← 返回 JSSC 论文列表JSSC 2023第5期Memory40nm
An In-Memory-Computing Charge-Domain Ternary CNN Classifier Xiangxing Y ang Memb
提出一种基于电荷域计算的三元神经网络分类器，实现高效低功耗的MNIST分类。
40nm LP CMOS, 549 FPS, 96 µW, 97.1% MNIST准确率, 0.18 µJ/分类
电荷域计算三元神经网络电容开关低功耗MNIST分类
▸1.5-b分辨率的三元权重和激活
▸基于VCM的电容开关方案实现MAC
▸训练中引入稀疏性降低切换率
Abstract
The article presents a charge-domain comput-
ing ternary neural network (TNN) classifier with a complete
four-layer neural network (NN) on a chip. The proposed ternary
network provides 1.5-b resolution (0/+1/−1) for weights and acti-
vations, leading to 3.9× fewer operations (OPs) per inference than
binary neural network (BNN) for the same Modified National
Institute of Standards and Technology (MNIST) accuracy. The
1.5-b multiply-and-accumulate (MAC) is implemented by V
CM-
based capacitor switch