This paper presents a new edge extraction based compressed domain deblocking algorithm and its ASIP (Application specific instruction processor) architecture for video decoding based on algorithm and architecture co-design methodology. Our algorithm works very stable and robust no mater under low bit rate compression or high bit rate compression scenarios. The VLIW (Very long instruction world) based ASIP implementation of our algorithm proves high performance with very limited hardware cost. The Algorithm and architecture co-design (AAC) concept is highly emphasized in this paper. We provide some quantitative example to show the necessity of algorithm and architecture co-design.