Researchers at Brown University have developed a new technique that could pave the way for significantly more energy-efficient computer chips, potentially reducing the electricity consumption of artificial intelligence systems. The breakthrough involves multiferroics, a class of materials combining electrical and magnetic properties that can be controlled using very small voltages. Until now, their practical application in computer chips has been limited by the so-called ‘dead layer’ effect, which causes these properties to disappear when the materials are made extremely thin. The findings were published in the journal Nature Materials.