Enhancing spatial locality via data layout optimizations

Mahmut Kandemir, A. Choudhary, J. Ramanujam, N. Shenoy, P. Banerjee

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

This paper aims to improve locality of references by suitably choosing array layouts. We use a new definition of spatial reuse vectors that takes into account memory layout of arrays. This capability creates two opportunities. First, it allows us to develop an array restructuring framework based on a combination of hyperplane theory and reuse vectors. Second, it allows us to observe the effect of different array layout optimizations on spatial reuse vectors. Since the iteration space based locality optimizations also change the spatial reuse vectors, our approach allows us to compare the iteration-space based and data-space based approaches in terms of their effects on spatial reuse vectors. We illustrate the effectiveness of our technique using an example from the BLAS library on the SGI Origin distributed shared-memory machine.

Original languageEnglish (US)
Title of host publicationEuro-Par 1998 Parallel Processing - 4th International Conference, Proceedings
Pages422-434
Number of pages13
StatePublished - Dec 1 1998
Event4th International Conference on Parallel Processing, Euro-Par 1998 - Southampton, United Kingdom
Duration: Sep 1 1998Sep 4 1998

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1470 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other4th International Conference on Parallel Processing, Euro-Par 1998
CountryUnited Kingdom
CitySouthampton
Period9/1/989/4/98

Fingerprint

Data Locality
Reuse
Layout
Optimization
Locality
Iteration
Data storage equipment
Distributed Shared Memory
Hyperplane

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Kandemir, M., Choudhary, A., Ramanujam, J., Shenoy, N., & Banerjee, P. (1998). Enhancing spatial locality via data layout optimizations. In Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings (pp. 422-434). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1470 LNCS).
Kandemir, Mahmut ; Choudhary, A. ; Ramanujam, J. ; Shenoy, N. ; Banerjee, P. / Enhancing spatial locality via data layout optimizations. Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings. 1998. pp. 422-434 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{e216dea751cf4cd19dfd61e8a1954fe3,
title = "Enhancing spatial locality via data layout optimizations",
abstract = "This paper aims to improve locality of references by suitably choosing array layouts. We use a new definition of spatial reuse vectors that takes into account memory layout of arrays. This capability creates two opportunities. First, it allows us to develop an array restructuring framework based on a combination of hyperplane theory and reuse vectors. Second, it allows us to observe the effect of different array layout optimizations on spatial reuse vectors. Since the iteration space based locality optimizations also change the spatial reuse vectors, our approach allows us to compare the iteration-space based and data-space based approaches in terms of their effects on spatial reuse vectors. We illustrate the effectiveness of our technique using an example from the BLAS library on the SGI Origin distributed shared-memory machine.",
author = "Mahmut Kandemir and A. Choudhary and J. Ramanujam and N. Shenoy and P. Banerjee",
year = "1998",
month = "12",
day = "1",
language = "English (US)",
isbn = "3540649522",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "422--434",
booktitle = "Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings",

}

Kandemir, M, Choudhary, A, Ramanujam, J, Shenoy, N & Banerjee, P 1998, Enhancing spatial locality via data layout optimizations. in Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1470 LNCS, pp. 422-434, 4th International Conference on Parallel Processing, Euro-Par 1998, Southampton, United Kingdom, 9/1/98.

Enhancing spatial locality via data layout optimizations. / Kandemir, Mahmut; Choudhary, A.; Ramanujam, J.; Shenoy, N.; Banerjee, P.

Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings. 1998. p. 422-434 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1470 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Enhancing spatial locality via data layout optimizations

AU - Kandemir, Mahmut

AU - Choudhary, A.

AU - Ramanujam, J.

AU - Shenoy, N.

AU - Banerjee, P.

PY - 1998/12/1

Y1 - 1998/12/1

N2 - This paper aims to improve locality of references by suitably choosing array layouts. We use a new definition of spatial reuse vectors that takes into account memory layout of arrays. This capability creates two opportunities. First, it allows us to develop an array restructuring framework based on a combination of hyperplane theory and reuse vectors. Second, it allows us to observe the effect of different array layout optimizations on spatial reuse vectors. Since the iteration space based locality optimizations also change the spatial reuse vectors, our approach allows us to compare the iteration-space based and data-space based approaches in terms of their effects on spatial reuse vectors. We illustrate the effectiveness of our technique using an example from the BLAS library on the SGI Origin distributed shared-memory machine.

AB - This paper aims to improve locality of references by suitably choosing array layouts. We use a new definition of spatial reuse vectors that takes into account memory layout of arrays. This capability creates two opportunities. First, it allows us to develop an array restructuring framework based on a combination of hyperplane theory and reuse vectors. Second, it allows us to observe the effect of different array layout optimizations on spatial reuse vectors. Since the iteration space based locality optimizations also change the spatial reuse vectors, our approach allows us to compare the iteration-space based and data-space based approaches in terms of their effects on spatial reuse vectors. We illustrate the effectiveness of our technique using an example from the BLAS library on the SGI Origin distributed shared-memory machine.

UR - http://www.scopus.com/inward/record.url?scp=22044440681&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=22044440681&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:22044440681

SN - 3540649522

SN - 9783540649526

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 422

EP - 434

BT - Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings

ER -

Kandemir M, Choudhary A, Ramanujam J, Shenoy N, Banerjee P. Enhancing spatial locality via data layout optimizations. In Euro-Par 1998 Parallel Processing - 4th International Conference, Proceedings. 1998. p. 422-434. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).