The Information reported on Sept. 16 that Apple is developing enterprise servers based on Nvidia networking equipment and its own chips, citing sources familiar with internal details.

It is part of an effort to expand its stake as demand surges for computers used for AI.

Apple sold its own server product, Xserve, from 2002 to 2011, but it did not achieve major results due to a lack of support for corporate customers. Attention is on whether it can secure a meaningful presence this time.

According to the report, Apple plans to sell the servers to AI developers, businesses and governments. Two types are under review: a smaller model combining two M8 Ultra chips and a larger model equipped with four.

The M-series Ultra chip is Apple’s highest-performance chip and is used in products such as the Mac Studio.

Apple has been considering using Nvidia’s NVLink Fusion to connect M8 chips. NVLink Fusion was originally designed to link Nvidia chips, but it can now connect chips from other companies. Some people involved in the Apple project see NVLink as the most suitable connection method, the sources said.

Apple is developing the servers with a 2029 launch in mind, but the project could be cancelled or could proceed without using Nvidia technology, The Information said.

Apple and Nvidia began cooperating in 2001, but their relationship cooled under Steve Jobs due to patent disputes and other issues. Recently, however, ties have improved as Apple processes its new AI-based Siri on Nvidia chips in Google data centers.

At Apple, the Mac computer lineup posted a 29 percent jump in revenue in the previous quarter to $10.4 billion, emerging as the company’s fastest-growing area. AI developers are buying Mac minis and Mac Studios in bulk, leading to supply shortages. But Macs are still closer to consumer products, and some point to a lack of enterprise features such as remote management.