Custom M8 Ultra chip cluster with NVIDIA interconnect eyed for Apple server push

According to The Information's reporting, the Cupertino technology manufacturer is designing enterprise-class infrastructure hardware around its forthcoming M8 Ultra processors. This system would potentially incorporate NVIDIA's NVLink Fusion technology as a link between multiple chips. The ambition represents an expansion of the company's semiconductor strategy into data centre equipment, a market segment the manufacturer abandoned when Xserve production ended over a decade ago. Sources indicate the effort commenced roughly 12 months prior with endorsement from its current chief executive John Ternus. Timeline expectations suggest no product availability until 2029 at the earliest, with possibilities ranging from full cancellation to pivoting toward competing connection technologies.

1. The manufacturer is developing an AI server using planned M8 Ultra chips, according to The Information
2. NVIDIA's NVLink Fusion technology is under consideration as a processor interconnect solution for the device
3. Development commenced approximately one year ago with current CEO John Ternus providing support
4. No commercial launch is anticipated prior to 2029
5. The initiative faces uncertainty and could be abandoned entirely or modified to use alternative interconnect approaches