A 22nm 0.26nW/Synapse Spike-Driven Spiking Neural Network Processing Unit Using Time-Step-First Dataflow and Sparsity-Adaptive In-Memory Computing

📋 论文概要

本文提出了一种基于22nm工艺的脉冲神经网络处理单元，采用时间步优先数据流和稀疏自适应技术，实现了每突触0.26nW的超低功耗。该设计针对动态视觉传感器等边缘AI应用，在保持高精度的同时大幅降低能耗。

💡 主要创新点

• 提出时间步优先数据流架构，优化了SNN中时间维度的计算顺序，减少冗余操作。
• 引入稀疏自适应机制，动态利用脉冲稀疏性，进一步降低功耗和计算量。

🏷 关键词

📄 原文摘要

Science and Technology University, Beijing, China *Equally Credited Authors (ECA) 1 2 Recently, brain-inspired spiking neural networks (SNNs) have demonstrated tremendous improvement in energy efficiency (EE) and low power by exploiting highly sparse spikes and event-driven design [1-2]. (top of Fig. 30.2.1) With the same spike-based information carrier, the combination of an SNN and dynamic vision sensor (DVS) [3] offers a promising solution for edge AI applications. Typically, the greater the number of synapses in an SNN can improve inference accuracy, but induce more significant power consumption and memory size. Therefore, deploying more synapses in an SNN chip while maintaining low power, memory size and high EE requires co-design of algorithm,

👥 作者与机构

Ying Liu*1, Yufei Ma*1, Ninghui Shang1, Tianhao Zhao2, Peiyu Chen1, Meng Wu1,

Jiayoon Ru1, Tianyu Jia1, Le Ye1, Zhixuan Wang3, Ru Huang1 Peking University, Beijing, China Nano Core Chip Electronic Technology, Hangzhou, China 3 Beijing Information