Dynamic Ferroelectric Transistor?Based Reservoir Computing for Spatiotemporal Information Processing

Reservoir computing (RC) architecture which mimics the human brain is a fundamentally preferred method to process dynamical systems that evolve with time. However, difficulty in generating rich reservoir states using two-terminal devices remains challenging, hinders its hardware implementation. Herein, 1D array of ferroelectric field-effect transistor (Fe-FET) based on ?-In2Se3 channel, shows volatile memory effect for realizing various RC systems, demonstrated. The fading sufficiently investigated by polarization dynamic model. proposed Fe-FET capable experimentally classifying images MNIST dataset high accuracy 91%. Furthermore, time-series real-life chaotic system, example, Earth's weather, can be accurately forecasted our Ferro-RC Jena climate recorded 1 year period. Remarkable determination coefficient (R 2) 0.9983 and normalized root mean square error (NRMSE) 8.3 × 10?3 are achieved minimized readout network. demonstration integrated computation opens route compact system.