Design of Power-efficient Approximate Multipliers for Approximate Artificial Neural Networks
Created by W.Langdon from
gp-bibliography.bib Revision:1.8010
- @InProceedings{Mrazek:2016:ICCAD,
-
author = "Vojtech Mrazek and Syed Shakib Sarwar and
Lukas Sekanina and Zdenek Vasicek and Kaushik Roy",
-
title = "Design of Power-efficient Approximate Multipliers for
Approximate Artificial Neural Networks",
-
booktitle = "Proceedings of the 35th International Conference on
Computer-Aided Design",
-
year = "2016",
-
pages = "81:1--81:7",
-
articleno = "81",
-
address = "Austin, Texas, USA",
-
month = nov # " 7-10",
-
publisher = "ACM",
-
keywords = "genetic algorithms, genetic programming, cartesian
genetic programming, Approximate computing, Neural
networks, Logic synthesis, Low power",
-
acmid = "2967021",
-
isbn13 = "978-1-4503-4466-1",
-
URL = "http://www.fit.vutbr.cz/~sekanina/pubs.php.en?id=11142",
-
URL = "http://doi.acm.org/10.1145/2966986.2967021",
-
DOI = "doi:10.1145/2966986.2967021",
-
size = "7 pages",
-
abstract = "Artificial neural networks (NN) have shown a
significant promise in difficult tasks like image
classification or speech recognition. Even
well-optimized hardware implementations of digital NNs
show significant power consumption. It is mainly due to
non-uniform pipeline structures and inherent redundancy
of numerous arithmetic operations that have to be
performed to produce each single output vector. This
paper provides a methodology for the design of
well-optimized power-efficient NNs with a uniform
structure suitable for hardware implementation. An
error resilience analysis was performed in order to
determine key constraints for the design of approximate
multipliers that are employed in the resulting
structure of NN. By means of a search based
approximation method, approximate multipliers showing
desired tradeoffs between the accuracy and
implementation cost were created. Resulting approximate
NNs, containing the approximate multipliers, were
evaluated using standard benchmarks (MNIST dataset) and
a real-world classification problem of Street-View
House Numbers. Significant improvement in power
efficiency was obtained in both cases with respect to
regular NNs. In some cases, 91percent power reduction
of multiplication led to classification accuracy
degradation of less than 2.80percent. Moreover, the
paper showed the capability of the back propagation
learning algorithm to adapt with NNs containing the
approximate multipliers.",
- }
Genetic Programming entries for
Vojtech Mrazek
Syed Shakib Sarwar
Lukas Sekanina
Zdenek Vasicek
Kaushik Roy
Citations