Apple has officially approved drivers that allow Nvidia and AMD external graphics processing units (GPUs) to work with Mac computers running Apple silicon chips, making it possible for users to boost their AI computing power without buying a new machine. The approval marks a significant shift in how Mac users can expand their hardware capabilities for artificial intelligence tasks like running large language models (LLMs), which are AI systems trained on vast amounts of text data to understand and generate human language.

Why Is This Such a Big Deal for Mac Users?

Until now, connecting external GPUs to Apple silicon Macs required workarounds that involved disabling critical security features. Tiny Corp, the company behind the tinybox AI accelerator, announced the approval in April 2026, noting that the installation process is now straightforward enough that "a Qwen could do it," referencing the popular open-source AI model . This development removes the friction that previously prevented mainstream adoption of external GPU setups on Macs.

The timing is particularly important because high-end Apple computers have become increasingly popular for AI work. The surge in demand has created real shortages, with delivery windows for Macs with large amounts of unified memory (a feature that allows the processor and graphics to share the same high-speed memory pool) stretching from six days to six weeks . Apple has even discontinued the 512GB unified memory option for the Mac Studio and raised prices on the 256GB model by $400, reflecting the intense demand from AI developers and researchers .

What External GPU Options Are Actually Available Right Now?

Tiny Corp currently offers two main products that can now work seamlessly with Apple silicon Macs. The company sells the red v2, powered by four AMD 9070XT GPUs and priced at $12,000, and the green v2 Blackwell, which features four RTX Pro 6000 Blackwell GPUs and costs $65,000 . These are high-end solutions designed specifically for serious AI work, not casual computing tasks.

It is important to note that this custom driver was developed by Tiny Corp itself, not by the GPU manufacturers. This means the driver is optimized specifically for running AI LLMs and inference tasks, not for gaming or graphics-intensive entertainment applications . For people hoping to use the same setup for both work and play, this remains a limitation worth considering.

How to Set Up an External GPU on Your Apple Silicon Mac

• Check Compatibility: Verify that your Mac has a Thunderbolt or USB4 port, which is required to connect external GPU enclosures to your computer.
• Install the Approved Driver: Download and install the officially approved driver from Tiny Corp, which now works without requiring you to disable System Integrity Protection or other security features.
• Connect Your External GPU: Attach your compatible Nvidia or AMD external GPU enclosure to your Mac using the Thunderbolt or USB4 connection, then configure it for AI workloads through the software interface.

The approval of these drivers represents a practical solution for Mac users who want to expand their AI computing capabilities without purchasing entirely new hardware. Rather than buying a dedicated AI supercomputer like Tiny Box, users can now pair their existing Mac with an external GPU setup to handle training or inference tasks with some limitations .

Looking ahead, Tiny Corp is planning to launch the exabox in 2027, which will feature 720 RDNA5 AT0 XL GPUs and deliver approximately 1 exaflop of computing power, roughly equivalent to one quintillion calculations per second, for around $10 million . This trajectory shows how the external GPU ecosystem for professional computing continues to evolve.

For Mac users working in artificial intelligence, this approval removes a significant barrier to entry. The combination of Apple's unified memory architecture and now-supported external GPUs creates a flexible pathway for AI development that didn't exist just months ago.