Apple weighs its first server since the Xserve, and as always, Nvidia may hold the key

Apple is reportedly exploring the development of its first dedicated enterprise AI server since the discontinuation of the Xserve in 2011. According to reports, the company is considering configurations featuring its future M8 Ultra chips, aimed at businesses seeking to run AI inference on-premises rather than in the public cloud. A potential launch is projected for 2029, though the project remains unfinalized. To overcome current bandwidth limitations inherent in Thunderbolt-based clustering, Apple is allegedly considering a partnership with Nvidia to utilize NVLink Fusion technology. This move follows a surge in demand for Apple’s Mac hardware among AI firms like OpenAI and Anthropic. While Apple previously exited the server market due to low demand, the current AI-driven landscape has prompted a strategic re-evaluation. However, the project remains in early stages and faces potential cancellation before reaching the market.