Quantized Neural Network via Synaptic Segregation Based on Ternary Charge‐Trap Transistors
Quantized Neural Network via Synaptic Segregation Based on Ternary Charge‐Trap Transistors
复制标题
DOI：
10.1002/aelm.202300303
复制
发表时间：
2023-09
影响因子：
6.2
通讯作者：
Y. Baek;B. Bae;Jeongyong Yang;Doeon Lee;H. Lee;Minseong Park;Taegeon Kim;Sihwan Kim;Bo-In Park;Geonwook Yoo;Kyusang Lee
中科院分区：
文献类型：
--
作者：
Y. Baek;B. Bae;Jeongyong Yang;Doeon Lee;H. Lee;Minseong Park;Taegeon Kim;Sihwan Kim;Bo-In Park;Geonwook Yoo;Kyusang Lee
人工神经网络（ANN）广泛用于许多基于人工智能的应用。然而，计算单元和存储之间传输的大量数据限制了人工神经网络在人工智能物联网（AIoT）和功率受限设备应用中的广泛部署。因此，在各种人工神经网络算法中，量化神经网络（QNN）因其需要较少的计算资源和最小的能量消耗而受到相当大的关注。在此，介绍了一种基于氧化物的三元电荷陷阱晶体管（CTT），其提供三个离散状态和非易失性存储器特性，这对于QNN计算是期望的。通过采用一对三进制CTT的差分，证明了QNN的多级量化值的人工突触分离。该方法建立了一个平台，该平台结合了多状态和对内存计算噪声的鲁棒性的优势，以在硬件中实现可靠的QNN性能，从而促进节能AIoT的开发。
Artificial neural networks (ANNs) are widely used in numerous artificial intelligence‐based applications. However, the significant amount of data transferred between computing units and storage has limited the widespread deployment of ANN for the artificial intelligence of things (AIoT) and power‐constrained device applications. Therefore, among various ANN algorithms, quantized neural networks (QNNs) have garnered considerable attention because they require fewer computational resources with minimal energy consumption. Herein, an oxide‐based ternary charge‐trap transistor (CTT) that provides three discrete states and non‐volatile memory characteristics are introduced, which are desirable for QNN computing. By employing a differential pair of ternary CTTs, an artificial synaptic segregation with multilevel quantized values for QNNs is demostrated. The approach establishes a platform that combines the advantages of multiple states and robustness to noise for in‐memory computing to achieve reliable QNN performance in hardware, thereby facilitating the development of energy‐efficient AIoT.