Nvidia Commits up to $100 Billion in Partnership With OpenAI to Build Massive AI Infrastructure

Rabat – Nvidia will invest as much as $100 billion in OpenAI to develop one of the largest computing infrastructures ever attempted, marking a major step in the race to power next-generation artificial intelligence models.

The partnership aims to deliver up to 10 gigawatts of advanced computing capacity. The first phase is expected to begin in the second half of 2026.

“Everything starts with compute,” OpenAI CEO Sam Altman said in a statement. “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.”

The agreement places Nvidia not only as a hardware supplier but also as a general strategic partner in shaping OpenAI’s growth.

The deal highlights how central computing power has become in the AI industry and how costly it is to build the systems that train and run large-scale models.

The scale of the investment

The partnership sets a target of constructing data centers that will be capable of supporting 10 gigawatts of computing power. To put it in perspective, a single gigawatt can supply electricity to hundreds of thousands of homes.

These facilities will be filled with Nvidia’s latest chips and networking technology, designed to handle the immense processing demands of AI systems.

Nvidia will provide funding in non-voting shares, meaning it does not gain direct control over OpenAI’s governance.

Instead, the financial commitment ensures OpenAI has the resources to build and expand infrastructure without facing immediate cash constraints. This arrangement also allows Nvidia to benefit from strong demand for its processors, as OpenAI is expected to purchase significant quantities of its hardware.

The first major deployment will use Nvidia’s upcoming “Vera Rubin” platform, with initial systems planned to go live in 2026. Each subsequent phase will be rolled out as part of the long-term strategy to reach the full 10-gigawatt capacity.

Strategic importance for both companies

For OpenAI, the agreement secures a steady pipeline of computing resources essential for training larger and more sophisticated models. The scale of these models has grown rapidly, and without massive infrastructure, continued progress would be limited.

By aligning with Nvidia, OpenAI positions itself to sustain growth in an increasingly competitive sector.

Nvidia, on the other hand, strengthens its role as the backbone of AI infrastructure worldwide. Its chips already dominate the industry, and this deal ensures that one of the world’s most prominent AI companies will remain a committed customer.

“For Nvidia, the prize is huge — every gigawatt of AI data centre capacity is worth about $50 billion in revenue, meaning this project could be worth as much as $500 billion,” said Matt Britzman, a Senior Equity Analyst at Hargreaves Lansdown, MarketScreener UK.

Broader implications and challenges

The size of the investment raises questions about energy use, land, and the sustainability of operating data centers on this scale.

Building facilities capable of handling 10 gigawatts will require careful planning to manage electricity consumption and cooling systems. Environmental and regulatory reviews are expected to play a major role in how quickly the project can advance.

According to Decrypt, deploying 10 gigawatts of computing won’t be easy. Cooling alone can use nearly 40% of a data center’s power.

Deloitte warns AI growth could push global electricity use from data centers from 536 to over 1,000 terawatt-hours by 2030. Rising water use for cooling and strain on electric grids add to the environmental concerns.