An Accurate, Error-Tolerant, and Energy-Efficient Neural Network Inference Engine Based on SONOS Analog Memory

T. Patrick Xiao,Ben Feinberg,Christopher H. Bennett,Vineet Agrawal,Prashant Saxena,Venkatraman Prabhakar, Krishnaswamy Ramkumar,Harsha Medu,Vijay Raghavan,Ramesh Chettuvetty,Sapan Agarwal,Matthew J. Marinella

IEEE Transactions on Circuits and Systems I: Regular Papers（2022）

引用 10|浏览17

暂无评分

摘要

We demonstrate SONOS (silicon-oxide-nitride-oxide-silicon) analog memory arrays that are optimized for neural network inference. The devices are fabricated in a 40nm process and operated in the subthreshold regime for in-memory matrix multiplication. Subthreshold operation enables low conductances to be implemented with low error, which matches the typical weight distribution of neural networks, w...

查看译文

关键词

SONOS devices,Neural networks,Transistors,Logic gates,Programming,Memristors,Analog memory

AI 理解论文

溯源树

样例

生成溯源树，研究论文发展脉络

Chat Paper

正在生成论文摘要