Researchers Create Brain-Like Transistor for AI
-
Researchers developed a new synaptic transistor that mimics the brain's integrated processing and memory capabilities and can perform complex cognitive tasks like associative learning.
-
The transistor operates at room temperature, consumes little energy, retains information when powered off, and is stable, making it practical for real-world applications.
-
The device combines graphene and boron nitride layered to create a moiré pattern, allowing researchers to achieve neuromorphic properties by tuning the twist between the layers.
-
Experiments showed the transistor could be trained to recognize and categorize similar patterns, demonstrating capabilities beyond simple machine learning.
-
The technology marks significant progress toward energy-efficient hardware that can advance artificial intelligence and perform human-like information processing.