INST’s new artificial synaptic chip to help transform information tech

NEW DELHI — Scientists at the Institute of Nano Science and Technology (INST), Mohali, an autonomous institution of the Department of Science and Technology (DST), have created an artificial synaptic device that emulates the behaviour of biological synapses to transform information technology through more efficient computing models. 

The team focussed on the workings of the human brain, which is dynamic and has a reconfigurable system. It has direct memory access, with neurons conducting the computational operations.

“Neuromorphic electronics, inspired by the intricate workings of the biological brain, offer the potential to transform information technology through more efficient computing models,” the researchers said.

On the other hand, modern computers have instructions that tell the computer how to handle information and perform tasks. This works by creating separate units for reading and executing complex operations, and then returning the results to memory. This can slow down computing and create a “bottleneck” when handling a lot of instructions and data at the same time.

INST scientists used a novel approach that follows the working principle of human brains to overcome the challenges of modern computing. They utilised the two-dimensional electron gas (2DEG) within the EuO-KTaO3 (KTO) oxide heterostructure to develop a chip which exhibited neuromorphic properties. It also showed a large change in resistivity known as resistive switching behaviour.

“The developed chip not only mimics the short- and long-term plasticity observed in biological synapses but also performs logic gate operations, significantly enhancing its versatility and potential for integration in advanced neuromorphic systems,” said the team.

Further, the neuromorphic design in oxide interfaces can facilitate more energy-efficient and quicker information processing. It also comes with enhanced AI capabilities and improved device miniaturisation. These systems may learn and change with time, resulting in more personalised and responsive technology.

The research was published in the journal ‘Applied Physics Letter’, and has important applications to improve everyday life, including healthcare, education, and environmental sustainability, said the researchers.