Efficient Parallel Processing of Multimedia Applications on Multi-core Architectures

The well-known wave-front parallelization is proposed for parallel H.264/AVC videoprocessing. Under this approach, groups of independent macro-blocks (MBs) are simultaneously processed, one group after another. Barrier mechanism is employedto synchronize processing of the independent MBs. This approach, however, has asubstantial synchronization overhead that significantly affects the throughput performance. A novel dynamic scheduling scheme with recursive tail submit provides agood throughput performance by exploiting macro-block level parallelism and alleviating the synchronization overhead and thread contention. Nevertheless, it fails toachieve an optimal performance due to the use of a global queue, and an unawarenessof cache locality of the underlying multi-core architecture. I propose an adaptive dynamic scheduling scheme that employs distribued queues, and dynamically schedulestasks in a cache locality-aware and load-balancing fashion.As a graphics accelerator, GPGPU is able to off-loads compute intensive functions. In H.264 video encoding, hierarchical search is a widely proposed for the mostexpensive motion estimation. GPGPU is suitable, especially with full search-basedapproaches as the process can be efficiently parallelized. However, their fixed pyramidstructure lacks a mechanism to select the best multiple-candidate schemes considering diverse video encoding characteristics. I propose profiled-based fixed multiple candidate motion vector selection scheme, and an efficient dynamic multiple candidatemotion vector selection scheme to dynamically select best multiple-candidate motionvector schemes at runtime