Publication detail

Evolutionary Design of Approximate Multipliers Under Different Error Metrics

VAŠÍČEK, Z. SEKANINA, L.

Original Title

Evolutionary Design of Approximate Multipliers Under Different Error Metrics

Type

conference paper

Language

English

Original Abstract

Approximate circuits are digital circuits which are intentionally designed in such a way that the specification is not met in terms of functionality in order to obtain some improvements in power consumption, performance or area, in comparison with fully functional circuits. In this paper, we propose to design approximate circuits using evolutionary design techniques. In particular, different error metrics are utilized to assess the circuit functionality. The proposed method begins with a fully functional circuit which is then intentionally degraded by Cartesian genetic programming (CGP) to obtain a circuit with a predefined error. In the second phase, CGP is used to minimize the number of gates or another error criterion. The effect of various error metrics on the search performance, area and power consumption is evaluated in the task of multiplier design.

Keywords

approximate circuit, multiplier, evolutionary design

Authors

VAŠÍČEK, Z.; SEKANINA, L.

RIV year

2014

Released

16. 5. 2014

Publisher

IEEE Computer Society

Location

Warsaw

ISBN

978-1-4799-4558-0

Book

17th IEEE Symposium on Design and Diagnostics of Electronic Circuits and Systems

Pages from

135

Pages to

140

Pages count

6

URL

BibTex

@inproceedings{BUT111522,
  author="Zdeněk {Vašíček} and Lukáš {Sekanina}",
  title="Evolutionary Design of Approximate Multipliers Under Different Error Metrics",
  booktitle="17th IEEE Symposium on Design and Diagnostics of Electronic Circuits and Systems",
  year="2014",
  pages="135--140",
  publisher="IEEE Computer Society",
  address="Warsaw",
  doi="10.1109/DDECS.2014.6868777",
  isbn="978-1-4799-4558-0",
  url="https://www.fit.vut.cz/research/publication/10513/"
}