E-NPU: A 34~126nJ/Class Event-Driven Adaptive Neural SoC with Signal-Dynamics-Aware Feature Clustering and Multi-Model In-Memory Inference/Training for Personalized Medical Wearables
Fengshi Tian, Jinbo Chen, Kunming Shao, Zilu Liu, Jiakun Zheng, Hui Wu, Chaoming Fang, Xiaomeng Wang, Ziyang Shen, Pingcheng Dong, Yuan Yao, Xuliang Wang, Jie Yang, Mohamad Sawan, Chi-Ying Tsui, Kwang-Ting Tim Cheng · 2025
Biomedical signal processing systems-on-chip (SoC) have shown significant promise in human-machine interfaces and closed-loop neuromodulation applications [1]–[5]. Future personalized medical wearables are anticipated to provide long-term monitoring and accurate interpretation while maintaining minimal energy consumption [6]–[8]. Lightweight neural networks, such as binary, spiking, and convolutional neural networks (BNN/SNN/CNN), are promising for neural signal analysis, with neural architecture search (NAS) techniques enabling the deployment of task-specific models optimized for accuracy and efficiency [9]–[12]. However, developing an accurate, ultra-energy-efficient medical SoC that seamlessly integrates signal feature extraction and processing with multi-model reconfigurability remains challenging. Furthermore, inter-patient variance adversely affects the accuracy of biosignal classifications in real-world scenarios, thereby limiting the reliability of these devices [13], [14].