NVIDIA is using money to carve out an AI moat. In the first few months of this year, the chip giant has committed $40 billion to equity investments, all targeting key players in the AI ecosystem.


The biggest bet went to OpenAI — $30 billion. This sum will be used to support the latter’s expansion of large model training and inference capabilities. In addition, NVIDIA has made seven intensive multi-billion-dollar investments in listed companies, including $3.2 billion to glass manufacturer Corning and $2.1 billion to data center operator IREN.

NVIDIA is also active in the startup space. According to statistics, they participated in 67 funding rounds for AI startups in 2025, and have already invested in about 20 unlisted AI companies in the first few months of this year.

Many of these invested companies are also NVIDIA’s customers, which has sparked discussions about “circular transactions.” Critics argue that money flowing back and forth between NVIDIA and its customers may distort the market’s judgment of real demand.

But Wall Street has a different view. Matthew Bryson, an analyst at Wedbush Securities, believes that this investment strategy fully aligns with the circular investment logic. If these invested companies can gain a foothold in their respective fields, they will build a wider moat for NVIDIA.

NVIDIA is transitioning from a chip supplier to an AI infrastructure operator. The company expects the cumulative revenue of its Blackwell and Rubin series chips to reach at least $1 trillion by the end of 2027. Through equity investments, NVIDIA is trying to tie itself to the core players of the next-generation AI infrastructure.

This is not a simple financial investment, but a carefully calculated strategic layout. While everyone is scrambling for GPUs, NVIDIA is already laying out the entire chessboard.