Related Experiment Video
Updated: Sep 3, 2025
08:07
Assembly and Characterization of Biomolecular Memristors Consisting of Ion Channel-doped Lipid Membranes
Published on: March 9, 2019
7.9K
Nanosecond protonic programmable resistors for analog deep learning
Murat Onen1,2, Nicolas Emond2,3, Baoming Wang2,3
1Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, 77 Massachusetts Ave., Cambridge, MA 02139, USA.
Science (New York, N.Y.)
|July 28, 2022
Summary
Researchers developed nanoscale protonic programmable resistors for analog deep learning. These artificial synapses operate efficiently under extreme electric fields, outperforming biological neurons and synapses in speed and energy usage.
Area of Science:
- Materials Science
- Neuroscience
- Electrical Engineering
Background:
- Nanoscale ionic programmable resistors are being explored for analog deep learning applications.
- The performance comparison between these artificial synapses and biological neurons/synapses, especially concerning speed, remains unclear.
- Operation under extreme electric fields within solid electrolytes is crucial for understanding device behavior.
Purpose of the Study:
- To develop silicon-compatible nanoscale protonic programmable resistors capable of operating under extreme electric fields.
- To investigate the operational characteristics, including speed and energy efficiency, of these artificial synapses.
- To assess the potential for these devices to surpass the performance of biological neurons and synapses.
Main Methods:
- Performed scaling analyses of ionic transport and charge-transfer reaction rates.
- Fabricated silicon-compatible nanoscale protonic programmable resistors.
- Tested device performance under extreme electric fields, focusing on proton shuttling and intercalation.
- Evaluated modulation characteristics, dynamic range, and energy efficiency at room temperature.
Main Results:
- Successfully generated nanoscale protonic programmable resistors with desirable characteristics under extreme electric fields.
- Achieved controlled proton shuttling and intercalation in nanoseconds at room temperature, demonstrating energy efficiency.
- Observed symmetric, linear, and reversible modulation with a 20x dynamic range.
- Indicated that the space-time-energy performance can significantly exceed biological counterparts.
Conclusions:
- Nanoscale protonic programmable resistors operating under extreme electric fields offer a promising pathway for advanced analog deep learning.
- The developed artificial synapses demonstrate superior speed and energy efficiency compared to biological neurons and synapses.
- These findings pave the way for next-generation neuromorphic computing hardware with enhanced capabilities.
