Researchers at the University of Pittsburgh developed a graphene-based artificial synapse that does not process information such as a digital computer, but mimics the analog path that the human brain has completed its tasks. The synapse showed excellent energy efficiency compared to biological synapses.

In this study, "The analog nature and massive paralysis of the brain is in part part of what people can outperform even the most powerful computers when it comes to complex cognitive functions such as voice recognition or pattern recognition in complex and diverse datasets." Dr. Xiong says.

A new area called a neuromorphic calculation focuses on computational hardware design inspired by the human brain. The conductive properties of graphene allow researchers to precisely adjust the electrical conductivity of the synaptic connection or the power of the synaptic weight.

In the recent revival of artificial intelligence, computers can reproduce the brain in certain ways, but about a dozen digital devices are needed to mimic an analog synapse. The human brain has hundreds of trillion synapses to transmit information, so it is impossible to construct a brain with digital devices or at least not be scaled. Xiong Lab's approach provides a possible way to implement hardware for large-scale artificial neural networks.

According to Xiong, artificial neural networks based on existing CMOS (complementary metal oxide semiconductor) technology will always have limited functionality in terms of energy efficiency, scalability and packaging density. "It's really important that we develop new device concepts for synaptic electronics that are suitable for analog, scalable and large-scale integrations in nature." Xiong says.

By strengthening the primitive intelligence level in wearable electronic devices and sensors, we can monitor our health with intelligent sensors, provide preventive maintenance and timely diagnostics, monitor the growth of plants and detect possible harmful problems and regulate and optimize the production process. This will significantly contribute to improving overall productivity and quality of life in our society.

The development of an artificial brain that functions like an analogue human brain is not yet appropriate and requires a series of breakthroughs. Researchers should find the correct configurations to optimize these new artificial synapses. They will need to make them compatible with a range of other devices to create neural networks and ensure that all artificial synapses in a large-scale neural network behave the same way. Despite the difficulties, Dr. Xiong says he is optimistic about the direction.

Advanced Materials

About | Contact | Privacy | Terms     © 2014-2019 JungleVoy All Rights Reserved.

Developed and designed by

Pixabay and Unsplash sites have been used for some images on our site.