Physics-driven human-like working memory outperforms digital networks in dynamic vision

Researchers have developed a physics-driven AI system called Intrinsic Plasticity Network (IPNet) that uses magnetic tunnel junctions to create human-like working memory. The system demonstrates 18x error reduction in dynamic vision tasks while reducing memory-energy overhead by over 90,000x compared to traditional digital AI systems.

• →IPNet leverages Joule-heating relaxation dynamics of magnetic tunnel junctions to create neuromorphic working memory.
• →The system achieves 18x error reduction compared to spatiotemporal convolutional models in dynamic vision tasks.
• →Memory-energy overhead is reduced by more than 90,000x compared to traditional digital AI systems.
• →In autonomous driving applications, IPNet reduces prediction errors by 12.4% versus recurrent networks.
• →The technology represents a potential breakthrough in sustainable AI computing by addressing energy consumption challenges.