Apple Trims Mac mini, Mac Studio Lineup as AI Demand Spikes

Apple has removed several high-memory configurations of its Mac mini and Mac Studio, the latest sign that the AI-fueled memory shortage is starting to bite.

Mac mini models featuring 32GB and 64GB of RAM, and Mac Studio models featuring 128GB and 256GB, were quietly removed from Apple’s US online store over the weekend. This comes a week after Apple removed the highest memory variant, 512GB, of the Mac Studio from its store. It also increased the price of the 256GB option by 25 percent last week.

Apple has not commented on the removal of these options, leaving the door open to speculation. If it’s not due to a memory shortage, there have been rumors of refreshed Mac mini and Mac Studio models potentially coming at WWDC 2026 in June, so Apple could simply be clearing inventory ahead of the release.

In these cases, the higher-memory variants are typically taken off the market first.

There has been a surge in interest in Apple’s Mac mini, with several AI developers and enthusiasts suggesting it is a cheap and powerful hub for an always-on AI agent. Perplexity recently launched Personal Computer, which lets an AI agent take over a computer, and said the Mac mini was the preferred device to run it.

The AI memory crunch

Apple is not the only tech manufacturer having to either increase prices or remove high-memory options.

• Dell and Lenovo have both raised prices in response to the RAM shortage
• HP has said that RAM now makes up to 35 percent of the PC build cost
• RAM suppliers have also reworked their operations, shifting focus from consumers to data centers
• Micron recently shut down its Crucial DRAM line, mainly used for gaming PCs
• Samsung has reallocated capacity from its consumer memory unit toward AI services

It does not look like this memory shortage will be short-lived either. SK Hynix, one of the largest suppliers of HBM, a specialized form of DRAM suitable for AI workloads, has announced plans to invest $13 billion in a new facility to meet demand from AI customers.

Some are exploring ways to reduce the memory burden, either through new memory types being developed by Intel and SoftBank or through new systems designed to lower AI memory requirements.

It is just the latest market to be squeezed by the huge demand for AI infrastructure buildout. Nvidia has been inundated with demand for its Blackwell chips since launch, with the chipmaker completely sold out of units for the next 12 months. Even though the memory crunch is severe, AI developers remain primarily focused on securing GPU capacity, and Nvidia supplies most of it.

It is not clear how far across the computing spectrum this crunch will spread. There is already strong demand for AI accelerators such as Google’s TPU and Amazon’s in-house chips. These server racks also contain CPUs, SSDs, and networking equipment. While these components are less critical than GPUs and memory, they could still face their own supply constraints if demand continues to outpace supply.

More Apple news: For more on Apple’s AI push, check out how its next-gen Siri chatbot could reshape iOS 27.