As mobile displays proliferate across the world, and fixed video displays expand in size, consumers demand enhanced visual experiences while watching such displays. Competitive low-end to high-end markets are migrating to high-definition (HD) output displays driven by consumer demand. In order to generate high quality visual experiences on HD displays, video processing algorithms are becoming increasingly complex. Such complex HD algorithms require up to tera-operations per second to generate acceptable HD video output. Silicon Hive's HiveFlex Video VSP1 processor services the needs all low to high end video displays by using a unique scaleable tiled architecture. The architecture of each tile is flexible, low-cost and low-power, resulting in an attractive IP solution targeting consumer video signal processing (VSP) systems on chips (SoCs). The Video VSPI tile presented here performs a variety of video coding algorithms commonly used in post processing of HDTV signals such as H.264 decoding, de-interlacing, picture-rate up-conversion and others. H.264 or MPEG4-AVC is an advanced video coding standard targeting multiple markets such as broadcast and hand-held entertainment. In partnership with Silicon Hive, AllGo Embedded Systems has developed an H.264 decoder that is optimized for the HiveFlex VSP architecture. Multiple VSP tiles with configurable number of issue slots and SIMD architecture are used efficiently to implement H.264 video decoding at HD resolution
[1]
E. Salami,et al.
A performance characterization of high definition digital video decoding using H.264/AVC
,
2005,
IEEE International. 2005 Proceedings of the IEEE Workload Characterization Symposium, 2005..
[2]
G. de Haan,et al.
Tackling occlusion in scan rate conversion systems
,
2003,
2003 IEEE International Conference on Consumer Electronics, 2003. ICCE..
[3]
A. Beric,et al.
Heterogeneous multiprocessor for high definition video
,
2006,
2006 Digest of Technical Papers International Conference on Consumer Electronics.
[4]
Gerard de Haan,et al.
A two-dimensional generalized sampling theory and application to de-interlacing
,
2004,
IS&T/SPIE Electronic Imaging.