image: Figure | Analogue optoelectronic memristor platform for in-sensor computing. a, Schematic illustration and optical photograph of the crossbar array incorporating optoelectronic memristors. b, Conceptual diagram of the analogue architecture for under-display fingerprint recognition. c, Broadband photoresponsivity measured under uniform light intensity. d, Linear current–voltage (I–V) characteristics at relatively low driving voltages enable direct use of analogue voltage inputs, while the nonlinear behaviour at higher voltages facilitates effective reservoir module implementation.
Credit: Mingdong Yi et al.
With the growing demand for edge computing across mobile devices and security systems, conventional electronics are increasingly hindered by high power consumption and complex integration requirements. In response to these challenges, researchers from Nanjing University of Posts and Telecommunications, Nanjing Tech University, and The Hong Kong University of Science and Technology have developed a novel solution based on ultra-low voltage optoelectronic polymer memristors.
With the growing demand for edge computing across mobile devices and security systems, conventional electronics are increasingly hindered by high power consumption and complex integration requirements. In response to these challenges, researchers from Nanjing University of Posts and Telecommunications, Nanjing Tech University, and The Hong Kong University of Science and Technology have developed a novel solution based on ultra-low voltage optoelectronic polymer memristors.
Utilising these devices, the researchers constructed a reservoir computing framework which achieved a fingerprint recognition accuracy of 97.15%. The memristors are fabricated via scalable solution-based techniques on flexible substrates, offering both low-cost production and compatibility with large-area flexible electronics, highlighting their potential for widespread adoption in future portable smart systems.
Journal
Light Science & Applications
Article Title
Optoelectronic polymer memristors with dynamic control for power-efficient in-sensor edge computing