An Energy-efficient Parallel H.264/AVC Baseline Encoder on a Fine-grained Many-core System

Zhibin Xiao
Stephen, Le
Bevan Baas
VLSI Computation Laboratory
Department of Electrical and Computer Engineering
University of California, Davis

Abstract

The emerging many-core architecture provides a flexible solution for the rapid evolving multimedia applications demanding both high performance and high energy-efficiency. However, developing parallel multimedia applications that can efficiently harness and utilize many-core architectures is the key challenge for scalable computing. We contribute to this challenge by presenting a fully-parallel H.264/AVC baseline encoder on a 167-core asynchronous array of simple processors(AsAP) computation platform. By exploiting fine-grained data and task level parallelism in the algorithms, we partition and map the dataflow of the H.264/AVC encoder to an array of 115 small processors coupled with two shared memories and a hardware accelerator for motion estimation. Due to the large number of independent processors available, the video encoding process can be divided into three main stages: prediction, entropy encoding, and reconstruction, with the entropy encoding and reconstruction stages done in parallel and pipelined with the prediction stage. Within each stage, each independent procedure is mapped onto an individual processor for greater parallelization and efficiency. The proposed parallel H.264/AVC encoder is capable of encoding video sequences with variable frame sizes. The preliminary implemenation is capable of encoding CIF (352x288) video at 54 frames per second (fps) with 925 mW average power consumption by adjusting each processor to workload-based optimal clock frequencies and dual supply voltages with less than 1dB loss in resolution.

Paper

Poster Slides

Reference

Zhibin Xiao, Stephen Le, Bevan M. Baas, "An Energy-Efficient Parallel H.264/AVC Baseline Encoder on a Fine-grained Many-core System", Technology and Talent for the 21st Century (TECHCON 2010), Austin, TX, Sep. 2010.

BibTeX Entry

@INPROCEEDINGS{H264:TECHCON2010,
   author={Zhibin Xiao, Stephen Le, Bevan M. Baas},
   booktitle={Technology and Talent for the 21st Century (TECHCON 2010)},
   title={An Energy-Efficient Parallel H.264/AVC Baseline Encoder on a Fine-grained Many-core System},
   year={2010},
   month={Sep.}
}
VCL Lab | ECE Dept. | UC Davis

Last update: Jun 17, 2010