Synchrony in silicon: the gamma rhythm.

Area of Science:

• Neuroscience
• Neuromorphic Engineering
• Computational Neuroscience

Background:

• Gamma rhythm (20-80 Hz) is crucial for neuronal computation but underrepresented in neuromorphic systems.
• Neuromorphic systems use mixed analog and digital circuits to mimic neurobiology.
• Synchronized neuronal activity is key to brain function, including object binding.

Purpose of the Study:

• To present a network of silicon interneurons synchronized in the gamma frequency range.
• To investigate the role of shunting inhibition and synaptic rise time in neuronal synchrony.
• To implement object binding using synchronized neuromorphic interneurons.

Main Methods:

• Fabrication of interneurons with soma and synapse circuits in 0.25-microm CMOS technology.
• Utilizing shunting inhibition with synaptic rise time for neuronal synchronization.
• Characterization of a network of 256 synchronized interneurons.

Main Results:

• The silicon interneurons synchronize effectively in the gamma frequency range.
• Synaptic rise time was found to be proportional to the spiking period of synchronized interneurons.
• The synchronized interneuron network was used to entrain model excitatory principal neurons.

Conclusions:

• The developed silicon interneurons successfully synchronize at gamma frequencies, mimicking biological neural networks.
• The integration of shunting inhibition and synaptic rise time is an effective strategy for achieving neuronal synchrony in neuromorphic systems.
• This work demonstrates a novel approach for implementing object binding in neuromorphic hardware.