← 返回 JSSC 论文列表JSSC 2025第10期Digital Circuits
An Analog Neuromorphic On-Chip Training System With IGZO TFT-Based 6T1C Synaptic
提出基于IGZO TFT的6T1C突触结构模拟神经形态芯片训练系统，实现高效多比特状态存储。
367 distinct states, R2=0.99, ENOB=8.95, 97.1% MNIST accuracy
IGZO TFT6T1C突触模拟神经形态芯片上训练MNIST
▸采用IGZO TFT突触单元实现低漏电流多比特状态存储
▸6T1C结构实现高线性度367个状态
▸首次实现模拟芯片上训练，MNIST准确率97.1%
Abstract
This article proposes an analog synapse-based neu-
romorphic on-chip training system that uses emerging indium
gallium zinc oxide (IGZO) thin film transistor (TFT) synapse cells
to store multi-bit states for deep neural networks (DNNs). IGZO
TFT demonstrates extremely low leakage currents, preserving the
charge stored in capacitors during prolonged training periods.
The 6 transistor 1 capacitor (6T1C) structure, characterized
by its symmetrical design and current sources configuration,
achieves