Hexagonal magnetic defects could lead to energy-efficient neuromorphic computing

Artificial intelligence applications are experiencing a boom and expected to be mainstream technologies in the near future. However, these applications run on classic computing hardware and are extremely power-hungry.