Scientists have developed an innovative synaptic transistor inspired by the human brain, demonstrating the ability to process and store information simultaneously. Unlike previous brain-like computing devices, this transistor remains stable at room temperature, operates efficiently, consumes minimal energy, and retains information even when powered off.
READ: “Taylor Swift Drops ‘1989 (Taylor’s Version)’: What’s Inside the New Album Will Shock You!
Human Brain
The breakthrough technology combines bilayer graphene and hexagonal boron nitride in a moiré pattern to achieve neuromorphic functionality, recognizing patterns and demonstrating associative learning.
This advancement marks a significant departure from traditional transistor-based computing, aiming to enhance energy efficiency and cognitive capabilities for AI and machine learning tasks.
The researchers, from Northwestern University, Boston College, and MIT, designed the synaptic transistor to mimic the brain’s integrated architecture, achieving concurrent memory and information processing functionality. Northwestern’s Mark C. Hersam, co-leading the research, emphasized the distinction between the brain and digital computers, highlighting the higher energy efficiency of the brain’s co-located memory and information processing. The new device, stable at room temperature, operates rapidly, consumes minimal energy, and retains information even without power, making it well-suited for practical applications.
The study, set to be published in the journal Nature, showcases the transistor’s ability to surpass simple machine-learning tasks, categorizing data and performing associative learning even when faced with incomplete patterns. With conventional computing systems consuming significant energy due to the separation of processing and storage units, the synaptic transistor offers a promising solution for more energy-efficient and advanced AI systems. Hersam and the team explored moiré patterns, stacking atomically thin materials, such as bilayer graphene and hexagonal boron nitride, to achieve unprecedented tunability of electronic properties in the new device.
