Nvidia plans to invest up to US$100 billion in OpenAI, with most of the funds expected to be spent on Nvidia’s own graphics processing units (GPUs), according to CNBC.

The investment will be deployed as new AI supercomputing facilities come online, starting with a site in Abilene, Texas, in the second half of 2026.

OpenAI, an AI startup based in the US, will lease Nvidia’s chips instead of buying them outright, allowing the company to spread costs over several years.

Nvidia’s CEO Jensen Huang estimated that a gigawatt-scale AI data center could cost US$50 billion, with about US$35 billion for GPUs.

The first US$10 billion from Nvidia will be available to OpenAI soon to support its data center expansion.

The OpenAI-Nvidia deal has raised some concerns about the sustainability of the AI boom.

🔗 Source: CNBC

Recent Nvidia developments