Related Experiment Video
Updated: Jun 28, 2025
08:07
Assembly and Characterization of Biomolecular Memristors Consisting of Ion Channel-doped Lipid Membranes
Published on: March 9, 2019
7.8K
Enhanced read resolution in reconfigurable memristive synapses for Spiking Neural Networks
Hritom Das1, Catherine Schuman2, Nishith N Chakraborty2
1Department of Electrical Engineering and Computer Science, University of Tennessee, Knoxville, TN, 37996, USA. hdas@utk.edu.
Scientific Reports
|April 17, 2024
Summary
This study enhances memristive synapse reliability by improving READ current resolution. Techniques like voltage scaling and body biasing significantly boost performance in neuromorphic computing applications.
Area of Science:
- Neuromorphic Engineering
- Materials Science
Background:
- Memristive synapses are crucial for neuromorphic computing but face challenges like low READ resolution, impacting reliability.
- Degraded distinguishability of stored weights in memristive synapses hinders performance.
Purpose of the Study:
- To analyze and enhance the READ current resolution in current-controlled memristor-based synapses.
- To improve the reliability and accuracy of memristive synapses for neuromorphic applications.
Main Methods:
- Utilized an empirical model to characterize the memristive device.
- Implemented scaling of specific device stages to improve READ current margin.
- Applied run-time adaptation techniques including READ voltage scaling and body biasing.
Main Results:
- Achieved enhanced READ current margin up to 4.3% and 21% through device stage scaling.
- Improved READ current resolution by approximately 46% with READ voltage scaling and 15% with body biasing.
- Demonstrated improved classification, control, and reservoir computing accuracy with higher READ current resolution.
Conclusions:
- Enhanced READ current resolution is critical for reliable memristive synapse operation.
- Proposed techniques effectively improve READ current resolution and overall neuromorphic system performance.
- Higher resolution leads to better accuracy, even under noisy conditions.
