Timeline

NVIDIA signs a letter of intent to invest up to $100 billion in OpenAI

Tied to ten gigawatts of deployment starting in late 2026, with OpenAI paying Nvidia in cash for chips while Nvidia takes a non-controlling equity stake.

  • Money & business
  • Compute & infrastructure
  • Major

NVIDIA and OpenAI signed a letter of intent for a strategic partnership under which NVIDIA would deploy at least 10 gigawatts of its systems for OpenAI’s next-generation infrastructure, and would invest up to $100 billion in OpenAI progressively as each gigawatt came online. The companies said the first gigawatt, built on NVIDIA’s Vera Rubin platform, was targeted for the second half of 2026.

The structure drew immediate scrutiny. OpenAI would pay NVIDIA in cash for the hardware, while NVIDIA’s investment returned to OpenAI as non-controlling equity — meaning a large share of NVIDIA’s revenue growth would be financed by NVIDIA’s own capital going back into its largest customer. Commentators and some investors described this as a pattern of circular financing running through the AI supply chain, in which chipmakers, cloud providers and model developers increasingly funded one another’s purchases rather than transacting with independent capital. The letter of intent was explicitly non-binding, with details to be finalised in the weeks that followed, and the scale involved — description of the buildout as equivalent to four to five million GPUs — was itself unverified beyond the companies’ own framing.

The announcement came three weeks before OpenAI struck comparable multi-gigawatt compute deals with AMD and Broadcom, part of a run of announcements that collectively committed OpenAI to tens of gigawatts of future capacity without disclosed financing for the underlying capital expenditure. Together with NVIDIA’s own rise to a $5 trillion market capitalisation five weeks later, the deal became a reference point in a wider debate over whether AI infrastructure spending was outrunning demonstrated revenue.

Referenced by