A 6K-MAC Feature-Map-Sparsity-Aware Neural Processing Unit in 5nm Flagship Mobile SoC

📋 论文概要

该论文提出了一款6K-MAC的特征图稀疏性感知神经处理单元（NPU），集成于5nm旗舰移动SoC中，旨在解决移动设备上实时机器学习应用的计算资源、功耗和内存带宽限制问题。通过利用特征图稀疏性，实现了高效的计算加速。

💡 主要创新点

• 提出特征图稀疏性感知架构，利用激活值稀疏性跳过无效计算，降低功耗和延迟。
• 实现6K MAC（乘累加）单元规模，在5nm工艺下平衡性能与面积效率。
• 集成于旗舰移动SoC，支持实时AI应用（如AI相机），兼顾响应速度、隐私保护和低功耗。

🏷 关键词

📄 原文摘要

Hanwoong Jung2, Seungwon Lee2, Suknam Kwon1, Kyungah Jeong1, Joon-Ho Song2, SukHwan Lim1, Inyup Kang1 Samsung Electronics, Hwaseong, Korea Samsung Advanced Institute of Technology, Suwon, Korea 1 2 On-device machine learning is critical for mobile products as it enables real-time applications (e.g. AI-powered camera applications), which need to be responsive, always available (i.e. do not require network connectivity) and privacy preserving. The platforms used in such situations have limited computing resources, power, and memory bandwidth. Enabling such on-device machine learning has triggered wide development of efficient neural-network accelerators that promise high energy and area efficiency compared to general-purpose processors, such as CPUs. The need to support a comprehensive range of neural networks has been important as well because the field of deep learning is evolving rapidly as depicted in Fig. 9.5.1. Recent work on neuralnetwork accelerators has focused on improving ener

👥 作者与机构

Jun-Seok Park1, Jun-Woo Jang2, Heonsoo Lee1, Dongwoo Lee1, Sehwan Lee2,