Neuromorphic behaviour in discontinuous metal films.

Area of Science:

• Materials Science
• Neuroscience
• Computer Science

Background:

• Neuromorphic computing aims to create brain-like computational systems.
• Understanding physical systems that mimic neural behavior is crucial for advancing this field.

Purpose of the Study:

• To investigate discontinuous metal films as a potential platform for neuromorphic computing.
• To analyze the electrical signal characteristics and underlying mechanisms in these films.

Main Methods:

• Fabrication of discontinuous metal films using simple evaporation.
• Experimental measurement of electrical signals and their avalanches.
• Analysis of atomic-scale switching processes.
• Numerical simulations and circuit modeling.

Main Results:

• Discontinuous metal films exhibit correlated electrical signal avalanches, similar to cortical activity.
• These signals meet established criteria for criticality.
• Atomic-scale switching processes mimic the integrate-and-fire mechanism of biological neurons.
• Switching event characteristics depend on network state and local position.

Conclusions:

• Discontinuous metal films demonstrate brain-like behavior and criticality.
• The observed switching mechanisms align with biological neural processes.
• These films represent a promising and accessible platform for brain-inspired computing.