A team of scientists from the Jawaharlal Nehru Center for Advanced Scientific Research (JNCASR) in Bengaluru has developed a device that can simulate the cognitive functions of the human brain, and is more efficient than conventional techniques. New AI technology increases computational speed and power consumption efficiency. The team created an artificial synaptic network (ASN) similar to a biological neural network, through a simple self-forming method. The scientists said that the structure of the device is formed automatically when heated. Aiming to develop a synaptic device for neuromorphic applications, the team also discovered a physical system that mimics neuronal bodies and axonal network connectivity, similar to a biological system.
Scientists said that there are 100 billion neurons in the human brain that contain axons and dendrites. These neurons connect with each other via axons and dendrites and form huge junctions called synapses. This complex bio-neural network gives rise to improved cognitive abilities, scientists believe, adding to artificial neural networks (ANNs) based on software, beating humans in games such as AlphaGo and AlphaZero, and even beating humans. That was seen helping in handling the COVID-19 situation.
The researchers said that while the power-hungry von Neumann computer architecture slows down ANN performance due to the serial processing available, the brain operates through parallel processing, consuming only 20W of power.
The scientists said the JNCASR team dewetted silver metal to form branched islands and nanoparticles with nanogap separations resembling bio neurons and neurotransmitters, where dewetting is the process of breaking up of the continuous film into disconnected/isolated islands or spherical particles.
"With such an architecture, many higher-order cognitive activities are simulated," the team explained. Ministry of Science and Technology Press release.
They postulated that using programmed electrical signals as real-world stimuli, this hierarchical structure influenced various learning activities such as short-term memory (STM), long-term memory (LTM), ability, depression, associative learning, interest. -based learning. , supervision, impression of supervision, and more.
Not only that, synaptic fatigue was also mimicked after excessive learning and its self-retrieval, and remarkably, these behaviors were simulated in a single material system without the aid of external CMOS circuits, the scientists said.
The team has developed a prototype kit to emulate Pavlov's dog behavior that demonstrates the potential of this device towards neuromorphic artificial intelligence. This is a remarkable achievement as the JNCASR team has taken a step forward towards accomplishing advanced neuromorphic artificial intelligence by organizing a nanomaterial similar to biological neural material.