Apple weighs Nvidia technology for potential server market return, The Information reports

Apple is reportedly exploring the possibility of returning to the server market with AI-focused hardware, according to a report from The Information. The company may introduce a server utilizing M8 series chips and Nvidia networking hardware, potentially bringing it back into a business it left when it discontinued Xserve in 2011.

This server would target businesses seeking to run AI models on their own equipment, with a specific focus on inference and generating responses from trained models. Nvidia's NVLink Fusion technology is under consideration, which would enable data exchange between chips. While the server and its final form are not confirmed, a potential release date is mentioned as 2029.

Apple CEO John Ternus reportedly backed the initiative when it began approximately a year ago during his time as the company's hardware engineering chief. The project comes after a surge in interest from AI businesses purchasing Apple hardware like Mac mini and Mac Studio systems. Currently, Apple provides server hardware for Private Cloud Compute, which handles AI tasks demanding processing beyond a user's device.

Scaling existing chip connections poses cost and speed challenges, hence the consideration of Nvidia's technology. Apple has also turned down partners' requests to use Private Cloud Compute servers. The collaboration with Nvidia signifies a potential shift in their long-standing disputes. Apple has announced plans to extend Private Cloud Compute to Google Cloud using Nvidia GPUs and has explored other AI-related collaborations, including Nvidia's open-source models.

However, the project still faces hurdles such as the need for stronger support for business customers and additional software resources for AI developers, including a further investment in its MLX framework.