Peng Wang, Biyi Jiang, Wuhong Xue, Jingyuan Qu, Wei Zhao, Tianqi Liu, Yan Han, Rongkai Jiao, Ruilong Yang, Feichi Zhou, Xiaohong Xu
Ferroelectric transistors have the potential to meet the computing demands of data-intensive applications such as artificial intelligence. However, the existing research lacks functional reconfigurability and multi-task parallel processing capability. Here, we report a dual-ferroelectric stacked transistor that operates through opto-electric dual-mode induced ferroelectric polarization reversal, which features configurable multi-mode functionality encompassing electronic storage, multi-level optical memory, and visual functional emulation. The dual-ferroelectric architecture significantly enhances non-volatile memory performances, demonstrating an ultra-high current on/off ratio, ultra-low dark current and energy consumption, over 8-bit conductance states, multiple analog states and large dynamic range. Furthermore, the simulated artificial visual system enables multi-task parallel processing of static and motion mixed-color information. This work demonstrates that dual-ferroelectric stacked architecture holds potential in parallel processing of multidimensional information.