TY - GEN
T1 - Variation-aware binarized memristive networks
AU - Lammie, Corey
AU - Krestinskaya, Olga
AU - James, Alex
AU - Azghadi, Mostafa Rahimi
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/11
Y1 - 2019/11
N2 - The quantization of weights to binary states in Deep Neural Networks (DNNs) can replace resource-hungry multiply accumulate operations with simple accumulations. Such Binarized Neural Networks (BNNs) exhibit greatly reduced resource and power requirements. In addition, memristors have been shown as promising synaptic weight elements in DNNs. In this paper, we propose and simulate novel Binarized Memristive Convolutional Neural Network (BMCNN) architectures employing hybrid weight and parameter representations. We train the proposed architectures offline and then map the trained parameters to our binarized memristive devices for inference. To take into account the variations in memristive devices, and to study their effect on the performance, we introduce variations in RON and ROFF. Moreover, we introduce means to mitigate the adverse effect of memristive variations in our proposed networks. Finally, we benchmark our BMCNNs and variation-aware BMCNNs using the MNIST dataset.
AB - The quantization of weights to binary states in Deep Neural Networks (DNNs) can replace resource-hungry multiply accumulate operations with simple accumulations. Such Binarized Neural Networks (BNNs) exhibit greatly reduced resource and power requirements. In addition, memristors have been shown as promising synaptic weight elements in DNNs. In this paper, we propose and simulate novel Binarized Memristive Convolutional Neural Network (BMCNN) architectures employing hybrid weight and parameter representations. We train the proposed architectures offline and then map the trained parameters to our binarized memristive devices for inference. To take into account the variations in memristive devices, and to study their effect on the performance, we introduce variations in RON and ROFF. Moreover, we introduce means to mitigate the adverse effect of memristive variations in our proposed networks. Finally, we benchmark our BMCNNs and variation-aware BMCNNs using the MNIST dataset.
UR - http://www.scopus.com/inward/record.url?scp=85079194825&partnerID=8YFLogxK
U2 - 10.1109/ICECS46596.2019.8964998
DO - 10.1109/ICECS46596.2019.8964998
M3 - Conference contribution
AN - SCOPUS:85079194825
T3 - 2019 26th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2019
SP - 490
EP - 493
BT - 2019 26th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2019
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 26th IEEE International Conference on Electronics, Circuits and Systems, ICECS 2019
Y2 - 27 November 2019 through 29 November 2019
ER -