Article

Researchers at Jiufengshan Lab in Hubei province have introduced a gallium nitride power module aimed at reducing energy consumption and costs for AI computing centers. Released on December 4, the module, developed under Li Sichao, replaces traditional silicon chips with gallium nitride semiconductors, achieving a 30% reduction in electrical power loss and a 50% decrease in production costs.

Implementing these modules in a 1-gigawatt AI data center could save approximately 300 million kilowatt-hours annually, resulting in a cost reduction of about 240 million yuan ($34 million). This innovation addresses the rising energy demands associated with the growing AI sector, where a 1 GW data center can consume energy comparable to a large nuclear plant.