89372e263c938e44,
title = "Selective block buffering TLB system for embedded processors",
abstract = "The authors present a translation lookaside buffer (TLB) system with low power consumption for embedded processors. The proposed TLB is constructed as multiple banks, each with an associated block buffer and a corresponding comparator. Either the block buffer or the main bank is selectively accessed on the basis of two bits in the tag buffer. Dynamic power savings are achieved by reducing the number of entries accessed in parallel, as a result of using the tag buffer as a filtering mechanism. The performance overhead of the proposed TLB is negligible compared with other hierarchical TLB structures. For example, the two-cycle overhead of the proposed TLB is only ∼1{\%}, as compared with 5{\%} overhead for a filter (micro)-TLB and 14{\%} overhead for a banked-TLB with block buffering. The authors show that the average hit ratios of the block buffers and the main banks of the proposed TLB are 94{\%} and 6{\%}, respectively. Dynamic power is reduced by ∼93{\%} with respect to a fully associative TLB, 87{\%} with respect to a filter-TLB and 60{\%} relative to a banked-TLB with block buffering. Therefore, significant power savings are achieved with only a small performance degradation.",
author = "Lee, {J. H.} and C. Weems and Shin-Dug Kim",
year = "2005",
month = "7",
day = "1",
doi = "10.1049/ip-cdt:20045025",
language = "English",
volume = "152",
pages = "507--516",
journal = "IEE Proceedings: Computers and Digital Techniques",
issn = "1350-2387",
publisher = "Institute of Electrical Engineers",
number = "4",

}

Selective block buffering TLB system for embedded processors. / Lee, J. H.; Weems, C.; Kim, Shin-Dug.

In: IEE Proceedings: Computers and Digital Techniques, Vol. 152, No. 4, 01.07.2005, p. 507-516.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Selective block buffering TLB system for embedded processors

AU - Lee, J. H.

AU - Weems, C.

AU - Kim, Shin-Dug

PY - 2005/7/1

Y1 - 2005/7/1

N2 - The authors present a translation lookaside buffer (TLB) system with low power consumption for embedded processors. The proposed TLB is constructed as multiple banks, each with an associated block buffer and a corresponding comparator. Either the block buffer or the main bank is selectively accessed on the basis of two bits in the tag buffer. Dynamic power savings are achieved by reducing the number of entries accessed in parallel, as a result of using the tag buffer as a filtering mechanism. The performance overhead of the proposed TLB is negligible compared with other hierarchical TLB structures. For example, the two-cycle overhead of the proposed TLB is only ∼1%, as compared with 5% overhead for a filter (micro)-TLB and 14% overhead for a banked-TLB with block buffering. The authors show that the average hit ratios of the block buffers and the main banks of the proposed TLB are 94% and 6%, respectively. Dynamic power is reduced by ∼93% with respect to a fully associative TLB, 87% with respect to a filter-TLB and 60% relative to a banked-TLB with block buffering. Therefore, significant power savings are achieved with only a small performance degradation.

AB - The authors present a translation lookaside buffer (TLB) system with low power consumption for embedded processors. The proposed TLB is constructed as multiple banks, each with an associated block buffer and a corresponding comparator. Either the block buffer or the main bank is selectively accessed on the basis of two bits in the tag buffer. Dynamic power savings are achieved by reducing the number of entries accessed in parallel, as a result of using the tag buffer as a filtering mechanism. The performance overhead of the proposed TLB is negligible compared with other hierarchical TLB structures. For example, the two-cycle overhead of the proposed TLB is only ∼1%, as compared with 5% overhead for a filter (micro)-TLB and 14% overhead for a banked-TLB with block buffering. The authors show that the average hit ratios of the block buffers and the main banks of the proposed TLB are 94% and 6%, respectively. Dynamic power is reduced by ∼93% with respect to a fully associative TLB, 87% with respect to a filter-TLB and 60% relative to a banked-TLB with block buffering. Therefore, significant power savings are achieved with only a small performance degradation.

UR - http://www.scopus.com/inward/record.url?scp=23744464009&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23744464009&partnerID=8YFLogxK

U2 - 10.1049/ip-cdt:20045025

DO - 10.1049/ip-cdt:20045025

M3 - Article

AN - SCOPUS:23744464009

VL - 152

SP - 507

EP - 516

JO - IEE Proceedings: Computers and Digital Techniques

JF - IEE Proceedings: Computers and Digital Techniques

SN - 1350-2387

IS - 4

ER -