Apple is developing enterprise-grade AI servers, signaling a return to the hardware market it exited after the Xserve. The rack-mounted systems will feature future M8 Ultra processors.
Configurations will include either two or four chips per server. Apple targets a release for the new hardware line around 2029.
The company has discussed incorporating Nvidia’s NVLink Fusion networking technology into the design. This high-bandwidth connection would integrate custom Apple silicon into broader data center infrastructures. The project leverages Apple’s chip expertise to compete in the AI hardware market.