ALGORITHMIC AND ARCHITECTURAL ENHANCEMENTS FOR REAL-TIME MPEG-1 DECODING ON A GENERAL-PURPOSE RISC WORKSTATION

Traditional video decoders require use of specially designed video dec ompression processors. We present novel algorithmic and architectural enhancements that allowed for the first time the real-time decompressi on of MPEG-1 video and audio streams on a low-end, general purpose RIS C processor, For video decompression, efficient algorithmic implementa tions were derived by examining the Huffman decoder, the inverse quant izer and the inverse DCT as a single system, For audio decompression, a new DCT based implementation of the subband filtering operation yiel ds 30% speed improvement in the audio decoding process and 17% speed i mprovement in overall audio and video decoding, Besides algorithmic en hancements, a new set of ''multimedia'' instructions and minor changes in the design of a traditional RISC ALU allowed increased parallelism of pixel-based operations with minimal design and control overhead, E xperimental results show that with the synergistic combination of algo rithmic and architectural enhancements a multimedia-enhanced RISC proc essor can achieve higher decoding rates than generic RISC and CISC pro cessors, even when these processors operate at higher clock rates and have larger instruction and data caches.