Why is Nvidia investing $100 bn in OpenAI?

Tech giant NVIDIA will invest $100 billion in AI startup OpenAI, the companies announced Monday. The companies said the partnership will involve deploying 10 GW of Nvidia systems powered by its upcoming Vera Rubin GPUs to meet OpenAI’s datacentre and power needs.

The first phase, with around 1 GW of capacity, is expected to be completed by the second half of 2026.

The partnership will enable OpenAI to scale its infrastructure to meet surging demand, as the company crosses 700 million weekly active users across enterprises, small businesses, and developers.

“Compute infrastructure will be the basis for the economy of the future, and we will utilize what we’re building with NVIDIA to both create new AI breakthroughs and empower people and businesses with them at scale,” said Sam Altman, co-founder and CEO, OpenAI, in a statement.

These million-GPU AI factories that OpenAI will be able to build with Nvidia, will support both its training and inference workloads for the next generation of agentic and reasoning models, central to its mission of developing artificial general intelligence.

What does Nvidia get out of this partnership?

The two companies share almost a decade-old partnership, beginning in 2016 when Nvidia founder and CEO Jensen Huang ‘hand-delivered’ the company’s first DGX supercomputer to OpenAI. That system went on to power the early research behind what later became ChatGPT.

“This is a billion times more computational power than that initial server,” said Greg Brockman, President, OpenAI.

The new agreement builds on this long-standing collaboration, expanding their work together into the next generation of AI infrastructure.

“NVIDIA and OpenAI have pushed each other for a decade, from the first DGX supercomputer to the breakthrough of ChatGPT. This investment and infrastructure partnership mark the next leap forward—deploying 10 gigawatts to power the next era of intelligence,” Huang said in a statement.

OpenAI will collaborate with Nvidia as its preferred compute and networking partner, jointly optimising former’s models and infrastructure software alongside the latter’s hardware and software roadmaps.

Although the agreement is still being finalised, it gives Nvidia a significant financial stake in OpenAI, which was recently reported to be valued at around $500 billion.

Beyond the equity stake, the deal also secures Nvidia a guaranteed customer for its upcoming systems, reinforcing the company’s dominance in AI data centres. With OpenAI’s facilities positioned to be among the industry’s most advanced, Nvidia will further be able to cement its indispensability in the sector with this partnership.

While the company has not clarified whether this investment will power OpenAI’s Stargate initiative, notably, when OpenAI first announced the Stargate project in January, it had identified Nvidia as a key technology partner.

“As part of Stargate, Oracle, NVIDIA, and OpenAI will closely collaborate to build and operate this computing system,” OpenAI wrote in a blogpost in January this year.

In July, OpenAI had announced another datacentre partnership with Oracle to develop 4.5 GW of additional Stargate data centre capacity in the U.S. Stargate is OpenAI and SoftBank’s most expensive bet on building AI infrastructure in the U.S. in 4 years with an investment of $500 billion.