Publication detail

Evolutionary Algorithms in Approximate Computing: A Survey

SEKANINA, L.

Original Title

Evolutionary Algorithms in Approximate Computing: A Survey

Type

journal article in Scopus

Language

English

Original Abstract

In recent years, many design automation methods have been developed to routinely create approximate implementations of circuits and programs that show excellent trade-offs between the quality of output and required resources. This paper deals with evolutionary approximation as one of the popular approximation methods. The paper provides the first survey of evolutionary algorithm (EA)-based approaches applied in the context of approximate computing. The survey reveals that EAs are primarily applied as multi-objective optimizers. We propose to divide these approaches into two main classes: (i) parameter optimization in which the EA optimizes a vector of system parameters, and (ii) synthesis and optimization in which EA is responsible for determining the architecture and parameters of the resulting system. The evolutionary approximation has been applied at all levels of design abstraction and in many different applications. The neural architecture search enabling the automated hardware-aware design of approximate deep neural networks was identified as a newly emerging topic in this area.

Keywords

evolutionary algorithm, approximate computing, digital circuit, neural network, optimization

Authors

SEKANINA, L.

Released

23. 8. 2021

ISBN

1872-0234

Periodical

Journal of Integrated Circuits and Systems

Year of study

16

Number

2

State

Federative Republic of Brazil

Pages from

1

Pages to

12

Pages count

12

URL

https://jics.org.br/ojs/index.php/JICS/article/view/499

BibTex

@article{BUT175795,
  author="Lukáš {Sekanina}",
  title="Evolutionary Algorithms in Approximate Computing: A Survey",
  journal="Journal of Integrated Circuits and Systems",
  year="2021",
  volume="16",
  number="2",
  pages="1--12",
  doi="10.29292/jics.v16i2.499",
  issn="1872-0234",
  url="https://jics.org.br/ojs/index.php/JICS/article/view/499"
}