Next-generation AI semiconductor devices mimic the human brain

A analysis group led by Prof. Kwon Hyuk-jun of the DGIST Division of Electrical Engineering and Laptop Science has developed a next-generation AI semiconductor expertise that mimics the human mind’s effectivity in AI and neuromorphic methods.

The development of AI has stimulated a quickly rising demand for energy-efficient semiconductor expertise with a quick operational velocity. Nonetheless, conventional computing gadgets with their von Neumann structure and separate computing and reminiscence items have velocity and power effectivity shortcomings related to knowledge processing bottlenecks. Consequently, analysis on neuromorphic gadgets that mimic organic neurons’ simultaneous computing and reminiscence capabilities is gaining consideration.

In opposition to this backdrop, Prof. Hyuk-Jun Kwon’s group developed synaptic field-effect transistors utilizing hafnium oxide, which has sturdy electrical properties, and skinny layers of tin disulfide. This resulted in a three-terminal neuromorphic machine able to storing a number of ranges of knowledge in a fashion much like neurons.

The analysis efficiently replicated organic traits corresponding to short- and long-term properties, yielding a extremely environment friendly machine that responds 10,000 instances quicker than human synapses and consumes little or no power.

Prof. Hyuk-Jun Kwon of the Division of Electrical Engineering and Laptop Science mentioned, “This analysis marks an necessary step towards next-generation computing structure, which requires low power consumption and high-speed computation. We now have developed high-performance neuromorphic {hardware} utilizing two-dimensional channels and ferroelectric hafnium oxide, and the innovation is predicted to have numerous AI and machine learning-related purposes sooner or later.”

The analysis is published within the journal Superior Science.