New Neuromorphic Chip Mimics Cerebellum for Efficient AI Control
TL;DR. Scientists developed a new AI chip mimicking the human cerebellum, designed for rapid, low-power processing of unusual events. - The brain-inspired device ignores routine data, only processing unexpected signals to conserve energy. - It identified irregular heartbeats with 98% accuracy, twice as fast as conventional AI systems. - This neuromorphic approach could lead to more efficient AI in health monitors and autonomous systems.
- A new AI chip mimics the human cerebellum for instinctive motor responses and event detection.
- The chip ignores routine information, focusing only on unexpected events to save power.
- It demonstrated 98% accuracy in flagging irregular heartbeats, performing twice as fast as conventional AI.
- This neuromorphic approach aims to reduce reliance on power-intensive data centers for AI processing.
- Potential applications include always-on health monitors, self-driving cars, and autonomous robots.