APPLICATIONS
Applications include pattern recognition, machine learning, artificial intelligence, robotics, and other areas where brain-inspired computing paradigms are beneficial.
APPLICATIONS
ADVANTAGES
Advantages include their low power consumption, high density integration, non-volatility, and compatibility with existing semiconductor fabrication processes. These characteristics make the proposed artificial neuron architecture suitable for large-scale neuromorphic systems and cognitive computing applications.
TECHNOLOGY DESCRIPTION
The invention addresses the need for efficient and scalable artificial neuron designs that can mimic the behavior of biological neurons in neural networks. It aims to provide a solution that allows for fast and energy-efficient information processing and learning capabilities.
The patent describes the use of diffusive memristors as key components in artificial neurons. Memristors are electronic devices that can change their resistance based on the history of applied voltages. Diffusive memristors, in particular, exhibit a gradual and reversible resistance change, making them suitable for emulating the synaptic connections between neurons.
The invention outlines a circuit architecture that integrates diffusive memristors with other electronic components to form artificial neurons. These neurons can receive input signals, perform computation, and produce output signals. The memristors play a crucial role in modulating the strength of the connections between artificial neurons, enabling synaptic plasticity and learning capabilities.
ABOUT THE INVENTOR
Dr. Qiangfei Xia is a professor of Electrical & Computer Engineering at UMass Amherst and head of the Nanodevices and Integrated Systems Laboratory. He received his Ph.D. in Electrical Engineering in 2007 from Princeton University
AVAILABILITY:
Available for Licensing and/or Sponsored Research
DOCKET:
UMA 17-036
PATENT STATUS:
US Patent Issued 11,586,864
NON-CONFIDENTIAL INVENTION DISCLOSURE
LEAD INVENTOR:
Dr. Qiangfei Xia
CONTACT: