Liu, Xin-YouXin-YouLiuShih, Chi-ShengChi-ShengShihLiu, Tsung-TeTsung-TeLiuTsung, Pei-KueiPei-KueiTsungChen, Chih-WeiChih-WeiChenTeng, Chieh-FangChieh-FangTeng2026-02-232026-02-232025-09-28[9798400719912]https://www.scopus.com/record/display.uri?eid=2-s2.0-105025359449&origin=resultslisthttps://scholars.lib.ntu.edu.tw/handle/123456789/735952Neural networks have demonstrated superior performance over rule-based and model-based approaches in processing noisy sensing data. However, their substantial computational and energy demands hinder deployment in battery-powered embedded systems. Computing-in-Memory (CIM) devices offer a promising alternative by significantly reducing energy consumption. Prior work [2] achieves this by leveraging a non-von Neumann architecture, which minimizes data movement between memory and compute units, thereby mitigating the memory wall bottleneck. Despite these advantages, analog CIM (ACIM) systems face several key challenges, including analog noise, limited numerical precision, and increased hardware complexity. While cloud-based neural networks are still dominant, emerging applications increasingly demand real-time, privacy-preserving inference on-device. For instance, facial authentication requires local execution on edge devices to ensure low-latency responsiveness and to protect user privacy. CIM architectures are particularly well-suited to these scenarios due to their tightly integrated memory-compute structure, offering low-latency and energy-efficient inference capabilities.trueconference paper10.1145/3742872.37570742-s2.0-105025359449