Journal
NANO LETTERS
Volume 23, Issue 19, Pages 9020-9025Publisher
AMER CHEMICAL SOC
DOI: 10.1021/acs.nanolett.3c02681
Keywords
neuromorphic devices; vanadium dioxide; artificialneural networks; photothermal electronics
Ask authors/readers for more resources
Biological nervous systems rely on complex, dynamic connectivity among billions of neurons, while artificial intelligence and neuromorphic computing platforms seek to mimic biological cognition. To incorporate the advantages of tunable dynamic software implementations into hardware, researchers have developed an artificial synapse with adaptable resistivity.
Biological nervous systems rely on the coordination of billions of neurons with complex, dynamic connectivity to enable the ability to process information and form memories. In turn, artificial intelligence and neuromorphic computing platforms have sought to mimic biological cognition through software-based neural networks and hardware demonstrations utilizing memristive circuitry with fixed dynamics. To incorporate the advantages of tunable dynamic software implementations of neural networks into hardware, we develop a proof-of-concept artificial synapse with adaptable resistivity. This synapse leverages the photothermally induced local phase transition of VO2 thin films by temporally modulated laser pulses. Such a process quickly modifies the conductivity of the film site-selectively by a factor of 500 to activate these neurons and store memory by applying varying bias voltages to induce self-sustained Joule heating between electrodes after activation with a laser. These synapses are demonstrated to undergo a complete heating and cooling cycle in less than 120 ns.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available