Heterogeneous acceleration of volumetric JPEG 2000 using OpenCL

Jan G. Cornelis; Jan Lemeire; Tim Bruylants; Peter Schelkens

doi:10.1177/1094342016646438

Heterogeneous acceleration of volumetric JPEG 2000 using OpenCL

Jan G. Cornelis, Jan Lemeire, Tim Bruylants, Peter Schelkens

Onderzoeksoutput: Article › peer review

223 Downloads (Pure)

Samenvatting

This paper discusses an OpenCL version of a volumetric JPEG 2000 codec that runs on GPUs, multi-core processors or a combination of both. Since the performance critical part consists of a fine-grained (discrete wavelet transform) and coarse-grained algorithm (Tier-1), the best performance is obtained with a hybrid execution in which the discrete wavelet transform is executed on a GPU and Tier-1 on a multi-core. Using an Intel i7 multi-core in combination with a modest NVIDIA Quadro K620 GPU yields speedups greater than 10 compared with the original sequential code. The performance bottlenecks that arise on GPUs when parallelizing algorithms that are coarse-grained by nature are discussed and also the optimizations that are possible. A performance analysis reveals the inefficiencies and explains the deviations from the GPU peak performance.

Originele taal-2	English
Pagina's (van-tot)	229-245
Aantal pagina's	17
Tijdschrift	International Journal of High Performance Computing Applications
Volume	31
Nummer van het tijdschrift	3
Vroegere onlinedatum	10 mei 2016
DOI's	https://doi.org/10.1177/1094342016646438
Status	Published - 1 mei 2017

Toegang tot document

10.1177/1094342016646438Licentie: Unspecified

ijhpca_final accepted versionAccepted author manuscript, 782 KBLicentie: Unspecified

Handle.net

Permanente koppeling

Andere bestanden en links

Link to publication in Scopus

Document onder embargo

Cornelis IJHPCA 2016
Final published version, 2,03 MB

EU465: Ijle signaalcodering voor interferentie-gebaseerde beeldvormingsmodaliteiten (INTERFERE)
Schelkens, P. (Administrative Promotor) & Schelkens, P. (Coördinator)
1/06/14 → 31/05/19
Project: Fundamenteel
SRP11: SRP (Zwaartepunt): Verwerking van grootschalige multi-dimensionale, multi-spectrale, multi-sensoriële en gedistribueerde gegevens (M³D²)
Schelkens, P. (Administrative Promotor), Deligiannis, N. (Co-Promoter), Jansen, B. (Co-Promoter), Kuijk, M. (Co-Promoter), Munteanu, A. (Co-Promoter), Sahli, H. (Co-Promoter), Steenhaut, K. (Co-Promoter), Stiens, J. (Co-Promoter), Schelkens, P. (Administrative Promotor), Cornelis, J. P. (Co-Promoter), Kuijk, M. (Co-Promoter), Munteanu, A. (Co-Promoter), Sahli, H. (Co-Promoter), Stiens, J. (Co-Promoter) & Vounckx, R. (Co-Promoter)
1/11/12 → 31/12/23
Project: Fundamenteel

Citeer dit

@article{66f7e98efb4a414cb8d65c0cf5c45dcb,

title = "Heterogeneous acceleration of volumetric JPEG 2000 using OpenCL",

abstract = "This paper discusses an OpenCL version of a volumetric JPEG 2000 codec that runs on GPUs, multi-core processors or a combination of both. Since the performance critical part consists of a fine-grained (discrete wavelet transform) and coarse-grained algorithm (Tier-1), the best performance is obtained with a hybrid execution in which the discrete wavelet transform is executed on a GPU and Tier-1 on a multi-core. Using an Intel i7 multi-core in combination with a modest NVIDIA Quadro K620 GPU yields speedups greater than 10 compared with the original sequential code. The performance bottlenecks that arise on GPUs when parallelizing algorithms that are coarse-grained by nature are discussed and also the optimizations that are possible. A performance analysis reveals the inefficiencies and explains the deviations from the GPU peak performance. ",

keywords = "GPU, Hybrid, multi-core, OpenCL, volumetric JPEG 2000",

author = "Cornelis, {Jan G.} and Jan Lemeire and Tim Bruylants and Peter Schelkens",

year = "2017",

month = may,

day = "1",

doi = "10.1177/1094342016646438",

language = "English",

volume = "31",

pages = "229--245",

journal = "International Journal of High Performance Computing Applications",

issn = "1094-3420",

publisher = "SAGE Publications Ltd",

number = "3",

}

TY - JOUR

T1 - Heterogeneous acceleration of volumetric JPEG 2000 using OpenCL

AU - Cornelis, Jan G.

AU - Lemeire, Jan

AU - Bruylants, Tim

AU - Schelkens, Peter

PY - 2017/5/1

Y1 - 2017/5/1

N2 - This paper discusses an OpenCL version of a volumetric JPEG 2000 codec that runs on GPUs, multi-core processors or a combination of both. Since the performance critical part consists of a fine-grained (discrete wavelet transform) and coarse-grained algorithm (Tier-1), the best performance is obtained with a hybrid execution in which the discrete wavelet transform is executed on a GPU and Tier-1 on a multi-core. Using an Intel i7 multi-core in combination with a modest NVIDIA Quadro K620 GPU yields speedups greater than 10 compared with the original sequential code. The performance bottlenecks that arise on GPUs when parallelizing algorithms that are coarse-grained by nature are discussed and also the optimizations that are possible. A performance analysis reveals the inefficiencies and explains the deviations from the GPU peak performance.

AB - This paper discusses an OpenCL version of a volumetric JPEG 2000 codec that runs on GPUs, multi-core processors or a combination of both. Since the performance critical part consists of a fine-grained (discrete wavelet transform) and coarse-grained algorithm (Tier-1), the best performance is obtained with a hybrid execution in which the discrete wavelet transform is executed on a GPU and Tier-1 on a multi-core. Using an Intel i7 multi-core in combination with a modest NVIDIA Quadro K620 GPU yields speedups greater than 10 compared with the original sequential code. The performance bottlenecks that arise on GPUs when parallelizing algorithms that are coarse-grained by nature are discussed and also the optimizations that are possible. A performance analysis reveals the inefficiencies and explains the deviations from the GPU peak performance.

KW - GPU

KW - Hybrid

KW - multi-core

KW - OpenCL

KW - volumetric JPEG 2000

UR - http://www.scopus.com/inward/record.url?scp=85019932776&partnerID=8YFLogxK

U2 - 10.1177/1094342016646438

DO - 10.1177/1094342016646438

M3 - Article

SN - 1094-3420

VL - 31

SP - 229

EP - 245

JO - International Journal of High Performance Computing Applications

JF - International Journal of High Performance Computing Applications

IS - 3

ER -

Heterogeneous acceleration of volumetric JPEG 2000 using OpenCL

Samenvatting

Toegang tot document

Handle.net

Andere bestanden en links

Document onder embargo

Vingerafdruk

Projecten

EU465: Ijle signaalcodering voor interferentie-gebaseerde beeldvormingsmodaliteiten (INTERFERE)

SRP11: SRP (Zwaartepunt): Verwerking van grootschalige multi-dimensionale, multi-spectrale, multi-sensoriële en gedistribueerde gegevens (M³D²)

Citeer dit