Nanocrystal-embedded-insulator (NEI) ferroelectric field-effect transistor featuring low operating voltages and improved synaptic behavior

Y Peng, W Xiao, G Han, Y Liu, J Wu… - IEEE Electron …, 2019 - ieeexplore.ieee.org
Y Peng, W Xiao, G Han, Y Liu, J Wu, K Wang, Y He, Z Yu, X Wang, N Xu, TJK Liu, Y Hao
IEEE Electron Device Letters, 2019•ieeexplore.ieee.org
A novel nanocrystal-embedded-insulator (NEI) ferroelectric field-effect transistor (FeFET) is
demonstrated to function as a synaptic device for analog neural network (NN) applications.
The NEI layer (down to 3.6 nm in thickness) comprises ferroelectric nanocrystals embedded
in amorphous Al 2 O 3, resulting in reduced operating voltages and depolarization effects as
compared to conventional doped-HfO 2 films. With fixed-amplitude 100 ns potentiation/
depression pulses, an NEI FeFET synapse achieves weight update with small non-linearity …
A novel nanocrystal-embedded-insulator (NEI) ferroelectric field-effect transistor (FeFET) is demonstrated to function as a synaptic device for analog neural network (NN) applications. The NEI layer (down to 3.6 nm in thickness) comprises ferroelectric nanocrystals embedded in amorphous Al 2 O 3 , resulting in reduced operating voltages and depolarization effects as compared to conventional doped-HfO 2 films. With fixed-amplitude 100 ns potentiation/depression pulses, an NEI FeFET synapse achieves weight update with small non-linearity ( , ) and asymmetry factors, advantageous for analog-style NNs with online training. A convolutional NN is designed and emulated for an MNIST dataset, projecting an online training accuracy of 92%.
ieeexplore.ieee.org
展开收起
被引用次数：23 相关文章 所有 2 个版本
以上显示的是最相近的搜索结果。 查看全部搜索结果