Nvidia Invests Over $40B in AI Firms, Including $30B in OpenAI

Nvidia has invested more than $40 billion in AI and technology companies so far in 2026, surpassing its total investment activity for 2025.

The largest single equity investment is an approximately $30 billion stake in OpenAI. That deal is the biggest equity purchase in Nvidia’s history.

The remainder of the capital is allocated across startups and firms focused on AI infrastructure, including a financing and infrastructure agreement with IREN. The IREN arrangement includes a $2.1 billion equity warrant and a $3.4 billion AI compute infrastructure contract.

IREN previously operated Bitcoin-mining facilities. Those sites already have high-power capacity, cooling systems and round-the-clock operations, features that can be converted into GPU-focused data centers.

Nvidia links many of its equity stakes to expected future demand for its accelerators. OpenAI and several portfolio companies operate or plan large GPU clusters for model training and serving.

Company executives and market analysts describe the strategy as financing customers that will buy Nvidia hardware. Many of the deals are tied directly to firms building or expanding data-center capacity for model training and inference.

Some observers compare the structure to vendor financing seen in the late 1990s, when equipment makers provided capital to customers that then purchased their products. Proponents point to reported revenue at companies such as OpenAI and to growing demand for AI compute.

Analysts warn that heavy equity exposure to customers could create financial links that amplify losses if AI spending declines. In that scenario, Nvidia could face pressure on chip sales and on the value of its equity holdings at the same time.

The investments give Nvidia a role in deciding which firms receive capital and where new AI compute capacity is built. The company’s spending focuses on infrastructure that will consume large amounts of Nvidia hardware.