Apple Reportedly Developing AI Servers Powered by M-series Ultra Chips

Apple is reportedly working on artificial intelligence (AI) servers that would leverage its high-performance M-series Ultra chips, similar to those found in Mac desktops. The potential product, expected to launch in 2029, would mark Apple's first enterprise server in nearly two decades, aiming to capitalize on the growing popularity of Apple hardware among AI developers.

This move comes as AI developers and companies are increasingly adopting Mac computers, particularly the Mac mini and Mac Studio, for running AI workloads.

Highlights

• Apple is reportedly developing an AI server that will utilize its M-series Ultra chips.
• The servers are expected to debut in 2029, potentially marking Apple's return to the enterprise server market after a long hiatus.
• The planned enterprise server configurations include either two or four of Apple's future M8 Ultra chips.
• AI companies like OpenAI and Anthropic have reportedly purchased or rented tens of thousands of Mac minis and Mac Studios for AI training.
• Apple is also considering integrating Nvidia's data center networking equipment and NVLink Fusion technology for connecting its M8 chips.

Details

According to The Information, the AI server project reportedly gained support from current Apple CEO John Ternus when it began a year ago, during his tenure leading Apple's hardware engineering efforts. The servers are expected to come in two configurations, featuring either two or four of Apple's upcoming M8 Ultra chips.

AI companies, including OpenAI, have reportedly acquired "tens of thousands" of Mac minis and Mac Studios to train AI agents through reinforcement learning. Anthropic has also rented Mac minis from Amazon Web Services. This surge in demand for M-series chip-powered devices has likely encouraged Apple's pursuit of an enterprise server using similar technology.

Apple is also said to be exploring the use of Nvidia's data center networking equipment and NVLink Fusion technology to connect its M8 chips. However, The Information cautioned that the project could still be canceled or proceed without Nvidia's technology.

Why it matters

If Apple's AI server project comes to fruition, it would signify a notable return to the enterprise server market for the company, following the discontinuation of its Xserve in 2011. This initiative could be a strategic move to meet the escalating demand in the AI sector and expand Apple's hardware ecosystem. However, the ongoing shortage of memory chips, driven by the rapid construction of AI data centers, presents a significant challenge for the project.