Related Experiment Video
Updated: Aug 6, 2026
06:36
Optrode Array for Simultaneous Optogenetic Modulation and Electrical Neural Recording
Published on: September 1, 2022
Reconfigurable ferroelectric transistor array for embodied neuromorphic vision with hardware-native sensing,
Yuan Li1,2, Zhi-Cheng Zhang1, Zhaolong Chen3
1The Key Laboratory of Weak Light Nonlinear Photonics, Ministry of Education, School of Physics and TEDA Institute of Applied Physics, Nankai University, Tianjin 300071, China.
Science Advances
|July 17, 2026
Summary
Researchers developed a reconfigurable ferroelectric transistor (Fe-FET) array for edge artificial intelligence (AI). This novel hardware integrates sensing, computation, and activation, enabling dynamic task allocation for energy-efficient neuromorphic vision systems.
Area of Science:
- Materials Science
- Artificial Intelligence
- Neuromorphic Engineering
Background:
- Edge AI and embodied vision require efficient hardware integrating sensing, computation, and activation.
- Current in-sensor computing (ISC) and in-memory computing (IMC) platforms lack reconfigurability and external activation.
- Fixed functional partitions in existing platforms hinder analog signal paths and system flexibility.
Purpose of the Study:
- To develop a reconfigurable hardware platform for edge AI and embodied vision.
- To overcome limitations of current ISC and IMC platforms by integrating functions within a single device.
- To enable dynamic reallocation of hardware functions for adaptive neuromorphic systems.
Main Methods:
- Fabrication of a reconfigurable ferroelectric transistor (Fe-FET) array using ambipolar tungsten diselenide (WSe2).
- Utilized polarization-programmed local fields for junction-barrier engineering in Fe-FETs.
- Demonstrated co-programming of photoresponsivity, multilevel conductance, and nonlinear transport within individual Fe-FET cells.
Main Results:
- Each Fe-FET cell was reconfigured for weighted sensing (ISC), linear accumulation (IMC), and hardware-native activation.
- The Fe-FET array functioned as a uniform pool of dynamically allocatable physical units.
- An end-to-end analog neuromorphic vision system was implemented on the Fe-FET platform, executing sensing, computation, and activation natively.
Conclusions:
- The Fe-FET array offers a task-adaptive and energy-efficient solution for neuromorphic vision hardware.
- This technology enables scalable hardware for edge intelligence by allowing dynamic function allocation without hardware changes.
- The developed platform paves the way for more flexible and efficient AI systems at the edge.
