Bit-width-aware constant-delay run-time Accuracy Programmable Adder for error-resilient applications

Abstract Error-resilient applications such as image, audio and video processing adopt the concept of approximate computing to decorate delay, power and area metrics at the cost of accuracy. Approximate computing relaxes the exact equivalence between the design specifications and the design implementation to achieve Speed-Power-Accuracy-Area (SPAA) trade-off. In this paper, we propose a bit-width-aware constant-delay run-time Accuracy Programmable Adder (APA) in which the probability of input combinations exhibiting accurate results is programmable and adaptively controlled by the Number of Iterations (NoIs). Simulation results based on the PTM 32 nm CMOS technology suggest that the proposed approach attains tremendous improvements in delay, power and area metrics with a trivial degradation in the output quality. The proposed APA shows 3.4 × improvement in performance and 41.5% reduction in area over the best known accuracy configurable adder. Even with 100% accuracy, a 32-bit APA improves delay, power and area by 33.92%, 23.04% and 17.44%, respectively, over Ripple Carry Adder (RCA). We also demonstrate an APA embedded error-resilient JPEG encoder architecture in order to inspect the efficacy of the proposed approach in real-time Digital Signal Processing (DSP) applications.

[1]  Sandeep K. Gupta,et al.  A Re-design Technique for Datapath Modules in Error Tolerant Applications , 2008, 2008 17th Asian Test Symposium.

[2]  John Lach,et al.  Exploring the fidelity-efficiency design space using imprecise arithmetic , 2011, 16th Asia and South Pacific Design Automation Conference (ASP-DAC 2011).

[3]  George A. Constantinides,et al.  Overclocking datapath for latency-error tradeoff , 2013, 2013 IEEE International Symposium on Circuits and Systems (ISCAS2013).

[4]  Chi-Ying Tsui,et al.  Low-power VLSI design for motion estimation using adaptive pixel truncation , 2000, IEEE Trans. Circuits Syst. Video Technol..

[5]  Andrew B. Kahng,et al.  Accuracy-configurable adder for approximate arithmetic designs , 2012, DAC Design Automation Conference 2012.

[6]  Rakesh Kumar,et al.  On reconfiguration-oriented approximate adder design and its application , 2013, 2013 IEEE/ACM International Conference on Computer-Aided Design (ICCAD).

[7]  H. T. Kung,et al.  A Regular Layout for Parallel Adders , 1982, IEEE Transactions on Computers.

[8]  Zhi-Hui Kong,et al.  Design of Low-Power High-Speed Truncation-Error-Tolerant Adder and Its Application in Digital Signal Processing , 2010, IEEE Transactions on Very Large Scale Integration (VLSI) Systems.

[9]  Shih-Lien Lu.  Speeding Up Processing with Approximation Circuits , 2004, Computer.

[10]  Fabrizio Lombardi,et al.  New Metrics for the Reliability of Approximate and Probabilistic Adders , 2013, IEEE Transactions on Computers.

[11]  Ilia Polian,et al.  Adaptive voltage over-scaling for resilient applications , 2011, 2011 Design, Automation & Test in Europe.

[12]  Wojciech Maly,et al.  Cost of Silicon Viewed from VLSI Design Perspective , 1994, 31st Design Automation Conference.

[13]  Harold S. Stone,et al.  A Parallel Algorithm for the Efficient Solution of a General Class of Recurrence Equations , 1973, IEEE Transactions on Computers.

[14]  Gang Wang,et al.  Enhanced low-power high-speed adder for error-tolerant application , 2009, 2010 International SoC Design Conference.

[15]  Paolo Ienne,et al.  Variable Latency Speculative Addition: A New Paradigm for Arithmetic Circuit Design , 2008, 2008 Design, Automation and Test in Europe.

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