Abstract
Approximate/inexact computing has become an attractive approach for designing high performance and low power arithmetic circuits. Floating-point (FLP) arithmetic is required in many applications, such as digital signal processing, image processing and machine learning. Approximate FLP multipliers with variable accuracy are proposed in this paper; the accuracy and the circuit requirements of these designs are analyzed and assessed according to different metrics. It is shown that the proposed approximate FLP multiplier designs further reduce delay, area, power consumption and power-delay product (PDP) while incurring about half of the normalized mean error distance (NMED) compared with the previous designs. The proposed IFLPM24–15 is the most efficient design when considering both PDP and NMED. Case studies with three error-tolerant applications show the validity of the proposed approximate designs.
Similar content being viewed by others
References
Venkataramani, S., Chakradhar, S.T., Roy, K., & Raghunathan, A. (2015). Computing approximately, and efficiently. Design, Automation & Test in Europe Conference & Exhibition (pp. 748–751).
Palem, K., & Ingamneni, A. (2013). Ten years of building brokenchips: The physics and engineering of inexact computing. ACM Transactions on Embedded Computing Systems, 12(5), 1–23.
Liu, W., Chen, L., Wang, C., O'Neill, M., & Lombardi, F. (2016). Design and analysis of inexact floating-point adders. IEEE Transactions on Computers, 65(1), 308–314.
Liang, J., Han, J., & Lombardi, F. (2013). New metrics for the reliability of approximate and probabilistic adders. IEEE Transactions on Computers, 62(9), 1760–1771.
Liu, W., Qian, L., Wang, C., Jiang, H., Han, J., & Lombardi, F. (2017). Design of Approximate Radix-4 booth multipliers for error-tolerant computing. IEEE Transactions on Computers, 66(8), 1435–1441.
Liu, C., Han, J., & Lombardi, F. (2014). A low-power, high-performance approximate multiplier with configurable partial error recovery. Design, Automation and Test in Europe Conference and Exhibition (DATE) (pp. 14).
Momeni, A., Han, J., Montuschi, P., & Lombardi, F. (2015). Design and analysis of approximate compressor for multiplication. IEEE Transactions on Computers, 64(4), 984–994.
Bhardwaj, K., Mane, P.S., & Henkel, J. (2014). Power-and area-efficient approximate Wallace tree multiplier for error-resilient systems. 15th International Symposium on Quality Electronic Design (ISQED) (pp. 263269).
Kulkarn, P., Gupta, P., & Ercegovac, M. (2011). Trading accuracy for power with an Underdesigned multiplier architecture. 24th International Conference on VLSI Design (VLSI Design), (pp. 346351).
Sullivan, M.B., & Swartzlander, E.E. Jr. (2012). Truncated error correction for flexible approximate multiplication. Forty-Sixth Asilomar Conference on Signals, Systems and Computers, (pp. 355359).
Park, H., & Swartzlander, Jr. E.E. (2006). Truncated multiplication with symmetric correction. Fortieth Asilomar Conference on Signals, Systems and Computers, (pp. 931934).
Swartzlander, Jr. E.E. (1999). Truncated multiplication with approximate rounding. Thirty-Third Asilomar Conference on Signals, Systems and Computers, (pp. 14801483).
King, E.J., & Swartzlander, Jr. E.E. (1997). Data-dependent truncation scheme for parallel multipliers. 31st Asilomar Conference on Signals, Systems and Computers, (pp. 11781182).
Schulte M.J., & Swartzlander, Jr. E.E. (1993). Truncated multiplication with correction constant. IEEE Workshop on VLSI Signal Processin, (pp. 388-396).
Tong, J. Y., Nagle, D., & Rutenbar, R. (2000). Reducing power by optimizing the necessary precision/range of floating-point arithmetic. IEEE Transactions on Very Large Scale Integrated System, 8(3), 273–286.
Fang, F., Chen, T., & Rutenbar, R. (2002). Floating-point bit-width optimization for low-power signal processing applications. Proc. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), (pp.32083211).
Eilert, J., Ehliar, A., & Liu, D. (2004). Using low precision floating point numbers to reduce memory cost for MP3 decoding. Proc. 6th IEEE Workshop on Multimedia Signal Processing, (pp.119122).
Gupta, A., Mandavalli, S., Mooney, V., Ling, K., Basu, A., Johan, H., & Tandianus, B. (2011). Low power probabilistic floating point multiplier design. 2011 I.E. Computor Society Annual Symposium on VLSI, (pp. 182187).
Zhang, H., Putic, M., & Lach, J. (2014). Low power GPGPU computation with imprecise hardware. 51st ACM/EDAC/IEEE Design Automation Conference (DAC), (pp.16).
Zhang, H., Zhang, W., & Lach, J. (2014). A low-power accuracy-configurable floating PointMultiplier. 32nd IEEE International Conference on Computer Design (ICCD), (pp. 4854).
Yin, P., Wang, C., Liu, W., & Lombardi, F. (2016). Design and performance evaluation of approximate floating-point multipliers. 2016 I.E. Computer Society Annual Symposium on VLSI, (pp. 296301).
IEEE Computer Society (2008) IEEE standard for floating-point arithmetic. IEEE Standard 754-2008.
Yeh, W., & Jen, C. (2000). High-speed booth encoded parallel multiplier design. IEEE Transactions on Computers, 49(7), 692–701.
Industrial Light & Magic, http://www.openexr.com. 2014.
Mantiuk, R., Kim, K., Rempel, A., & Heidrich, W. (2011). HDR-VDP-2: A calibrated visual metric for visibility and quality predictions in all luminance conditions. ACM Transactions on Graphics, 30(4), 76–79.
Kung, J., Kim, D., & Mukhopadhyay, S. (2015). A power-aware digital feedforward neural network platform with backpropagation driven approximate synapses. 2015 IEEE/ACM International Symposium on Low Power Electronics and Design (ISLPED), (pp. 8590).
Du, Z., Lingamneni, A., Chen, Y., Palem, K., Temam, O., & Wu, C. (2015). Leveraging the error resilience of neural Networksfor designing highly energy efficient accelerators. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 34(8), 1223–1235.
Asuncion, A., & Newman, D. UCI machine learning repository. http://archive.ics.uci.edu/ml/index.php. 2007.
Acknowledgements
This work is supported by grants from National Natural Science Foundation of China (No. 61401197), Natural Science Foundation of Jiangsu Province (BK20151477) and Fundamental Research Funds for the Central Universities China (NS2017024).
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
ESM 1
(PDF 712 kb)
Rights and permissions
About this article
Cite this article
Yin, P., Wang, C., Liu, W. et al. Designs of Approximate Floating-Point Multipliers with Variable Accuracy for Error-Tolerant Applications. J Sign Process Syst 90, 641–654 (2018). https://doi.org/10.1007/s11265-017-1280-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11265-017-1280-4