AI for Business

Nvidia in Advanced Talks for Major OpenAI Stake, Sources Say

Nvidia is negotiating a potential investment of up to $30 billion in OpenAI, according to a source with direct knowledge of the confidential talks. The deal, if finalized, would value the AI...

Share:

Nvidia is negotiating a potential investment of up to $30 billion in OpenAI, according to a source with direct knowledge of the confidential talks. The deal, if finalized, would value the AI leader at $730 billion before the new capital is added.

This proposed investment is distinct from the $100 billion infrastructure partnership the two companies announced last September, the source clarified. That earlier agreement, which triggered a wave of similar mega-deals across the industry, tied Nvidia's funding to the completion of specific supercomputing facilities. The $30 billion under discussion is not contingent on such milestones.

While not yet final, the negotiations have intensified in recent weeks. The terms remain fluid, the source noted, and Nvidia could still choose to structure future commitments within the original framework. When reached for comment, Nvidia declined to elaborate.

Speculation about the status of the broader $100 billion pact has grown since a January report suggested it had stalled. In a November financial filing, Nvidia itself stated there was "no assurance" definitive agreements would be reached.

Both CEOs have publicly downplayed tensions. OpenAI's Sam Altman recently posted on X that the company "loves working with Nvidia." Nvidia's Jensen Huang, in a CNBC interview this month, stated there was "no question" his firm would invest in OpenAI's next funding round.

OpenAI is concurrently speaking with other investors, including Amazon and Microsoft, with the total round potentially reaching $100 billion. The process may conclude in two phases, starting with strategic partners before expanding to a wider investor group.

Source: CNBC

Ready to Modernize Your Business?

Get your AI automation roadmap in minutes, not months.

Analyze Your Workflows →