'Today is the day you've been waiting for': eGPUs can now officially turn a humble Mac Mini into an AI powerhouse
External GPUs now turbocharge Mac Minis for AI workloads
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
You are now subscribed
Your newsletter sign-up was successful
Join the club
Get full access to premium articles, exclusive features and a growing list of member rewards.
- TinyGPU enables external GPUs to run AI models directly on Macs
- Users can now leverage AMD and Nvidia cards without bypassing protections
- TinyGPU supports macOS 12.1 and later with Thunderbolt or USB4
Apple Silicon Macs have long been praised for their efficiency and tight hardware-software integration, but they have not been ideal for external GPU use.
That changed when Apple officially approved TinyGPU, a driver enabling eGPUs to function as AI accelerators on Apple Silicon machines.
The approval means that users can now leverage high-performance AMD and Nvidia cards for AI tasks without needing to bypass system protections like SIP.
Article continues belowOfficial Approval Unlocks AI Potential for Mac Users
“If you have a Thunderbolt or USB4 eGPU and a Mac, today is the day you've been waiting for! Apple finally approved our driver for both AMD and NVIDIA,” TinyCorp said on X.
TinyGPU’s functionality focuses exclusively on AI workloads rather than traditional graphics output, allowing complex models to run directly on a Mac Mini or other supported Apple Silicon devices.
The TinyGPU driver supports macOS 12.1 or later, along with devices equipped with USB4 or Thunderbolt 3 and 4 ports.
AMD GPUs from the RDNA3 generation onward and Nvidia cards from the Ampere series are compatible.
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Users can run AMD workloads natively, while Nvidia GPUs require a Docker Desktop setup to execute AI computations through NVCC.
Once the driver is installed and approved in the system, even demanding AI models such as Qwen 2.5 27B can run effectively.
TinyGPU’s framework, tinygrad, provides the necessary computational interface, offering a streamlined approach for AI acceleration that had previously been impossible.
The timing of this development coincides with the permanent discontinuation of the Mac Pro.
Apple has removed the Mac Pro from its website, redirecting the flagship model’s page to the general Mac homepage.
Reports indicate that no new Mac Pro hardware is planned, effectively ending a line that had only seen three updates over the past fourteen years.
Rumors of an M4 Ultra variant never materialized, suggesting limited sales and a perceived lack of demand for high-end desktop configurations.
For many, this shift leaves Apple Silicon Macs without a modular workstation alternative — until now, eGPUs provide a viable route for users who require substantial AI performance.
The combination of TinyGPU with a Mac Mini or other Apple Silicon systems opens new possibilities for professionals and enthusiasts alike.
The ability to connect external GPUs and run AI tools directly on macOS introduces a level of flexibility.
As TinyCorp noted, “It’s so easy to install now a Qwen could do it, then it can run that Qwen,” reflecting the driver’s accessibility and impact.
While Apple has moved away from traditional workstation hardware, eGPU support now lets users turn modest devices into AI-capable systems.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.

Efosa has been writing about technology for over 7 years, initially driven by curiosity but now fueled by a strong passion for the field. He holds both a Master's and a PhD in sciences, which provided him with a solid foundation in analytical thinking.
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.