Conference paper

MRÁZEK Vojtěch, HANIF Muhammad A., VAŠÍČEK Zdeněk, SEKANINA Lukáš and SHAFIQUE Muhammad. autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components. In: The 56th Annual Design Automation Conference 2019 (DAC '19). Las Vegas: Association for Computing Machinery, 2019, pp. 1-6. ISBN 978-1-4503-6725-7. Available from: https://arxiv.org/abs/1902.10807
Publication language:english
Original title:autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components
Title (cs):autoAx: Automatická metodologie pro vytváření obvodů s využitím knihoven aproximačních komponent
Pages:1-6
Proceedings:The 56th Annual Design Automation Conference 2019 (DAC '19)
Conference:Design Automation Conference
Place:Las Vegas, US
Year:2019
URL:https://arxiv.org/abs/1902.10807
ISBN:978-1-4503-6725-7
DOI:10.1145/3316781.3317781
Publisher:Association for Computing Machinery
Keywords
approximate computing, design space exploration, approximate components, machine learning
Annotation
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.
BibTeX:
@INPROCEEDINGS{
   author = {Vojt{\v{e}}ch Mr{\'{a}}zek and A. Muhammad Hanif
	and Zden{\v{e}}k Va{\v{s}}{\'{i}}{\v{c}}ek and
	Luk{\'{a}}{\v{s}} Sekanina and Muhammad Shafique},
   title = {autoAx: An Automatic Design Space Exploration and
	Circuit Building Methodology utilizing Libraries
	of Approximate Components},
   pages = {1--6},
   booktitle = {The 56th Annual Design Automation Conference 2019 (DAC '19)},
   year = 2019,
   location = {Las Vegas, US},
   publisher = {Association for Computing Machinery},
   ISBN = {978-1-4503-6725-7},
   doi = {10.1145/3316781.3317781},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=11862}
}

Your IPv4 address: 18.232.124.77
Switch to https