Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2003 · 76 citations · 1 references
EngineeringComputer EngineeringComputer ScienceComputer ArchitectureHardware Description LanguagesHardware-software Co-designData CompressionHardware AccelerationImage CompressionInstruction SchedulingSource CodingCoding StandardsDecoder DesignMotion CompensationSequence DecodingVideo Compression
H.264 targets high‑quality video at low bit‑rates but demands significantly more computation than earlier standards. The study aims to analyze and accelerate a real‑time H.264 decoder on general‑purpose processors, particularly Intel Pentium 4, and to offer design recommendations for future video/image applications. The authors dissect the reference implementation to pinpoint bottlenecks, then apply media‑instruction optimizations to key modules and discuss implications for efficient hardware‑aware design. Optimizations reduced the most time‑consuming modules—motion compensation and inverse integer transform—yielding a 3× speedup to 48 fps at 720×480 on a 2.4 GHz Pentium 4, and the techniques generalize to other image/video processing tasks.
As emerging video coding standards, e.g. H.264, aim at high-quality video contents at low bit-rates, the encoding and decoding processes require much more computation than most existing standards do. This paper analyzes software implementation of a real-time H.264 decoder on general-purpose processors with media instructions. Specifically, we discuss how to optimize the speed of H.264 decoders on Intel Pentium 4 processors. This paper first analyzes the reference implementation to identify the time-consuming modules. Our study shows that a number of components, e.g., motion compensation and inverse integer transform, are the most time-consuming modules in the H.264 decoder. Second, we present a list of performance optimization methods using media instructions to improve the efficiency of these modules. After appropriate optimizations, the decoder speed improved by more than 3x---it can decode a 720×480 resolution video sequence at 48 frames per second on 2.4GHz Intel Pentium 4 processors compared to reference software's 12 frames per second. The optimization techniques demonstrated in this paper can also be applied to other video/image processing applications. Additionally, after presenting detailed application behavior on general-purpose processors, this paper discusses a few recommendations on how to design future efficient/powerful video/image applications/standards with given hardware implications.
1