Aggressive dynamic execution of decoded traces

Benjamin Bishop, Robert Owens, Mary Jane Irwin

Research output: Contribution to conferencePaper

Abstract

In this paper, we consider the increased performance that can be obtained by using in concert, three previously proposed (and in two cases used in commercial systems) ideas. These ideas are aggressive dynamic (run time) instruction scheduling, reuse of decoded instructions, and trace scheduling. We show that these ideas complement and support one another. Hence, while each of these ideas has been shown to have merit in it's own right, when used in concert, we claim the overall advantage is greater than that obtained by using any one singly. To support this claim, we present the results from running several common multimedia kernels. Overall, these results show an average speedup of 3.50 times what can be had by using dynamic instruction scheduling alone.

Original languageEnglish (US)
Pages253-262
Number of pages10
StatePublished - Dec 1 1997
EventProceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation - Leicester, UK
Duration: Nov 3 1997Nov 5 1997

Other

OtherProceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation
CityLeicester, UK
Period11/3/9711/5/97

Fingerprint

Scheduling

All Science Journal Classification (ASJC) codes

  • Signal Processing
  • Media Technology

Cite this

Bishop, B., Owens, R., & Irwin, M. J. (1997). Aggressive dynamic execution of decoded traces. 253-262. Paper presented at Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation, Leicester, UK, .
Bishop, Benjamin ; Owens, Robert ; Irwin, Mary Jane. / Aggressive dynamic execution of decoded traces. Paper presented at Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation, Leicester, UK, .10 p.
@conference{17de841aafe441c28e990b014ab0cfee,
title = "Aggressive dynamic execution of decoded traces",
abstract = "In this paper, we consider the increased performance that can be obtained by using in concert, three previously proposed (and in two cases used in commercial systems) ideas. These ideas are aggressive dynamic (run time) instruction scheduling, reuse of decoded instructions, and trace scheduling. We show that these ideas complement and support one another. Hence, while each of these ideas has been shown to have merit in it's own right, when used in concert, we claim the overall advantage is greater than that obtained by using any one singly. To support this claim, we present the results from running several common multimedia kernels. Overall, these results show an average speedup of 3.50 times what can be had by using dynamic instruction scheduling alone.",
author = "Benjamin Bishop and Robert Owens and Irwin, {Mary Jane}",
year = "1997",
month = "12",
day = "1",
language = "English (US)",
pages = "253--262",
note = "Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation ; Conference date: 03-11-1997 Through 05-11-1997",

}

Bishop, B, Owens, R & Irwin, MJ 1997, 'Aggressive dynamic execution of decoded traces', Paper presented at Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation, Leicester, UK, 11/3/97 - 11/5/97 pp. 253-262.

Aggressive dynamic execution of decoded traces. / Bishop, Benjamin; Owens, Robert; Irwin, Mary Jane.

1997. 253-262 Paper presented at Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation, Leicester, UK, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Aggressive dynamic execution of decoded traces

AU - Bishop, Benjamin

AU - Owens, Robert

AU - Irwin, Mary Jane

PY - 1997/12/1

Y1 - 1997/12/1

N2 - In this paper, we consider the increased performance that can be obtained by using in concert, three previously proposed (and in two cases used in commercial systems) ideas. These ideas are aggressive dynamic (run time) instruction scheduling, reuse of decoded instructions, and trace scheduling. We show that these ideas complement and support one another. Hence, while each of these ideas has been shown to have merit in it's own right, when used in concert, we claim the overall advantage is greater than that obtained by using any one singly. To support this claim, we present the results from running several common multimedia kernels. Overall, these results show an average speedup of 3.50 times what can be had by using dynamic instruction scheduling alone.

AB - In this paper, we consider the increased performance that can be obtained by using in concert, three previously proposed (and in two cases used in commercial systems) ideas. These ideas are aggressive dynamic (run time) instruction scheduling, reuse of decoded instructions, and trace scheduling. We show that these ideas complement and support one another. Hence, while each of these ideas has been shown to have merit in it's own right, when used in concert, we claim the overall advantage is greater than that obtained by using any one singly. To support this claim, we present the results from running several common multimedia kernels. Overall, these results show an average speedup of 3.50 times what can be had by using dynamic instruction scheduling alone.

UR - http://www.scopus.com/inward/record.url?scp=0031347581&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031347581&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0031347581

SP - 253

EP - 262

ER -

Bishop B, Owens R, Irwin MJ. Aggressive dynamic execution of decoded traces. 1997. Paper presented at Proceedings of the 1997 IEEE Workshop on Signal Processing Systems, SiPS 97: Design and Implementation, Leicester, UK, .