Key Highlights

  • Apple introduced new chips, called M6 and M5 Ultra, for its Mac mini and Mac Studio desktop computers, built to run AI programs much faster directly on the device.
  • The new Mac mini starts at $899, a $100 price increase from before, while the most powerful new Mac Studio starts at $5,499.
  • Apple says the new chips can run AI language models up to 9.8 times faster than older models, letting people work with huge AI models without needing the internet or paying for cloud computing.
  • Both new computers are available to order now and will ship to customers starting September 22.

Apple Bets Big on Desktop Computers for Running AI at Home

Apple (NASDAQ:AAPL) introduced new versions of its Mac mini and Mac Studio desktop computers, built around two new chips designed to make AI programs run much faster directly on the computer itself, rather than relying on the internet or a distant data center. The announcement comes just weeks before Apple is expected to unveil new iPhones, and it highlights how important these desktop computers have become to Apple as more people use them specifically to build and run their own AI programs.

Many developers who build AI "agents," programs that can complete tasks on their own like writing code or organizing files, prefer running them on a dedicated desktop computer like the Mac mini rather than paying for cloud servers. Similarly, researchers who want to experiment with Training and running their own AI models often turn to the more powerful Mac Studio. Apple said its computer chips include special built-in parts specifically for running AI, along with a memory system designed to avoid the slowdowns that typically happen when a computer needs to constantly access data from elsewhere.

Mac mini Gets Its First-Ever 2-Nanometer Chip

The Mac mini can now be configured with two chip options: the all-new M6, or the more powerful M5 Pro. This marks the first time an Apple computer has used a chip built on a 2-nanometer Manufacturing process, a more advanced and efficient way of packing components onto a chip. Apple says the M6 chip delivers up to 4 times faster AI performance and 2 times faster graphics compared to the previous Mac mini chip.

For users who need even more power, the Mac mini can be configured with the M5 Pro chip instead, which Apple says processes AI language model requests 8.5 times faster than the Mac mini model it replaces. However, prices have also increased. The Mac mini now starts at $899, up $100 from its previous starting price, following an earlier price increase over the summer that Apple attributed to rising memory costs.

Mac Studio Becomes Apple's Most Powerful Computer Yet

The Mac Studio, Apple's most powerful computer without a built-in screen, received an even bigger upgrade. It can now be configured with the M5 Max chip, which Apple says runs AI models nearly 4 times faster than the previous generation, or the brand-new M5 Ultra chip, described by Apple as the most powerful chip it has ever made.

The M5 Ultra is built differently than any previous Apple chip, combining four separate chip dies into a single unified processor for the first time. Apple says this chip can run AI language models up to 9.8 times faster than older Mac Studio models. Multiple Mac Studio computers with the M5 Ultra chip can even be linked together to work as one machine, combining their memory to run extremely large AI models that would normally require powerful cloud servers.

Mac Studio prices with the M5 Max chip remain unchanged at $2,499. However, the top-tier Mac Studio with the M5 Ultra chip now starts at $5,499, higher than the previous starting price of $5,299, and it also comes with a newer generation chip than before.

Conclusion

Apple's decision to upgrade its desktop lineup weeks ahead of its typical iPhone announcement signals how seriously the company is treating the shift toward AI programs that run directly on personal devices rather than in the cloud. The steep price increases, following an earlier hike blamed on memory costs, suggest AI-driven Demand for computer memory continues to squeeze margins across the industry. Whether developers and researchers embrace these new machines at their higher price points may depend on how much value they place on running AI privately and locally versus paying ongoing Cloud Computing costs.