Low power data processing by elimination of redundant computations

Mir Azam; Paul Franzon; Wentai Liu

doi:10.1145/263272.263348

Low power data processing by elimination of redundant computations

Mir Azam
, Paul Franzon
, Wentai Liu

Department of Biomedical Engineering

Research output: Contribution to conference › Paper › peer-review

23 Citations (Scopus)

Abstract

We suggest a new technique to reduce energy consumption in the processor datapath without sacrificing performance by exploiting operand value locality at run time. Data locality is one of the major characteristics of video streams as well as other commonly used applications. We use a cache-like scheme to store a selective history of computation results, and the resultant re-use leads to power savings. The cache is indexed by the operands. Based on our model, an 8 to 128 entry execution cache reduces power consumption by 20% to 60%.

Original language	English
Pages	259-264
Number of pages	6
DOIs	https://doi.org/10.1145/263272.263348
Publication status	Published - 1997
Event	Proceedings of the 1997 International Symposium on Low Power Electronics and Design - Monterey, CA, USA Duration: 1997 Aug 18 → 1997 Aug 20

Conference

Conference	Proceedings of the 1997 International Symposium on Low Power Electronics and Design
City	Monterey, CA, USA
Period	97-08-18 → 97-08-20

All Science Journal Classification (ASJC) codes

General Engineering

Access to Document

10.1145/263272.263348

Cite this

@conference{66c18aab3fbe45be8102838edd559861,

title = "Low power data processing by elimination of redundant computations",

abstract = "We suggest a new technique to reduce energy consumption in the processor datapath without sacrificing performance by exploiting operand value locality at run time. Data locality is one of the major characteristics of video streams as well as other commonly used applications. We use a cache-like scheme to store a selective history of computation results, and the resultant re-use leads to power savings. The cache is indexed by the operands. Based on our model, an 8 to 128 entry execution cache reduces power consumption by 20\% to 60\%.",

author = "Mir Azam and Paul Franzon and Wentai Liu",

year = "1997",

doi = "10.1145/263272.263348",

language = "English",

pages = "259--264",

note = "Proceedings of the 1997 International Symposium on Low Power Electronics and Design ; Conference date: 18-08-1997 Through 20-08-1997",

}

TY - CONF

T1 - Low power data processing by elimination of redundant computations

AU - Azam, Mir

AU - Franzon, Paul

AU - Liu, Wentai

PY - 1997

Y1 - 1997

N2 - We suggest a new technique to reduce energy consumption in the processor datapath without sacrificing performance by exploiting operand value locality at run time. Data locality is one of the major characteristics of video streams as well as other commonly used applications. We use a cache-like scheme to store a selective history of computation results, and the resultant re-use leads to power savings. The cache is indexed by the operands. Based on our model, an 8 to 128 entry execution cache reduces power consumption by 20% to 60%.

AB - We suggest a new technique to reduce energy consumption in the processor datapath without sacrificing performance by exploiting operand value locality at run time. Data locality is one of the major characteristics of video streams as well as other commonly used applications. We use a cache-like scheme to store a selective history of computation results, and the resultant re-use leads to power savings. The cache is indexed by the operands. Based on our model, an 8 to 128 entry execution cache reduces power consumption by 20% to 60%.

UR - https://www.scopus.com/pages/publications/0030684917

UR - https://www.scopus.com/pages/publications/0030684917#tab=citedBy

U2 - 10.1145/263272.263348

DO - 10.1145/263272.263348

M3 - Paper

AN - SCOPUS:0030684917

SP - 259

EP - 264

T2 - Proceedings of the 1997 International Symposium on Low Power Electronics and Design

Y2 - 18 August 1997 through 20 August 1997

ER -

Low power data processing by elimination of redundant computations

Abstract

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this