Apple Develops Enterprise AI Servers Using Nvidia Technology

Apple Inc. is developing a line of enterprise servers designed for artificial intelligence developers, corporate clients, and governments. The project, which has been in development for approximately one year, aims to meet rising demand from AI developers for Mac hardware. The servers are planned for a 2029 market entry, though the project remains subject to potential cancellation as plans are not yet finalized.

The servers will be available in two configurations powered by forthcoming M8 Ultra chips, featuring either two or four chips per unit. To facilitate high-speed communication between these chips, Apple is in discussions with Nvidia Corp. to incorporate NVLink Fusion, a networking platform consisting of software, chiplets, and switches.

Apple executive John Ternus is reportedly in favor of the project. Unlike internal infrastructure, these servers are intended for commercial sale to external businesses and government entities.