M8-powered AI server could mark Apple’s return to enterprise hardware

Apple is reportedly exploring a return to the server market with an AI‑focused system built around upcoming M8 series chips. The project, still early and not guaranteed to ship, would target companies that want to run AI inference workloads on their own hardware rather than relying on cloud GPUs. Apple is considering using Nvidia’s NVLink Fusion to link multiple M8 processors, as its current interconnect solutions may be too slow or costly at server scale. The system could launch in 2029, though cancellation remains possible. CEO John Ternus is said to support the initiative, which follows rising demand from AI firms for Mac mini and Mac Studio systems. Apple already operates its own server hardware for Private Cloud Compute but does not offer it to external customers. A commercial server would mark Apple’s first return to enterprise infrastructure since discontinuing Xserve in 2011. The project also signals a warming relationship between Apple and Nvidia, with both companies exploring deeper collaboration in AI. Apple still faces challenges, including enterprise support expectations and the need to expand its MLX machine‑learning framework.

Read the full story on MacRumors →