An Efficient Majority-Based Compressor For Approximate Computing In The Nano Era

Abstract

Approximate computing is an effective paradigm for energy-efficient hardware design in nanoscale. In this study, an efficient 4:2 compressor for approximate computing in the nano era is proposed. The proposed design includes only two majority gates instead of AND-OR and XOR logics, which leads to circuit efficiency and lower energy consumption. Moreover, the majority operator is the natural logic primitive for several beyond-CMOS technologies such as quantum-dot cellular automata (QCA). The proposed approach is designed using FinFET as a current industrial technology and QCA as a promising emerging nanodevice. FinFETs show lower short channel effects and provide excellent electrostatic characteristics than bulk CMOS for sub-32 nm technologies. Furthermore, QCA will provide extremely high-density and energy-efficient digital circuits for the future VLSI design. In order to evaluate the performance of the proposed approach and make comparisons with the previous designs, extensive simulations are performed using HSPICE, QCADesigner and QCAPro tools. In addition, the proposed compressor is utilized efficiently in image processing applications and the critical metrics in measuring the quality of images are evaluated using MATLAB. The results indicate significant improvements in terms of different performance and accuracy metrics in comparison with the most efficient designs previously presented in the literature.

Publication Date

3-1-2018

Publication Title

Microsystem Technologies

Volume

24

Issue

3

Number of Pages

1589-1601

Document Type

Article

Personal Identifier

scopus

DOI Link

https://doi.org/10.1007/s00542-017-3587-2

Socpus ID

85031780942 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/85031780942

This document is currently not available here.

Share

COinS