Design of energy-efficient discrete cosine transform using pruned arithmetic circuits

Inexact circuits and approximate computing have been gaining a lot of interest in order to improve performances and energy efficiency beyond the boundaries of conventional digital circuits. Image and video processing is one of the best candidate for applying such techniques. As one of the key building blocks, Discrete Cosine Transform (DCT) accelerators are investigated using pruned arithmetic circuits. A design methodology is presented in order to optimize both image quality and circuit performances. This work demonstrates that with such technique, savings are possible not only on arithmetic units, but in the entire accelerator hardware. Simulations show up to 12 % area and 10 % power savings with less than 20 dB PSNR degradation compared to the conventional DCT design.

[1]  John Augustine,et al.  Opportunities for energy efficient computing: A study of inexact general purpose processors for high-performance and big-data applications , 2015, 2015 Design, Automation & Test in Europe Conference & Exhibition (DATE).

[2]  Christian Piguet,et al.  Why design reliable chips when faulty ones are even better , 2013, 2013 Proceedings of the ESSCIRC (ESSCIRC).

[3]  Kaushik Roy,et al.  CRISTA: A New Paradigm for Low-Power, Variation-Tolerant, and Adaptive Circuit Synthesis Using Critical Path Isolation , 2007, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[4]  Lingamneni Avinash,et al.  Parsimonious Circuits for Error-Tolerant Applications through Probabilistic Logic Minimization , 2011, PATMOS.

[5]  Richard M. Karp,et al.  Algorithmic methodologies for ultra-efficient inexact architectures for sustaining technology scaling , 2012, CF '12.

[6]  Christian C. Enz,et al.  Energy-efficient inexact speculative adder with high performance and accuracy control , 2015, 2015 IEEE International Symposium on Circuits and Systems (ISCAS).

[7]  Earl E. Swartzlander,et al.  DCT Implementation with Distributed Arithmetic , 2001, IEEE Trans. Computers.

[8]  Krishna V. Palem,et al.  Automatic generation of inexact digital circuits by gate-level pruning , 2015, 2015 IEEE International Symposium on Circuits and Systems (ISCAS).

[9]  Kaushik Roy,et al.  Low-Power Digital Signal Processing Using Approximate Adders , 2013, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[10]  Krishna V. Palem,et al.  Energy aware computing through probabilistic switching: a study of limits , 2005, IEEE Transactions on Computers.

[11]  David Blaauw,et al.  Razor: A Low-Power Pipeline Based on Circuit-Level Timing Speculation , 2003, MICRO.