Summary
Modern artificial intelligence (AI) is inspired by the brain’s cognitive functions but relies on models that differ greatly from biological systems and consume substantial energy during training and inference. According to the Semiconductor Research Corporation, continued scaling of logic devices and increasing model complexity could push machine learning energy consumption beyond global energy production capacity—an unsustainable trajectory. In contrast, the human brain performs complex computations with vastly lower energy. To bridge this gap, this project proposes a novel three-terminal tra
