Reconfigurable Five-In-One\nCarbon Nanotube Optoelectronic\nTransistor for Intelligent Computing and Communication

Advances in artificial general intelligence (AGI) necessitate\nthe\nintegration of diverse functionalities to address complex tasks. Carbon\nnanotubes (CNTs), with their unique physical properties, have broad\napplications in emerging research fields, providing a foundation for\nnext-generation devices that could overcome the limits of Moore’s\nLaw. This work demonstrates a novel intelligent device that integrates\nfive functionssensors, memory, neuromorphic computing, logic,\nand communicationusing CNT field-effect transistors (CNFETs)\ncompatible with CMOS processes. Through passivation and annealing\ntechniques, we have significantly enhanced the optoelectronic performance\nof CNFETs, leading to the development of multifunctional optoelectronic\nsynaptic transistors. These optimized CNFETs enable dual-mode weight-tunable\nsynaptic functions, including long-term plasticity and multilevel\nstorage. Additionally, a CNT-based neural network has achieved high\nrecognition accuracy on the MNIST data set, showcasing the potential\nof in-memory computing. This research also innovates by integrating\nlogic functions with optoelectronic communication capabilities, paving\nthe way for next-generation intelligent computing and communication\nintegrated systems.