We have developed synthetic domain computing and neural network based on our nonlinear phononic devices at GHz frequencies. Our device-aware neural network can perform a four-class classification task with an accuracy of 98.2%. Our devices feature a wide operation temperature range up to 192 °C. This work is published in Nature Electronics.
Congratulations to Dr. Jun Ji and co-authors!
Read the paper: https://doi.org/10.1038/s41928-025-01436-9
Full text link: https://rdcu.be/eDiGF