Cooperative parallelization

Praveen Yedlapalli, Emre Kultursay, Mahmut Kandemir

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

1 Citation (Scopus)

Abstract

We propose a cooperation between the programmer, the compiler and the runtime system to identify, exploit and efficiently exercise the parallelism available in many pointer based applications. Our parallelization strategy, called Cooperative Parallelization, is driven by programmer directives as well as runtime information. We show that minimal information from the programmer can be combined with runtime information to extract latent parallelism in many pointer intensive applications that involve trees and linked lists. We implemented a compilation framework which automatically parallelizes programs annotated with parallelism directives. We evaluated our approach on a collection of linked list and tree based applications. Our results show that we can achieve speedups of up to 15 on a sixteen-core platform. We also compared our approach to OpenMP both qualitatively and quantitatively.

Original languageEnglish (US)
Title of host publication2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011
Pages134-141
Number of pages8
DOIs
StatePublished - Dec 1 2011
Event2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011 - San Jose, CA, United States
Duration: Nov 7 2011Nov 10 2011

Publication series

NameIEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD
ISSN (Print)1092-3152

Other

Other2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011
CountryUnited States
CitySan Jose, CA
Period11/7/1111/10/11

All Science Journal Classification (ASJC) codes

  • Software
  • Computer Science Applications
  • Computer Graphics and Computer-Aided Design

Cite this

Yedlapalli, P., Kultursay, E., & Kandemir, M. (2011). Cooperative parallelization. In 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011 (pp. 134-141). [6105317] (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD). https://doi.org/10.1109/ICCAD.2011.6105317
Yedlapalli, Praveen ; Kultursay, Emre ; Kandemir, Mahmut. / Cooperative parallelization. 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. pp. 134-141 (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD).
@inproceedings{c02dbffff59a445ea4589dbeab224e9a,
title = "Cooperative parallelization",
abstract = "We propose a cooperation between the programmer, the compiler and the runtime system to identify, exploit and efficiently exercise the parallelism available in many pointer based applications. Our parallelization strategy, called Cooperative Parallelization, is driven by programmer directives as well as runtime information. We show that minimal information from the programmer can be combined with runtime information to extract latent parallelism in many pointer intensive applications that involve trees and linked lists. We implemented a compilation framework which automatically parallelizes programs annotated with parallelism directives. We evaluated our approach on a collection of linked list and tree based applications. Our results show that we can achieve speedups of up to 15 on a sixteen-core platform. We also compared our approach to OpenMP both qualitatively and quantitatively.",
author = "Praveen Yedlapalli and Emre Kultursay and Mahmut Kandemir",
year = "2011",
month = "12",
day = "1",
doi = "10.1109/ICCAD.2011.6105317",
language = "English (US)",
isbn = "9781457713989",
series = "IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD",
pages = "134--141",
booktitle = "2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011",

}

Yedlapalli, P, Kultursay, E & Kandemir, M 2011, Cooperative parallelization. in 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011., 6105317, IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD, pp. 134-141, 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011, San Jose, CA, United States, 11/7/11. https://doi.org/10.1109/ICCAD.2011.6105317

Cooperative parallelization. / Yedlapalli, Praveen; Kultursay, Emre; Kandemir, Mahmut.

2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. p. 134-141 6105317 (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD).

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

TY - GEN

T1 - Cooperative parallelization

AU - Yedlapalli, Praveen

AU - Kultursay, Emre

AU - Kandemir, Mahmut

PY - 2011/12/1

Y1 - 2011/12/1

N2 - We propose a cooperation between the programmer, the compiler and the runtime system to identify, exploit and efficiently exercise the parallelism available in many pointer based applications. Our parallelization strategy, called Cooperative Parallelization, is driven by programmer directives as well as runtime information. We show that minimal information from the programmer can be combined with runtime information to extract latent parallelism in many pointer intensive applications that involve trees and linked lists. We implemented a compilation framework which automatically parallelizes programs annotated with parallelism directives. We evaluated our approach on a collection of linked list and tree based applications. Our results show that we can achieve speedups of up to 15 on a sixteen-core platform. We also compared our approach to OpenMP both qualitatively and quantitatively.

AB - We propose a cooperation between the programmer, the compiler and the runtime system to identify, exploit and efficiently exercise the parallelism available in many pointer based applications. Our parallelization strategy, called Cooperative Parallelization, is driven by programmer directives as well as runtime information. We show that minimal information from the programmer can be combined with runtime information to extract latent parallelism in many pointer intensive applications that involve trees and linked lists. We implemented a compilation framework which automatically parallelizes programs annotated with parallelism directives. We evaluated our approach on a collection of linked list and tree based applications. Our results show that we can achieve speedups of up to 15 on a sixteen-core platform. We also compared our approach to OpenMP both qualitatively and quantitatively.

UR - http://www.scopus.com/inward/record.url?scp=84855810103&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84855810103&partnerID=8YFLogxK

U2 - 10.1109/ICCAD.2011.6105317

DO - 10.1109/ICCAD.2011.6105317

M3 - Conference contribution

AN - SCOPUS:84855810103

SN - 9781457713989

T3 - IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD

SP - 134

EP - 141

BT - 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011

ER -

Yedlapalli P, Kultursay E, Kandemir M. Cooperative parallelization. In 2011 IEEE/ACM International Conference on Computer-Aided Design, ICCAD 2011. 2011. p. 134-141. 6105317. (IEEE/ACM International Conference on Computer-Aided Design, Digest of Technical Papers, ICCAD). https://doi.org/10.1109/ICCAD.2011.6105317