Design, implementation and evaluation of a task-parallel JPEG decoder for the Libjpeg-turbo library

Jingun Hong, Wasuwee Sodsong, Seongwook Chung, Cheong Ghil Kim, Yeongkyu Lim, Shin-Dug Kim, bernd Burgstaller

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

In this paper, we propose a task-parallel programming extension for the JPEG decoder of the libjpeg-turbo library. Efficient JPEG decoding is especially important for resource-constrained mobile devices such as smartphones, where decoding (e.g., browsing of web pages containing images, image search aso) is far more common than image encoding. The aim of our work is to utilize multiple CPU cores for JPEG decompression from a single client thread. Our method is orthogonal to libjpeg-turbo's support for data-parallelism (SIMD). Experimental evaluation of our approach on a 4-core Intel i7-2600K CPU shows speed-ups of up to 2.5x over the sequential, and up to 34% over the SIMD-version of the libjpeg-turbo JPEG decoder.

Original languageEnglish
Pages (from-to)147-152
Number of pages6
JournalInternational Journal of Multimedia and Ubiquitous Engineering
Volume7
Issue number2
Publication statusPublished - 2012 Sep 6

Fingerprint

Program processors
Decoding
Parallel programming
Smartphones
Mobile devices
Websites

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

@article{171a927649d54ba691ba6c866a91a233,
title = "Design, implementation and evaluation of a task-parallel JPEG decoder for the Libjpeg-turbo library",
abstract = "In this paper, we propose a task-parallel programming extension for the JPEG decoder of the libjpeg-turbo library. Efficient JPEG decoding is especially important for resource-constrained mobile devices such as smartphones, where decoding (e.g., browsing of web pages containing images, image search aso) is far more common than image encoding. The aim of our work is to utilize multiple CPU cores for JPEG decompression from a single client thread. Our method is orthogonal to libjpeg-turbo's support for data-parallelism (SIMD). Experimental evaluation of our approach on a 4-core Intel i7-2600K CPU shows speed-ups of up to 2.5x over the sequential, and up to 34{\%} over the SIMD-version of the libjpeg-turbo JPEG decoder.",
author = "Jingun Hong and Wasuwee Sodsong and Seongwook Chung and Kim, {Cheong Ghil} and Yeongkyu Lim and Shin-Dug Kim and bernd Burgstaller",
year = "2012",
month = "9",
day = "6",
language = "English",
volume = "7",
pages = "147--152",
journal = "International Journal of Multimedia and Ubiquitous Engineering",
issn = "1975-0080",
publisher = "Science and Engineering Research Support Society",
number = "2",

}

Design, implementation and evaluation of a task-parallel JPEG decoder for the Libjpeg-turbo library. / Hong, Jingun; Sodsong, Wasuwee; Chung, Seongwook; Kim, Cheong Ghil; Lim, Yeongkyu; Kim, Shin-Dug; Burgstaller, bernd.

In: International Journal of Multimedia and Ubiquitous Engineering, Vol. 7, No. 2, 06.09.2012, p. 147-152.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Design, implementation and evaluation of a task-parallel JPEG decoder for the Libjpeg-turbo library

AU - Hong, Jingun

AU - Sodsong, Wasuwee

AU - Chung, Seongwook

AU - Kim, Cheong Ghil

AU - Lim, Yeongkyu

AU - Kim, Shin-Dug

AU - Burgstaller, bernd

PY - 2012/9/6

Y1 - 2012/9/6

N2 - In this paper, we propose a task-parallel programming extension for the JPEG decoder of the libjpeg-turbo library. Efficient JPEG decoding is especially important for resource-constrained mobile devices such as smartphones, where decoding (e.g., browsing of web pages containing images, image search aso) is far more common than image encoding. The aim of our work is to utilize multiple CPU cores for JPEG decompression from a single client thread. Our method is orthogonal to libjpeg-turbo's support for data-parallelism (SIMD). Experimental evaluation of our approach on a 4-core Intel i7-2600K CPU shows speed-ups of up to 2.5x over the sequential, and up to 34% over the SIMD-version of the libjpeg-turbo JPEG decoder.

AB - In this paper, we propose a task-parallel programming extension for the JPEG decoder of the libjpeg-turbo library. Efficient JPEG decoding is especially important for resource-constrained mobile devices such as smartphones, where decoding (e.g., browsing of web pages containing images, image search aso) is far more common than image encoding. The aim of our work is to utilize multiple CPU cores for JPEG decompression from a single client thread. Our method is orthogonal to libjpeg-turbo's support for data-parallelism (SIMD). Experimental evaluation of our approach on a 4-core Intel i7-2600K CPU shows speed-ups of up to 2.5x over the sequential, and up to 34% over the SIMD-version of the libjpeg-turbo JPEG decoder.

UR - http://www.scopus.com/inward/record.url?scp=84865634863&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865634863&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84865634863

VL - 7

SP - 147

EP - 152

JO - International Journal of Multimedia and Ubiquitous Engineering

JF - International Journal of Multimedia and Ubiquitous Engineering

SN - 1975-0080

IS - 2

ER -