If you have followed the AI landscape recently, you may be aware that Apple’s Mac mini and Mac Studio have become the go-to devices to run AI models locally. But as per reports, the demand for the Mac devices is not limited to local use. Instead, AI labs including OpenAI and Anthropic are now increasingly using Mac mini and Mac Studio devices to train AI models as they try to get more computing power.
As per a report from The Information, OpenAI has bought tens of thousands of Mac mini and Mac Studio systems, while Anthropic is renting Mac minis through Amazon Web Services (AWS). These moves point to a sharp rise in demand for compute and an unexpected role for Apple’s desktop hardware in AI development.
Do note that Apple recently released new Mac mini and Mac Studio devices. The Mac mini now gets the new M6 chip, while the Mac Studio comes with M5 Max or M5 Ultra chips.
OpenAI and Anthropic using Mac minis to train AI
According to the report, OpenAI is using the Macs for reinforcement learning (RL), a technique in which AI learns through trial and error, and for training computer-use AI agents. These AI agents can navigate software interfaces, carry out multi-step workflows and interact with computers in a human-like way.
Going by the report, OpenAI is trying hard to secure more of the devices, in addition to the tens of thousands it has purchased already. On the other hand, Anthropic is renting Mac minis through AWS, but is using the devices for similar work.
To give you some context, in recent months, developers have been turning to Macs for specialised AI workloads rather than relying only on large Nvidia GPU clusters.
Why are AI companies buying Mac minis?
Apple’s Unified Memory Architecture and the thermal design of the Mac mini and Mac Studio are considered quite useful for on-device AI work, allowing multiple local sessions as well as large models and desktop tasks to run on the same system without the usual latency and cost overhead.
The surge in interest has also fed into Apple’s hardware business. As per reports, Mac revenue rose nearly 29 per cent year on year to $10.3 billion in the latest quarter, making it Apple’s fastest-growing hardware line.
The unexpected enterprise demand is also said to have influenced unusually early refreshes for the Mac mini and Mac Studio. Apple typically unveils new Mac models in the autumn, closer to October or November, but the M6 Mac mini and the new M5 Max and Ultra chips for the Mac Studio were announced in late August instead.
Even so, the report said the broader enterprise rush towards powerful desktop Macs caught Apple off guard. The company reportedly did not have an engineering team dedicated to business customers, staff focused on developer relations, or an enterprise AI strategy. The rise in demand for Mac devices has come amidst a global memory shortage, leaving many high-end Mac mini and Mac Studio configurations out of stock for months.
As per reports, Nvidia, the maker of the most advanced GPU clusters for training AI models, now sees Apple as its biggest competitor in local AI and launched the DGX Spark in response.
– Ends
Leave a Reply