Bypass aware instruction scheduling for register file power reduction

Sanghyun Park, Alex Nicolau, Aviral Shrivastava, Yunheung Paek, Nikil Dutt, Eugene Earlie

Research output: Contribution to journalArticle

17 Citations (Scopus)

Abstract

Since register files suffer from some of the highest power densities within processors, designers have investigated several architectural strategies for register file power reduction, including "On Demand RF Read" where the register file is read only if the operand value is not available from the bypasses. However, we show in this paper that significant additional reductions in the register file power consumption can be obtained by scheduling instructions so that they transfer the operands via bypasses, rather than reading from the register file. Such instruction scheduling requires the compiler to be cognizant of the bypasses in the processor pipeline. In this paper, we develop several bypass aware instruction scheduling heuristics varying in time complexity, and study their effectiveness on the Intel XScale processor pipeline running MiBench benchmarks. Our experimental results show additional power consumption reductions of up to 26% and on average 12% over and above the register file power reduction achieved through existing techniques.

Original languageEnglish (US)
Pages (from-to)173-181
Number of pages9
JournalACM SIGPLAN Notices
Volume41
Issue number7
DOIs
StatePublished - Jul 2006
Externally publishedYes

Fingerprint

Scheduling
Electric power utilization
Pipelines

Keywords

  • Architecture-sensitive Compiler
  • Bypass-sensitive
  • Forwarding Paths
  • Operation Table
  • Power Consumption
  • Processor Bypasses
  • Register File
  • Reservation Table

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software

Cite this

Bypass aware instruction scheduling for register file power reduction. / Park, Sanghyun; Nicolau, Alex; Shrivastava, Aviral; Paek, Yunheung; Dutt, Nikil; Earlie, Eugene.

In: ACM SIGPLAN Notices, Vol. 41, No. 7, 07.2006, p. 173-181.

Research output: Contribution to journalArticle

Park, Sanghyun ; Nicolau, Alex ; Shrivastava, Aviral ; Paek, Yunheung ; Dutt, Nikil ; Earlie, Eugene. / Bypass aware instruction scheduling for register file power reduction. In: ACM SIGPLAN Notices. 2006 ; Vol. 41, No. 7. pp. 173-181.
@article{c1e02393020e421bbaf87030a1652d91,
title = "Bypass aware instruction scheduling for register file power reduction",
abstract = "Since register files suffer from some of the highest power densities within processors, designers have investigated several architectural strategies for register file power reduction, including {"}On Demand RF Read{"} where the register file is read only if the operand value is not available from the bypasses. However, we show in this paper that significant additional reductions in the register file power consumption can be obtained by scheduling instructions so that they transfer the operands via bypasses, rather than reading from the register file. Such instruction scheduling requires the compiler to be cognizant of the bypasses in the processor pipeline. In this paper, we develop several bypass aware instruction scheduling heuristics varying in time complexity, and study their effectiveness on the Intel XScale processor pipeline running MiBench benchmarks. Our experimental results show additional power consumption reductions of up to 26{\%} and on average 12{\%} over and above the register file power reduction achieved through existing techniques.",
keywords = "Architecture-sensitive Compiler, Bypass-sensitive, Forwarding Paths, Operation Table, Power Consumption, Processor Bypasses, Register File, Reservation Table",
author = "Sanghyun Park and Alex Nicolau and Aviral Shrivastava and Yunheung Paek and Nikil Dutt and Eugene Earlie",
year = "2006",
month = "7",
doi = "10.1145/1159974.1134675",
language = "English (US)",
volume = "41",
pages = "173--181",
journal = "SIGPLAN Notices (ACM Special Interest Group on Programming Languages)",
issn = "1523-2867",
publisher = "Association for Computing Machinery (ACM)",
number = "7",

}

TY - JOUR

T1 - Bypass aware instruction scheduling for register file power reduction

AU - Park, Sanghyun

AU - Nicolau, Alex

AU - Shrivastava, Aviral

AU - Paek, Yunheung

AU - Dutt, Nikil

AU - Earlie, Eugene

PY - 2006/7

Y1 - 2006/7

N2 - Since register files suffer from some of the highest power densities within processors, designers have investigated several architectural strategies for register file power reduction, including "On Demand RF Read" where the register file is read only if the operand value is not available from the bypasses. However, we show in this paper that significant additional reductions in the register file power consumption can be obtained by scheduling instructions so that they transfer the operands via bypasses, rather than reading from the register file. Such instruction scheduling requires the compiler to be cognizant of the bypasses in the processor pipeline. In this paper, we develop several bypass aware instruction scheduling heuristics varying in time complexity, and study their effectiveness on the Intel XScale processor pipeline running MiBench benchmarks. Our experimental results show additional power consumption reductions of up to 26% and on average 12% over and above the register file power reduction achieved through existing techniques.

AB - Since register files suffer from some of the highest power densities within processors, designers have investigated several architectural strategies for register file power reduction, including "On Demand RF Read" where the register file is read only if the operand value is not available from the bypasses. However, we show in this paper that significant additional reductions in the register file power consumption can be obtained by scheduling instructions so that they transfer the operands via bypasses, rather than reading from the register file. Such instruction scheduling requires the compiler to be cognizant of the bypasses in the processor pipeline. In this paper, we develop several bypass aware instruction scheduling heuristics varying in time complexity, and study their effectiveness on the Intel XScale processor pipeline running MiBench benchmarks. Our experimental results show additional power consumption reductions of up to 26% and on average 12% over and above the register file power reduction achieved through existing techniques.

KW - Architecture-sensitive Compiler

KW - Bypass-sensitive

KW - Forwarding Paths

KW - Operation Table

KW - Power Consumption

KW - Processor Bypasses

KW - Register File

KW - Reservation Table

UR - http://www.scopus.com/inward/record.url?scp=33748999034&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748999034&partnerID=8YFLogxK

U2 - 10.1145/1159974.1134675

DO - 10.1145/1159974.1134675

M3 - Article

VL - 41

SP - 173

EP - 181

JO - SIGPLAN Notices (ACM Special Interest Group on Programming Languages)

JF - SIGPLAN Notices (ACM Special Interest Group on Programming Languages)

SN - 1523-2867

IS - 7

ER -