New brain-like transistor performs energy-efficient associative learning at room temperature

In a breakthrough development, researchers at Northwestern University, Boston College, and MIT have created a brain-inspired synaptic transistor that mimics human intelligence. This transistor processes and stores information simultaneously, introducing a new era of neuromorphic computing that aligns closely with the brain’s architecture. Unlike previous brain-like computing devices, this transistor is stable at room temperature, operates at high speeds, consumes minimal energy, and retains stored information even when power is removed. The research, titled “Moiré synaptic transistor with room-temperature neuromorphic functionality,” published in the journal _Nature_, signifies a significant advancement in AI and machine learning technologies.

The team led by Northwestern’s Mark C. Hersam leveraged moiré patterns, a unique geometrical design arising from stacked two-dimensional materials, to engineer this innovative transistor. By combining bilayer graphene and hexagonal boron nitride and purposefully twisting them to form a moiré pattern, the researchers achieved unprecedented tunability of electronic properties at room temperature. The transistor was trained to recognize similar but not identical patterns, successfully displaying associative memory and higher-level cognition in experiments, setting it apart from traditional AI algorithms.

The new brain-like transistor addresses the inherent limitations of conventional digital computing systems by offering concurrent memory and information processing, significantly enhancing energy efficiency. Hersam and his team’s groundbreaking exploration of moiré physics and advanced materials presents a paradigm shift in computing hardware, especially for AI and machine-learning tasks. Through this innovative development, the researchers aim to steer AI technology towards higher-level thinking, which can better handle complex real-world conditions, making it a potential game-changer for various industries, including autonomous vehicles and data-intensive applications.