Publication Details

Improving the Accuracy and Hardware Efficiency of Neural Networks Using Approximate Multipliers

ANSARI, M.; MRÁZEK, V.; COCKBURN, B.; SEKANINA, L.; VAŠÍČEK, Z.; HAN, J. Improving the Accuracy and Hardware Efficiency of Neural Networks Using Approximate Multipliers. IEEE Trans. on VLSI Systems., 2020, vol. 28, no. 2, p. 317-328. ISSN: 1063-8210.
Czech title
Vylepšení přesnosti a obvodové realizace neuronových sítí pomocí aproximativních násobiček
Type
journal article
Language
English
Authors
ANSARI, M.
Mrázek Vojtěch, Ing., Ph.D. (DCSY)
COCKBURN, B.
Sekanina Lukáš, prof. Ing., Ph.D. (DCSY)
Vašíček Zdeněk, doc. Ing., Ph.D. (DCSY)
HAN, J.
Keywords

approximate multipliers, Cartesian genetic programming, convolutional neural network, multi-layer perceptron, neural networks

Abstract

Improving the accuracy of a neural network (NN) usually requires using larger hardware that consumes more energy. However, the error tolerance of NNs and their applications allow approximate computing techniques to be applied to reduce implementation costs. Given that multiplication is the most resource-intensive and power-hungry operation in NNs, more economical approximate multipliers (AMs) can significantly reduce hardware costs. In this article, we show that using AMs can also improve the NN accuracy by introducing noise. We consider two categories of AMs: 1) deliberately designed and 2) Cartesian genetic programing (CGP)-based AMs. The exact multipliers in two representative NNs, a multilayer perceptron (MLP) and a convolutional NN (CNN), are replaced with approximate designs to evaluate their effect on the classification accuracy of the Mixed National Institute of Standards and Technology (MNIST) and Street View House Numbers (SVHN) data sets, respectively. Interestingly, up to 0.63% improvement in the classification accuracy is achieved with reductions of 71.45% and 61.55% in the energy consumption and area, respectively. Finally, the features in an AM are identified that tend to make one design outperform others with respect to NN accuracy. Those features are then used to train a predictor that indicates how well an AM is likely to work in an NN.

Published
2020
Pages
317–328
Journal
IEEE Trans. on VLSI Systems., vol. 28, no. 2, ISSN 1063-8210
DOI
10.1109/TVLSI.2019.2940943
UT WoS
000510674300002
EID Scopus
2-s2.0-85078705685
BibTeX 
@article{BUT161464,
  author="ANSARI, M. and MRÁZEK, V. and COCKBURN, B. and SEKANINA, L. and VAŠÍČEK, Z. and HAN, J.",
  title="Improving the Accuracy and Hardware Efficiency of Neural Networks Using Approximate Multipliers",
  journal="IEEE Trans. on VLSI Systems.",
  year="2020",
  volume="28",
  number="2",
  pages="317--328",
  doi="10.1109/TVLSI.2019.2940943",
  issn="1063-8210",
  url="https://www.fit.vut.cz/research/publication/12066/"
}
Files
Back to top