A research team at Technische Universität Braunschweig’s Nitride Technology Centre is exploring innovative uses for energy-efficient microLEDs, aiming to revolutionize artificial intelligence hardware. By miniaturizing and scaling these microscopic LEDs, the team — in collaboration with Ostfalia University and ams OSRAM — is developing a neuromorphic computing system that mimics the human brain by utilizing optical components to reduce the energy demands typically required by conventional digital tech. This advancement in gallium nitride (GaN) technology signals a potential for significantly lower power consumption in AI applications, theoretically reducing energy usage by a factor of 10,000. Their initial success includes a demonstrator capable of recognizing handwritten numbers, showcasing the possibilities that lie ahead in AI processing and efficiency.