Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids

Changmin Lee, Won Woo Ro, Jean Luc Gaudiot

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

14 Citations (Scopus)

Abstract

This paper presents a cooperative heterogeneous computing framework which enables the efficient utilization of available computing resources of host CPU cores for CUDA kernels, which are designed to run only on GPU. The proposed system exploits at runtime the coarse-grain thread-level parallelism across CPU and GPU, without any source recompilation. To this end, three features including a work distribution module, a transparent memory space, and a global scheduling queue are described in this paper. With a completely automatic runtime workload distribution, the proposed framework achieves speedups as high as 3.08 compared to the baseline GPU-only processing.

Original languageEnglish
Title of host publicationProceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012
Pages33-40
Number of pages8
DOIs
Publication statusPublished - 2012 Dec 20
Event16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012 - New Orleans, LA, United States
Duration: 2012 Feb 252012 Feb 25

Publication series

NameProceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT
ISSN (Print)1550-6207

Other

Other16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012
CountryUnited States
CityNew Orleans, LA
Period12/2/2512/2/25

Fingerprint

Program processors
Processing
Scheduling
Data storage equipment
Graphics processing unit

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Lee, C., Ro, W. W., & Gaudiot, J. L. (2012). Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids. In Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012 (pp. 33-40). [6339624] (Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT). https://doi.org/10.1109/INTERACT.2012.6339624
Lee, Changmin ; Ro, Won Woo ; Gaudiot, Jean Luc. / Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids. Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012. 2012. pp. 33-40 (Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT).
@inproceedings{16adeba7153c45e29bca1b9dd3d9f10a,
title = "Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids",
abstract = "This paper presents a cooperative heterogeneous computing framework which enables the efficient utilization of available computing resources of host CPU cores for CUDA kernels, which are designed to run only on GPU. The proposed system exploits at runtime the coarse-grain thread-level parallelism across CPU and GPU, without any source recompilation. To this end, three features including a work distribution module, a transparent memory space, and a global scheduling queue are described in this paper. With a completely automatic runtime workload distribution, the proposed framework achieves speedups as high as 3.08 compared to the baseline GPU-only processing.",
author = "Changmin Lee and Ro, {Won Woo} and Gaudiot, {Jean Luc}",
year = "2012",
month = "12",
day = "20",
doi = "10.1109/INTERACT.2012.6339624",
language = "English",
isbn = "9781467326148",
series = "Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT",
pages = "33--40",
booktitle = "Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012",

}

Lee, C, Ro, WW & Gaudiot, JL 2012, Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids. in Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012., 6339624, Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT, pp. 33-40, 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012, New Orleans, LA, United States, 12/2/25. https://doi.org/10.1109/INTERACT.2012.6339624

Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids. / Lee, Changmin; Ro, Won Woo; Gaudiot, Jean Luc.

Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012. 2012. p. 33-40 6339624 (Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT).

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

TY - GEN

T1 - Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids

AU - Lee, Changmin

AU - Ro, Won Woo

AU - Gaudiot, Jean Luc

PY - 2012/12/20

Y1 - 2012/12/20

N2 - This paper presents a cooperative heterogeneous computing framework which enables the efficient utilization of available computing resources of host CPU cores for CUDA kernels, which are designed to run only on GPU. The proposed system exploits at runtime the coarse-grain thread-level parallelism across CPU and GPU, without any source recompilation. To this end, three features including a work distribution module, a transparent memory space, and a global scheduling queue are described in this paper. With a completely automatic runtime workload distribution, the proposed framework achieves speedups as high as 3.08 compared to the baseline GPU-only processing.

AB - This paper presents a cooperative heterogeneous computing framework which enables the efficient utilization of available computing resources of host CPU cores for CUDA kernels, which are designed to run only on GPU. The proposed system exploits at runtime the coarse-grain thread-level parallelism across CPU and GPU, without any source recompilation. To this end, three features including a work distribution module, a transparent memory space, and a global scheduling queue are described in this paper. With a completely automatic runtime workload distribution, the proposed framework achieves speedups as high as 3.08 compared to the baseline GPU-only processing.

UR - http://www.scopus.com/inward/record.url?scp=84871138761&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84871138761&partnerID=8YFLogxK

U2 - 10.1109/INTERACT.2012.6339624

DO - 10.1109/INTERACT.2012.6339624

M3 - Conference contribution

SN - 9781467326148

T3 - Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT

SP - 33

EP - 40

BT - Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012

ER -

Lee C, Ro WW, Gaudiot JL. Cooperative heterogeneous computing for parallel processing on CPU/GPU hybrids. In Proceedings - 16th Workshop on Interaction Between Compilers and Computer Architectures, INTERACT 2012. 2012. p. 33-40. 6339624. (Proceedings - Annual Workshop on Interaction between Compilers and Computer Architectures, INTERACT). https://doi.org/10.1109/INTERACT.2012.6339624