Energy-Efficient Temperature-Calibration Readout Circuits with Thermal-State Sensible Sampling and Zoom Window Switch Scheme for RRAM-Based Analog Computing-in-Memory

作者：Zhuoya Chen, Zongwei Wang, Haisu Zhang, Linbo Shan, Qishen Wang, Xiyuan Tang, Yunyi Fu, Yimao Cai, Ru Huang · 年份：2025 · DOI：10.1109/edtm61175.2025.11041619 · 被引用次数：2 · 研究领域：Advanced Memory and Neural Computing、Ferroelectric and Negative Capacitance Devices、Semiconductor materials and devices

In this work, we present a temperature calibration readout circuit for RRAM-based analog computing-inmemory (ACIM), which includes i) a thermal-state sensible sampling circuit with a compact in-array reference and resistance compensation column to mitigate the temperatureinduced resistance shift issue, ii) a zoom window switch scheme to improve the energy efficiency. Based on 40nm foundry PDK, the implemented sampling circuit enhances recognition accuracy by 16% at$125^{\circ} \mathrm{C}$. Meanwhile, zoom window switch scheme achieves an 93 % reduction in average capacitor switching energy compared to traditional methods.