Apple is considering a return to the commercial server market with an AI system built around future M8-series processors and Nvidia networking technology, according to The Information. The proposed hardware would be sold to outside companies that want to run artificial intelligence models on their own equipment.

The project is still under discussion and could be canceled before launch. If it moves forward, a release could arrive in 2029, placing Apple back in the server business more than a decade after it discontinued the Xserve line in 2011.

Apple’s potential AI server strategy

The proposed system would reportedly focus on AI inference, the process of using trained models to generate responses and other results. That could make it relevant to businesses seeking more control over their AI workloads, including how data is processed and where systems are operated.

Apple has not finalized the server or its relationship with Nvidia. One option under consideration is Nvidia’s NVLink Fusion platform, which combines hardware and software for moving data between processors. Apple could use the technology to connect multiple M8 chips in a larger system.

Scaling communication between Apple processors is reportedly one of the main technical and financial challenges. Nvidia’s interconnect technology could help address those constraints, although the details of any potential design remain unresolved.

Why Nvidia networking could matter

An agreement would give Nvidia a role in an Apple product that could compete with Nvidia’s own AI systems. NVLink Fusion is designed to let Nvidia provide networking technology to companies developing processors that are not based on Nvidia chips, broadening the company’s reach across the AI hardware market.

The discussions also point to a potentially closer relationship between the two companies after years of disagreements. Apple has already announced plans to extend its Private Cloud Compute infrastructure to Google Cloud using Nvidia GPUs. The companies have also reportedly considered other AI cooperation, including the use of Nvidia’s open-source models.

Apple’s existing server infrastructure

Apple already operates specialized servers for Private Cloud Compute, which handles Apple Intelligence requests that require processing beyond a user’s device. However, Apple has reportedly rejected requests from partners to use those servers.

The potential commercial server project would therefore represent a separate offering rather than a simple expansion of Private Cloud Compute. Apple would need to address the demands of business customers, including broader support and additional software tools for developers.

Software support remains a challenge

Apple’s MLX framework is another area where further investment may be needed. MLX supports machine learning development on Apple silicon, but a server business aimed at AI companies would require a strong software ecosystem as well as capable hardware.

AI businesses are already purchasing Mac mini and Mac Studio systems in large quantities, according to the report. That interest may be encouraging Apple to consider a more purpose-built system, but the possible 2029 launch remains distant and far from guaranteed.