Apple in talks with Foxconn to build AI servers in Taiwan

Apple wants to tap the talent at Foxconn to power its Apple Intelligence servers, though the supplier may already be at capacity.

Apple has been ramping up efforts for Apple Intelligence, its user-centric suite of artificial intelligence tools. Consequently, the company aims to manufacture specialized servers to enhance its computational power.

Currently, the company is in talks with Foxconn, its largest iPhone supplier. The Taiwanese supplier already provides a similar service for Nvidia.

Apple hopes to build its servers using Apple Silicon, which is already used on the user-side of Apple Intelligence features in Macs and iPads. According to Nikkei Asia, Apple's volume of servers would be relatively small compared to Nvidia's.

Nvidia's GB200 system is currently the world's most powerful AI computing vehicle. It follows that Apple would want to utilize Foxconn's resources to power its own.

However, Foxconn's relationship with Nvidia could potentially be the reason it can't commit to Apple.

According to anonymous sources from Nikkei Asia, Foxconn is currently preoccupied with constructing server centers for Nvidia. Furthermore, Apple's relatively recent entry into the market would likely require additional support during the initial stages, potentially straining the already limited resources available at Foxconn.

As a result, Apple has begun exploring other partnerships, including China's Lenovo Group and its subsidiary LCFC, and smaller suppliers such as Universal Scientific Industrial. Allegedly, Apple's talks with Lenovo involve building servers outside of China.

Apple has begun rolling out Apple Intelligence features to compatible devices, such as the iPhone 15 Pro line, the iPhone 16 lineup, and its M-series iPads and Macs. As of now, Apple Intelligence is limited to U.S. English, but the company plans to expand its language catalog soon.