Collective affinity aware computation mapping

Mahmut T. Kandemir; Jihyun Ryoo; Hui Zhao; Myoungsoo Jung; Mustafa Karakoy

doi:10.1145/3410463.3414661

Collective affinity aware computation mapping

Mahmut T. Kandemir, Jihyun Ryoo, Hui Zhao, Myoungsoo Jung, Mustafa Karakoy

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

This work defines the concept of collective affinity. It is claimedthat collective affinity has more potential than single core-centricaffinity, for data locality optimization in manycores. The reason isthat collective affinity captures the potential benefits of transferringcomputations originally assigned to one core to other cores. Next,building upon the collective affinity concept and a cache contentestimation strategy, it presents a computation-to-core mappingstrategy, specifically tuned for exploiting near data computing byreducing distance-to-data.

Original language	English (US)
Title of host publication	PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	343-344
Number of pages	2
ISBN (Electronic)	9781450380751
DOIs	https://doi.org/10.1145/3410463.3414661
State	Published - Sep 30 2020
Event	2020 ACM International Conference on Parallel Architectures and Compilation Techniques, PACT 2020 - Virtual, Online, United States Duration: Oct 3 2020 → Oct 7 2020

Publication series

Name	Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT
ISSN (Print)	1089-795X

Conference

Conference	2020 ACM International Conference on Parallel Architectures and Compilation Techniques, PACT 2020
Country/Territory	United States
City	Virtual, Online
Period	10/3/20 → 10/7/20

All Science Journal Classification (ASJC) codes

Software
Theoretical Computer Science
Hardware and Architecture

Access to Document

10.1145/3410463.3414661

Cite this

Kandemir, M. T., Ryoo, J., Zhao, H., Jung, M., & Karakoy, M. (2020). Collective affinity aware computation mapping. In PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques (pp. 343-344). (Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1145/3410463.3414661

Kandemir, Mahmut T. ; Ryoo, Jihyun ; Zhao, Hui et al. / Collective affinity aware computation mapping. PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 343-344 (Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT).

@inproceedings{15b0059dd75047bc8e79841db2c8e196,

title = "Collective affinity aware computation mapping",

abstract = "This work defines the concept of collective affinity. It is claimedthat collective affinity has more potential than single core-centricaffinity, for data locality optimization in manycores. The reason isthat collective affinity captures the potential benefits of transferringcomputations originally assigned to one core to other cores. Next,building upon the collective affinity concept and a cache contentestimation strategy, it presents a computation-to-core mappingstrategy, specifically tuned for exploiting near data computing byreducing distance-to-data.",

author = "Kandemir, {Mahmut T.} and Jihyun Ryoo and Hui Zhao and Myoungsoo Jung and Mustafa Karakoy",

note = "Publisher Copyright: {\textcopyright} 2020 Association for Computing Machinery.; 2020 ACM International Conference on Parallel Architectures and Compilation Techniques, PACT 2020 ; Conference date: 03-10-2020 Through 07-10-2020",

year = "2020",

month = sep,

day = "30",

doi = "10.1145/3410463.3414661",

language = "English (US)",

series = "Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "343--344",

booktitle = "PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques",

address = "United States",

}

Kandemir, MT, Ryoo, J, Zhao, H, Jung, M & Karakoy, M 2020, Collective affinity aware computation mapping. in PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques. Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT, Institute of Electrical and Electronics Engineers Inc., pp. 343-344, 2020 ACM International Conference on Parallel Architectures and Compilation Techniques, PACT 2020, Virtual, Online, United States, 10/3/20. https://doi.org/10.1145/3410463.3414661

Collective affinity aware computation mapping. / Kandemir, Mahmut T.; Ryoo, Jihyun; Zhao, Hui et al.
PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques. Institute of Electrical and Electronics Engineers Inc., 2020. p. 343-344 (Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Collective affinity aware computation mapping

AU - Kandemir, Mahmut T.

AU - Ryoo, Jihyun

AU - Zhao, Hui

AU - Jung, Myoungsoo

AU - Karakoy, Mustafa

PY - 2020/9/30

Y1 - 2020/9/30

N2 - This work defines the concept of collective affinity. It is claimedthat collective affinity has more potential than single core-centricaffinity, for data locality optimization in manycores. The reason isthat collective affinity captures the potential benefits of transferringcomputations originally assigned to one core to other cores. Next,building upon the collective affinity concept and a cache contentestimation strategy, it presents a computation-to-core mappingstrategy, specifically tuned for exploiting near data computing byreducing distance-to-data.

AB - This work defines the concept of collective affinity. It is claimedthat collective affinity has more potential than single core-centricaffinity, for data locality optimization in manycores. The reason isthat collective affinity captures the potential benefits of transferringcomputations originally assigned to one core to other cores. Next,building upon the collective affinity concept and a cache contentestimation strategy, it presents a computation-to-core mappingstrategy, specifically tuned for exploiting near data computing byreducing distance-to-data.

UR - http://www.scopus.com/inward/record.url?scp=85094213662&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85094213662&partnerID=8YFLogxK

U2 - 10.1145/3410463.3414661

DO - 10.1145/3410463.3414661

M3 - Conference contribution

AN - SCOPUS:85094213662

T3 - Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT

SP - 343

EP - 344

BT - PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 ACM International Conference on Parallel Architectures and Compilation Techniques, PACT 2020

Y2 - 3 October 2020 through 7 October 2020

ER -

Kandemir MT, Ryoo J, Zhao H, Jung M, Karakoy M. Collective affinity aware computation mapping. In PACT 2020 - Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques. Institute of Electrical and Electronics Engineers Inc. 2020. p. 343-344. (Parallel Architectures and Compilation Techniques - Conference Proceedings, PACT). doi: 10.1145/3410463.3414661

Collective affinity aware computation mapping

Abstract

Publication series

Conference

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this