Article
AI News AI Platforms

Nvidia's $13bn Hugging Face deal and the concept of reverse diversification

The chipmaker's move for the AI repository is regarded by analysts as insurance against rivals building its own hardware. Here, we explain why

by Ian Lyall
The image features wooden blocks arranged on a desk, prominently displaying the letters 'M & A' at the center, symbolizing mergers and acquisitions. Flanking the central blocks are icons representing companies and teamwork, underscoring the collaborative nature of business takeovers.

Nvidia is reportedly close to acquiring Hugging Face, the leading repository for open source and open weight artificial intelligence models, for close to $13 billion dollars.

The deal has been rumoured for some time before this reported confirmation.

Hugging Face is not a household name, but it is used constantly across the AI industry as a hub for sharing and downloading models.

Analysts are comparing the acquisition to Microsoft's purchase of GitHub, another developer platform that became strategically valuable well beyond its public profile.

Reverse diversification

The logic is described in some quarters as "reverse diversification", with Nvidia moving into software and platforms just as AI labs such as OpenAI and Anthropic explore building their own chips.

By controlling the primary platform where open models are hosted, Nvidia aims to protect long-term demand for its hardware.

The company has a track record of investing in smaller AI startups that later grow significantly, and this deal fits that pattern at a much larger scale.

The longer-term implications for OpenAI and Google remain unclear, since the impact of platform acquisitions often depends on decisions the acquiring company has not yet made.

by Ian Lyall