Details

Autor(en) / Beteiligte

• Yang, Shu‐Ting
• Li, Xing‐Yu
• Yu, Tong‐Liang
• Wang, Jie
• Fang, Hong
• Nie, Fang
• He, Bin
• Zhao, Le
• Lü, Wei‐Ming
• Yan, Shi‐Shen
• Nogaret, Alain
• Liu, Gang
• Zheng, Li‐Mei

Titel

High‐Performance Neuromorphic Computing Based on Ferroelectric Synapses with Excellent Conductance Linearity and Symmetry

Ist Teil von

• Advanced functional materials, 2022-08, Vol.32 (35), p.n/a

Ort / Verlag

Hoboken: Wiley Subscription Services, Inc

Erscheinungsjahr

2022

Quelle

Wiley Online Library - AutoHoldings Journals

Beschreibungen/Notizen

• Artificial synapses can boost neuromorphic computing to overcome the inherent limitations of von Neumann architecture. As a promising memristor candidate, ferroelectric tunnel junctions (FTJ) enable the authors to successfully emulate spike‐timing‐dependent synapses. However, the nonlinear and asymmetric synaptic weight update under repeated presynaptic stimulation hampers neuromorphic computing by favoring the runaway of synaptic weights during learning. Here, the authors demonstrate an FTJ whose conductivity varies linearly and symmetrically by judiciously combining ferroelectric domain switching and oxygen vacancy migration. The artificial neural network based on this FTJ‐synapse achieves classification accuracy of 96.7% during supervised learning, which is the closest to the maximum theoretical value of 98% achieved to date. This artificial synapse also demonstrates stable unsupervised learning in a noisy environment for its well‐balanced spike‐timing‐dependent plasticity response. The novel concept of controlling ionic migration in ferroelectric materials paves the way toward highly reliable and reproducible supervised and unsupervised learning strategies. By combining ferroelectric domain switching and oxygen vacancy migration, a ferroelectric tunnel junction artificial synapse with intrinsic nonlinearity as low as 0.13–0.17 and symmetric weight updating is developed, which greatly improved the classification accuracy of neural network hardware in supervised learning to 96.7% and enhanced robustness to noise during unsupervised learning.