Nvidia invests in OpenAI with an AI infrastructure agreement of $ 100 billion

OpenAi and Nvidia have beaten One of the largest partnerships In AI, with Nvidia who promises to invest up to $ 100 billion in OpenAI and at the same time to deliver the computing power that is needed to build the next generation models of the company.

The deal, announced on Monday in a declaration of intent, calls open to use at least 10 Gigawatt Nvidia systems in the coming years. The first phase, one gigawatt, is planned for the second half of 2026 on the upcoming Vera Rubin platform of Nvidia, named after the late dark astronomer. The investment of Nvidia will grow in scale as every new system is used.

(Disclosure: Ziff Davis, CNET’s parent company, brought a lawsuit against Chatgpt Maker OpenAi in April, who claimed that it infringed Ziff Davis authentic rights in training and operating his AI systems.)

Miss none of our impartial technical content and laboratory -based reviews. Add Cnet Like a preferred Google source.

“Everything starts with calculation,” said OpenAi CEO Sam Altman in a statement. “Calculate infrastructure will form the basis for the economy of the future, and we will use what we build with Nvidia to create new AI through -breaches and to enable people and companies to scale.”

Also read: Is AI able to “plan?” What opened found when testing difficult behavior

This partnership is remarkable, because AI research is increasingly limited by access to massive computer sources. By protecting a long-term pipeline from NVIDIA hardware, OpenAI tries to guarantee the ability to keep pace with rivals such as Google, Anthropic, Microsoft and Meta.

For Nvidia, the deal makes it more than just a supplier. By gradually taking a great interest in OpenAi, it positions itself in the middle of the AI ​​tree and buys in the largest AI company.

Also read: OpenAI is building a teenage friendly version of Chatgpt

The completion of this deal will take years. If the partnership applies, this could determine how fast AI is progressing, which types of models opened in the future and how accessible those models will be for the world population will be.