Nvidia will put 3 billion dollars into Meta's equity raise

The chip seller is now a shareholder in its biggest customer. Nvidia announced on Monday that it will invest 3 billion dollars in Meta's equity offering. The announcement came from Nvidia's own newsroom. It landed on the same day Meta disclosed its plan to raise up to 95 billion dollars.

The two companies also said they will deepen work on Meta's Muse family of AI models and agents. Muse is Meta's personal AI assistant, launched in September 2026. It handles tasks across Meta's apps.

The hardware list

Meta plans to deploy more Nvidia technology to run its AI training and inference. Nvidia's announcement names four product lines. Vera Rubin systems and Vera CPUs are the main compute. Spectrum-X Ethernet handles networking between them. BlueField-4 handles the data path on each server. The two companies say the goal is better training and inference performance, with lower cost per unit of work.

Nvidia's statement names three outcomes it is working on with Meta. Faster training. Faster inference. A lower cost of running the models in production.

The quote

Nvidia chief executive Jensen Huang said in the announcement that AI agents are a new way to use software. He called them "a new computing paradigm" that "gets work done on our behalf". Nvidia says it wants Muse to be a leading product in that group.

Where the equity sits inside the raise

Meta's equity raise is up to 95 billion dollars in total. On Monday it opened a 35 billion dollar underwritten offering. The rest is an at-the-market programme. It starts when Meta's new non-voting Class C shares begin trading, in the first quarter of 2027. Business Recorder reported this is the largest equity raise Meta has done. Nvidia's 3 billion dollars is a confirmed piece of the Monday round.

The Information reported last week on the chip makers' interest. Three names were on that list: Nvidia, AMD and Broadcom. Together they told Meta they might buy up to 5 billion dollars of Class A shares. Monday's announcement is the first firm number. Nvidia is in for 3 billion.

Why this matters for a team buying AI

A chip seller funding its largest customer is a signal about the shape of AI demand. Meta's own cash flow turned negative last quarter. The money for the next round of data centres has to come from somewhere. Monday's filings show it is now coming from the capital markets. The vendor selling the chips is helping to pay for them.

That is a tie between the two companies' futures. If Meta needs more Nvidia hardware, Nvidia's investment helps pay for it. If Meta cuts spending, Nvidia loses on the equity and on the sale. For a team that buys AI compute, the practical read is simple. The sticker price on the chip is only part of the real cost. A share of that price is being pre-funded by the seller through deals like this one. The question is what happens to prices when that funding stops.

Source

Nvidia and Meta to expand collaboration on Muse and Muse family of models amid robust growth, Nvidia. Reporting by CNBC and The Information.