Apple develops AI servers using M8 Ultra chips and NVIDIA technology

Apple is reportedly developing specialized enterprise AI servers, marking a potential return to the commercial server market for the first time since discontinuing its Xserve line in 2011. The project, which has been in development for approximately one year, is reportedly supported by CEO John Ternus.

The proposed servers would leverage Apple’s future M8 Ultra chips, with configurations potentially featuring either two or four chips per unit. To address the technical challenge of high-speed interconnectivity between multiple chips, Apple is in discussions with NVIDIA regarding the integration of its NVLink Fusion technology.

This strategic move is driven by surging demand for Apple Silicon in AI workloads. Companies like OpenAI and Anthropic have reportedly purchased tens of thousands of Mac mini and Mac Studio units for AI development and inference tasks. By offering dedicated server hardware, Apple aims to provide third-party developers, corporations, and government agencies with secure, on-premises infrastructure for running AI models.

While the project is ambitious, it remains in the early stages and faces uncertainties, including potential cancellation or changes in technical partnerships. If realized, a market launch is anticipated around 2029.