%100 original translation content

Researchers at the University of Pittsburgh have developed a graphene-based artificial synapse that mimics the analogue path that does not process information like a digital computer, but the tasks of the human brain. Synapse showed excellent energy efficiency when compared to biological synapses.
Dr. "The analog nature of the brain and massive paralysis can partly leave behind even the most powerful computers when it comes to higher levels of cognitive functions, such as voice recognition or pattern recognition in complex and diverse data sets of people," says Xiong.
A new field called neuromorphic computing focuses on the design of computational hardware inspired by the human brain. The conductive properties of the graphene allow researchers to precisely adjust the electrical conductivity, which is the power of synaptic connectivity or synaptic weight.
Recently , in the revival of artificial intelligence , computers can reproduce the brain in certain forms, but about a dozen digital devices are needed to mimic analog synapses. The human brain has hundreds of trillion synapses to transmit information, so building a brain with digital devices is impossible or at least not scalable. The Xiong Lab 's approach provides a possible way for hardware implementation of large-scale artificial neural networks.
According to Xiong, artificial neural networks based on current CMOS (complementary metal oxide semiconductor) technology will always have limited functionality in terms of energy efficiency, scalability and packaging density. Xiong "It's really important for us to develop new device concepts for nature analog, scalable and large-scale integrations for synaptic electronics. Grafen's synapse seems to be controlling all the boxes on these terms so far, "he says.
By strengthening the level of primitive intelligence in wearable electronic devices and sensors, we can monitor health through intelligent sensors, provide preventive maintenance and timely diagnostics, monitor plant growth, identify potential harmful problems, and regulate and optimize production processes. This will in large measure contribute to improving the 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 well suited and requires a series of breakthroughs. Researchers must find the right configurations to optimize these new artificial synapses. They will need to make them adapt to a number of other networks to form neural networks, and they need to be sure that all artificial synapses in a large-scale neural network behave the same. Despite the difficulties, Xiong says he is optimistic about the direction he is managing.
Source: poxox.com science

No comments:

Post a Comment

Bottom Ad [Post Page]