Apple is dropping Mac mini and Studio early after OpenAI bought tens of thousands for training
OpenAI bought tens of thousands of Macs for reinforcement learning, Aaron Tilley at The Information reports, via Techmeme. Anthropic rents them. Nvidia sees Apple as its main local-AI rival as Macs gain traction with AI developers.
Hartley Charlton at MacRumors, citing The Information, says Apple’s Mac mini and Mac Studio announcement this week is early, driven by unexpectedly strong enterprise appetite for AI hardware. Apple normally waits until autumn, closer to October or November.
A Mac mini is the small desktop Apple sells for a home office. Reinforcement learning is how labs train a model by letting it try, get scored, and try again. That work usually lives on rented data-center chips.
The labs are stacking those minis and using them as the computers that run the training.
Apple sold the box. It did not staff a team for the companies buying the boxes by the tens of thousands.
Tilley wrote on X, via Techmeme, that Apple never expected the Mac mini to run AI infrastructure, but neoclouds like Mount Thor are now being built exclusively on Apple hardware. A neocloud, here, is a company that rents other people’s computers to run AI. Mount Thor is doing that on Apple machines.
Apple promoted the ability to link multiple Mac Studios into a single system for running large frontier models, a feature aimed at business and developer customers. June’s “Business at the Park” event had executives from Ford, Disney, and Anthropic. The Mac mini was the “darling” of the event.
The broader rush still took Apple by surprise. The company reportedly did not have an engineering team dedicated to business customers or staff focused on developer relations, and lacked an enterprise AI strategy.
Private Cloud Compute is Apple’s own private server layer for running some Apple Intelligence work off the phone. Businesses that asked to buy access were turned down. Apple is leaning on partners such as WebAI and Mount Thor, which provide AI tools and execution environments built on Apple hardware.
Demand hit the global memory shortage. Many mini and Studio configurations have been out of stock for months. Some enterprise customers are turning to Nvidia’s DGX Spark, a compact AI desktop launched late last year in a form factor similar to the Mac mini, as Apple’s high-end configurations stay hard to get.
The labs treated the mini as a rack. Apple treated it as a desktop.
Private Cloud Compute stayed private.
Pingback: Infostealers stole Claude sessions. SB Energy handed OpenAI $5.5 billion in warrants.