Apple Reportedly Planning AI Servers Powered by M8 Ultra Chips for Enterprise Market
Apple is developing enterprise-grade AI servers powered by its upcoming M8 Ultra chips, according to reports from September 2026. The project targets AI developers, businesses, and government clients, and would mark Apple's return to the server market for the first time since it discontinued the Xserve in 2011. A potential release is targeted for 2029, though the project could still be canceled.
Apple is developing enterprise-grade AI servers powered by its upcoming M8 Ultra chips, according to reports published in September 2026. The project would mark the company's first entry into the dedicated server market since it discontinued the Xserve in January 2011.
The servers are being designed in two configurations. A smaller version would use two clustered M8 Ultra chips. A larger, more powerful model would pack four M8 Ultra processors. Apple has evaluated using Nvidia's NVLink Fusion technology to handle communication between chips within a data center environment, though the company could move forward without it.
The initiative gained internal momentum under John Ternus, who supported the project while serving as head of hardware engineering before becoming CEO in September 2026. The target market includes AI developers, startups, businesses, and government entities that currently use Apple hardware like the Mac Studio and Mac mini for local AI compute.
Apple's goal is to offer a scalable, integrated alternative to traditional server setups built around x86 processors and discrete graphics cards.
The project is currently targeted for a 2029 release, but it remains in an early planning phase. Apple has not finalized the plans, and the company could cancel the program. The initiative also faces broader industry challenges, including a global shortage of memory chips needed for large-scale AI data centers.
No official announcement has been made by Apple.