Publication detail

autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components

MRÁZEK, V. HANIF, M. VAŠÍČEK, Z. SEKANINA, L. SHAFIQUE, M.

Original Title

autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components

Type

conference paper

Language

English

Original Abstract

Approximate computing is an emerging paradigm for developing highly energy-efficient computing systems such as various accelerators. In the literature, many libraries of elementary approximate circuits have already been proposed to simplify the design process of approximate accelerators. Because these libraries contain from tens to thousands of approximate implementations for a single arithmetic operation it is intractable to find an optimal combination of approximate circuits in the library even for an application consisting of a few operations. An open problem is "how to effectively combine circuits from these libraries to construct complex approximate accelerators''. This paper proposes a novel methodology for searching, selecting and combining the most suitable approximate circuits from a set of available libraries to generate an approximate accelerator for a given application. To enable fast design space generation and exploration, the methodology utilizes machine learning techniques to create  computational models estimating the overall quality of processing and hardware cost without performing full synthesis at the accelerator  level. Using the methodology, we construct hundreds of approximate accelerators (for a Sobel edge detector) showing different but relevant tradeoffs between the quality of processing and hardware cost and identify a corresponding Pareto-frontier. Furthermore, when searching for  approximate implementations of a generic Gaussian filter consisting of 17 arithmetic operations, the proposed approach allows us to identify approximately 10^3 highly relevant implementations from 10^23 possible solutions in a few hours, while the exhaustive search would take four months on a high-end processor.

Keywords

approximate computing, design space exploration, approximate components, machine learning

Authors

MRÁZEK, V.; HANIF, M.; VAŠÍČEK, Z.; SEKANINA, L.; SHAFIQUE, M.

Released

20. 6. 2019

Publisher

Association for Computing Machinery

Location

Las Vegas

ISBN

978-1-4503-6725-7

Book

The 56th Annual Design Automation Conference 2019 (DAC '19)

Pages from

1

Pages to

6

Pages count

6

URL

https://arxiv.org/abs/1902.10807

BibTex

@inproceedings{BUT158069,
  author="MRÁZEK, V. and HANIF, M. and VAŠÍČEK, Z. and SEKANINA, L. and SHAFIQUE, M.",
  title="autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components",
  booktitle="The 56th Annual Design Automation Conference 2019 (DAC '19)",
  year="2019",
  pages="1--6",
  publisher="Association for Computing Machinery",
  address="Las Vegas",
  doi="10.1145/3316781.3317781",
  isbn="978-1-4503-6725-7",
  url="https://arxiv.org/abs/1902.10807"
}

Documents

DAC_2019_autoAx.pdf