Multi-Core Video Decoder Using Double Buffering for Parallel Slice Processing

The patent divides the video bitstream into multiple independent slices, where each slice can be decoded independently by different processor cores. This segmentation allows parallel processing while maintaining decoding accuracy, as each slice's decoding does not depend on others. The first core parses and allocates slices to the second core, enabling concurrent processing without compromising reliability.

The patent implements a universal decoding model where the first core performs parsing and allocation functions, while the second core performs decoding and reconstruction functions. This multi-functional design allows different cores to specialize in specific tasks while maintaining overall system versatility, avoiding the complexity of duplicating all functions in each core.

The patent implements continuous transfer of macroblocks from the auxiliary buffer to main memory during the decoding process. The second core continuously stores decoded macroblocks in the auxiliary buffer and transfers them to main memory without interruption, ensuring that memory bandwidth is fully utilized while maintaining continuous decoding operation without idle waiting time.

The patent segments the decoding process into distinct functional units: the first core handles parsing and slice allocation, while the second core handles decoding and reconstruction. This functional division simplifies implementation by assigning specific responsibilities to each core, while simultaneously enabling parallel processing performance through independent operation of these segmented functions.