Please use this identifier to cite or link to this item:
Main Title: Parallel H.264/AVC motion compensation for GPUs using OpenCL
Author(s): Wang, Biao
Álvarez-Mesa, Mauricio
Chi, Chi Ching
Juurlink, Ben
Type: Article
Language Code: en
Abstract: Motion compensation is one of the most compute-intensive parts in H.264/AVC video decoding. It exposes massive parallelism, which can reap the benefit from graphics processing units (GPUs). Control and memory divergence, however, may lead to performance penalties on GPUs. In this paper, we propose two GPU motion-compensation kernels, implemented with OpenCL, that mitigate the divergence effect. In addition, the motion-compensation kernels have been integrated into a complete and optimized H.264/AVC decoder that supports high-profile H.264/AVC. We evaluated our kernels on GPUs with different architectures from AMD, Intel, and Nvidia. Compared with the fastest CPU used in this paper, our kernel achieves 2.0 speedup on a discrete Nvidia GPU at kernel level. However, when the overheads of memory copy and OpenCL runtime are included, no speedup is gained at application level.
Issue Date: 2015
Date Available: 23-Feb-2017
DDC Class: 004 Datenverarbeitung; Informatik
620 Ingenieurwissenschaften und zugeordnete Tätigkeiten
Subject(s): GPU
graphics processing unit (GPU)
motion compensation
Sponsor/Funder: EC/FP7/288653/EU/Low-Power Parallel Computing on GPUs/LPGPU
Journal Title: IEEE transactions on circuits and systems for video technology : a publication of the Circuits and Systems Society
Publisher: IEEE
Publisher Place: New York, NY [u.a.]
Volume: 25
Issue: 3
Publisher DOI: 10.1109/TCSVT.2014.2344512
Page Start: 525
Page End: 531
EISSN: 1558-2205
ISSN: 1051-8215
Appears in Collections:FG Architektur eingebetteter Systeme » Publications

Files in This Item:
File Description SizeFormat 
Wang_et-al_2015.pdf1.64 MBAdobe PDFThumbnail

Items in DepositOnce are protected by copyright, with all rights reserved, unless otherwise indicated.