DocumentCode :
248110
Title :
A highly parallelized H.265/HEVC real-time UHD software encoder
Author :
Tse Kai Heng ; Asano, Wataru ; Itoh, Takayuki ; Tanizawa, Akiyuki ; Yamaguchi, Jun ; Matsuo, Takuya ; Kodama, Tomoya
Author_Institution :
Corp. R&D Center, Toshiba Corp., Kawasaki, Japan
fYear :
2014
fDate :
27-30 Oct. 2014
Firstpage :
1213
Lastpage :
1217
Abstract :
H.265/HEVC standard, promising up to twice the compression efficiency over H.264/AVC standard is suitable for encoding UHD videos and has garnered much attention since its inception. With increasing amount of devices supporting UHD, real-time H.265/HEVC encoder and decoder are needed to complete the UHD media ecosystem. In this paper, we present a highly parallelized HEVC software encoder suitable for broadcasting or network streaming applications. Real-time Main 10 profile encoding of 4K UHD videos at 60fps is achieved through an efficient parallel encoder platform comprising of CPUs interconnected by high speed network. The encoder core is optimized with data parallelism and GPU assisted motion estimation. Utilizing temporal parallelism of GOP and spatial parallelism of picture through slicing to encode, scalability and flexibility are achieved. Results show that our encoder system is about 15794 times faster than HEVC test model HM 11.0 and 13 times faster than ×265, an open source HEVC encoder.
Keywords :
graphics processing units; motion estimation; parallel processing; video coding; GPU assisted motion estimation; H.265-HEVC realtime UHD software encoder; H.265-HEVC standard; UHD media ecosystem; broadcasting application; compression efficiency; data parallelism; graphics processing unit; high efficiency video coding; highly parallelized HEVC software encoder; network streaming application; spatial parallelism; temporal parallelism; Encoding; Graphics processing units; Motion estimation; Parallel processing; Real-time systems; Video coding; Videos; GPU; HEVC; UHD; Video Codec;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Image Processing (ICIP), 2014 IEEE International Conference on
Conference_Location :
Paris
Type :
conf
DOI :
10.1109/ICIP.2014.7025242
Filename :
7025242
Link To Document :
بازگشت