Related Experiment Video
Updated: Sep 7, 2025
08:07
Assembly and Characterization of Biomolecular Memristors Consisting of Ion Channel-doped Lipid Membranes
Published on: March 9, 2019
7.9K
Neuromorphic Photonic Memory Devices Using Ultrafast, Non-Volatile Phase-Change Materials
Xiaozhang Chen1, Yuan Xue2, Yibo Sun1
1State Key Laboratory of ASIC and System, Department of Microelectronics, Fudan University, Shanghai, 200433, P. R. China.
Advanced Materials (Deerfield Beach, Fla.)
|June 17, 2022
Summary
Researchers developed ultrafast photonic memory using scandium-doped antimony telluride (SST), achieving a record 2 ns write/erase speed. This non-volatile memory shows potential for AI and photonic computing applications.
Area of Science:
- Photonics and Materials Science
- Integrated Optics
- Non-volatile Memory Technologies
Background:
- Growing demand for high-speed, low-energy data processing in cloud computing, AI, and supercomputing necessitates advanced memory solutions.
- Optical domain signal processing offers advantages in speed, bandwidth, and energy efficiency.
- Silicon photonics integrated with phase-change materials (PCMs) enable non-volatile photonic memory.
Purpose of the Study:
- To develop and demonstrate an ultrafast non-volatile photonic memory device.
- To investigate the potential of scandium-doped antimony telluride (SST) for high-speed phase-change applications.
- To explore the application of SST-based photonic memory in artificial neural networks and nanodisplays.
Main Methods:
- Fabrication of an integrated photonic memory device utilizing an SST thin film.
- Characterization of the SST thin film's phase transition dynamics and memory performance.
- Implementation of an artificial neural network for image classification using the photonic memory.
- Demonstration of a reflective nanodisplay application with optoelectronic modulation.
Main Results:
- Achieved an unprecedented 2 ns write/erase speed for integrated phase-change photonic devices, the fastest reported to date.
- Demonstrated multilevel capabilities and room-temperature stability for SST-based photonic memories.
- Successfully established an artificial neural network for image classification by mapping memory levels to synaptic weights.
- Showcased a reflective nanodisplay application leveraging SST's optoelectronic modulation.
Conclusions:
- SST is a promising material for ultrafast non-volatile photonic memory, offering significant improvements in switching speed.
- The developed photonic memory devices exhibit potential for advanced applications in photonic computing, neuromorphic computing, and optoelectronics.
- The integration of SST into silicon photonics paves the way for next-generation nanophotonic devices and systems.
