Synthesis and Comparison of Low-Power Architectures for SAD Calculation

This paper presents the standard-cells synthesis and comparison of parallel hardware architectures for the Sum of Absolute Differences (SAD) datapath, focusing on different design points such as the tradeoff between high-performance and low-power dissipation (isoperformance target). Multi-VDD, multi-VT and different combination of parallelism and pipeline architectural techniques were explored in this work. In order to generate the results, we used the IBM 65nm standard-cells library with typical voltage of 1 V and 1.2 V, and the back-end Cadence tools, e.g. Power Analysis, for the power measurements. We achieved significant power reduction for the architectures with low-frequency and high parallelism, with High-VT and mainly with only one pipeline stage and small power source.

[1]  Wang Qin,et al.  A High-Performance Low Cost SAD Architecture for Video Coding , 2007, IEEE Transactions on Consumer Electronics.

[2]  Sergio Bampi,et al.  Synthesis and comparison of low-power high-throughput architectures for SAD calculation , 2011, 2011 IEEE Second Latin American Symposium on Circuits and Systems (LASCAS).

[3]  Timo Hämäläinen,et al.  A High-Performance Sum of Absolute Difference Implementation for Motion Estimation , 2006, IEEE Transactions on Circuits and Systems for Video Technology.

[4]  Yang Song,et al.  32-Parallel SAD Tree Hardwired Engine for Variable Block Size Motion Estimation in HDTV1080P Real-Time Encoding Application , 2007, 2007 IEEE Workshop on Signal Processing Systems.

[5]  Sergio Bampi,et al.  Comparative analysis of parallel SAD calculation hardware architectures for H.264/AVC video coding , 2010, 2010 First IEEE Latin American Symposium on Circuits and Systems (LASCAS).

[6]  Liang-Gee Chen,et al.  Analysis and architecture design of an HDTV720p 30 frames/s H.264/AVC encoder , 2006, IEEE Transactions on Circuits and Systems for Video Technology.