Data compression for improving SPM behavior

O. Ozturk, Mahmut Kandemir, I. Demirkiran, G. Chen, Mary Jane Irwin

Research output: Contribution to journalConference article

11 Citations (Scopus)

Abstract

Scratch-pad memories (SPMs) enable fast access to time-critical data. While prior research studied both static and dynamic SPM management strategies, not being able to keep all hot data (i.e., data with high reuse) in the SPM remains the biggest problem. This paper proposes data compression to increase the number of data blocks that can be kept in the SPM. Our experiments with several embedded applications show that our compression-based SPM management heuristic is very effective and outperforms prior static and dynamic SPM management approaches. We also present an ILP formulation of the problem, and show that the proposed heuristic generates competitive results with those obtained through ILP, while spending much less time in compilation.

Original languageEnglish (US)
Pages (from-to)401-406
Number of pages6
JournalProceedings - Design Automation Conference
StatePublished - Sep 20 2004
EventProceedings of the 41st Design Automation Conference - San Diego, CA, United States
Duration: Jun 7 2004Jun 11 2004

Fingerprint

Data compression
Data storage equipment
Inductive logic programming (ILP)
Experiments

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Control and Systems Engineering

Cite this

Ozturk, O., Kandemir, M., Demirkiran, I., Chen, G., & Irwin, M. J. (2004). Data compression for improving SPM behavior. Proceedings - Design Automation Conference, 401-406.
Ozturk, O. ; Kandemir, Mahmut ; Demirkiran, I. ; Chen, G. ; Irwin, Mary Jane. / Data compression for improving SPM behavior. In: Proceedings - Design Automation Conference. 2004 ; pp. 401-406.
@article{644be5a0729f4d7fa15dd5f7ecff5317,
title = "Data compression for improving SPM behavior",
abstract = "Scratch-pad memories (SPMs) enable fast access to time-critical data. While prior research studied both static and dynamic SPM management strategies, not being able to keep all hot data (i.e., data with high reuse) in the SPM remains the biggest problem. This paper proposes data compression to increase the number of data blocks that can be kept in the SPM. Our experiments with several embedded applications show that our compression-based SPM management heuristic is very effective and outperforms prior static and dynamic SPM management approaches. We also present an ILP formulation of the problem, and show that the proposed heuristic generates competitive results with those obtained through ILP, while spending much less time in compilation.",
author = "O. Ozturk and Mahmut Kandemir and I. Demirkiran and G. Chen and Irwin, {Mary Jane}",
year = "2004",
month = "9",
day = "20",
language = "English (US)",
pages = "401--406",
journal = "Proceedings - Design Automation Conference",
issn = "0738-100X",

}

Ozturk, O, Kandemir, M, Demirkiran, I, Chen, G & Irwin, MJ 2004, 'Data compression for improving SPM behavior', Proceedings - Design Automation Conference, pp. 401-406.

Data compression for improving SPM behavior. / Ozturk, O.; Kandemir, Mahmut; Demirkiran, I.; Chen, G.; Irwin, Mary Jane.

In: Proceedings - Design Automation Conference, 20.09.2004, p. 401-406.

Research output: Contribution to journalConference article

TY - JOUR

T1 - Data compression for improving SPM behavior

AU - Ozturk, O.

AU - Kandemir, Mahmut

AU - Demirkiran, I.

AU - Chen, G.

AU - Irwin, Mary Jane

PY - 2004/9/20

Y1 - 2004/9/20

N2 - Scratch-pad memories (SPMs) enable fast access to time-critical data. While prior research studied both static and dynamic SPM management strategies, not being able to keep all hot data (i.e., data with high reuse) in the SPM remains the biggest problem. This paper proposes data compression to increase the number of data blocks that can be kept in the SPM. Our experiments with several embedded applications show that our compression-based SPM management heuristic is very effective and outperforms prior static and dynamic SPM management approaches. We also present an ILP formulation of the problem, and show that the proposed heuristic generates competitive results with those obtained through ILP, while spending much less time in compilation.

AB - Scratch-pad memories (SPMs) enable fast access to time-critical data. While prior research studied both static and dynamic SPM management strategies, not being able to keep all hot data (i.e., data with high reuse) in the SPM remains the biggest problem. This paper proposes data compression to increase the number of data blocks that can be kept in the SPM. Our experiments with several embedded applications show that our compression-based SPM management heuristic is very effective and outperforms prior static and dynamic SPM management approaches. We also present an ILP formulation of the problem, and show that the proposed heuristic generates competitive results with those obtained through ILP, while spending much less time in compilation.

UR - http://www.scopus.com/inward/record.url?scp=4444283877&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4444283877&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:4444283877

SP - 401

EP - 406

JO - Proceedings - Design Automation Conference

JF - Proceedings - Design Automation Conference

SN - 0738-100X

ER -

Ozturk O, Kandemir M, Demirkiran I, Chen G, Irwin MJ. Data compression for improving SPM behavior. Proceedings - Design Automation Conference. 2004 Sep 20;401-406.