Apple Approves Nvidia eGPU Drivers for M-Series Macs

Apple has officially approved drivers developed by Tiny Corp that enable AMD and Nvidia external graphics processing units (eGPUs) to function with Macs powered by M-series processors [1]. This authorization marks a significant shift for the Apple Silicon ecosystem, which previously lacked official support for such external hardware [2]. The new software is designed primarily to accelerate artificial intelligence tasks rather than improve gaming performance [1].

Key takeaways

• Apple approved Tiny Corp drivers for AMD and Nvidia eGPUs on M-series Macs [1].
• The software is optimized for AI workloads and Large Language Models, not gaming [2].
• Users no longer need to disable System Integrity Protection (SIP) to install the driver [3].
• Connectivity requires Macs equipped with Thunderbolt or USB4 ports [2].

A New Focus on Artificial Intelligence

While eGPUs have historically been used to boost gaming graphics, Tiny Corp explicitly designed these drivers to handle AI-based tasks, specifically running and training Large Language Models (LLMs) [1][2]. The company claims the installation process is now simple enough that "a Qwen could do it," though sources note the setup still requires compiling the driver via Docker rather than offering a traditional plug-and-play experience [1][3]. By signing the drivers, Apple has eliminated the previous requirement for users to disable System Integrity Protection (SIP), a core system security protocol, thereby creating a more stable environment for researchers [2][3].

Hardware Connectivity and Constraints

To utilize the external processing power, users must connect Thunderbolt or USB4 eGPUs to Mac models that possess these specific high-bandwidth ports [1][2]. This means compatibility is restricted to newer or higher-end devices, as older models with restricted connectivity cannot utilize the feature [2]. Despite the breakthrough in compute capability, current limitations exist regarding display output; the drivers do not currently support acceleration for direct video output to external monitors [2]. This development does not indicate a direct partnership between Apple and Nvidia, but rather an authorization of Tiny Corp’s independent software solution [3].

Why it matters

This approval addresses the growing demand for high-performance computing within the macOS environment as AI research intensifies [1]. By allowing researchers to leverage powerful external Nvidia and AMD hardware, Apple is providing a pathway for M-series Mac users to perform complex data processing and machine learning workflows that the integrated silicon might not handle as efficiently [1][2]. The move comes as high-end Macs with large unified memory have faced shortages due to increased demand from AI developers, suggesting a strategic effort to retain this professional demographic [4].