A 288μW Programmable Deep-Learning Processor with 270KB On-Chip Weight Storage Using Non-Uniform Memory Hierarchy for Mobile Intelligence

📋 论文概要

本文提出了一款功耗仅288μW的可编程深度学习处理器，集成了270KB片上权重存储，采用非均匀存储层次结构降低能耗，适用于移动物联网边缘智能。

💡 主要创新点

• 采用非均匀存储层次结构（non-uniform memory hierarchy），有效降低深度学习推理中的片上存储访问功耗。
• 实现极低功耗（288μW）的可编程深度学习加速，适合电池供电的物联网设备。

🏷 关键词

📄 原文摘要

Qing Dong1, Yen-Po Chen1, Laura Fick1, Xun Sun1, Ron Dreslinski1, Trevor Mudge1, Hun Seok Kim1, David Blaauw1, Dennis Sylvester1 University of Michigan, Ann Arbor, MI CubeWorks, Ann Arbor, MI 1 2 Deep learning has proven to be a powerful tool for a wide range of applications, such as speech recognition and object detection, among others. Recently there has been increased interest in deep learning for mobile IoT [1] to enable intelligence at the edge and shield the cloud from a deluge of data by only forwarding meaningful events. This hierarchical intelligence thereby enhances radio bandwidth and power efficiency by trading-off computation and communication at edge devices. Since many mobile applications are “always-on” (e.g., voice commands), low power is a critical design constraint. However, prior works have focused on high performance reconfigurable processors [2-3] optimized for large-scale deep neural networks (DNNs) that consume >50mW.

👥 作者与机构

Suyoung Bang1, Jingcheng Wang1, Ziyun Li1, Cao Gao1, Yejoong Kim1,2,