Nvidia and AMD to invest $4 billion in meta to secure compute architecture and fuel $97.5 billion AI buildout
Nvidia and Advanced Micro Devices (AMD) are investing a combined $4 billion in Meta as the Facebook and Instagram owner accelerates its artificial intelligence ambitions. Nvidia has committed $3 billion, while AMD will contribute another $1 billion through Meta's latest equity offering. The investments make the two leading AI chip manufacturers shareholders in a company that is also one of their biggest customers.
The announcements come as Meta seeks additional funding to support its rapidly expanding AI infrastructure, including data centres, advanced processors and systems designed to power its next generation of AI assistants.
Meta's share offering includes $25 billion in ordinary shares and a separate $12.5 billion preferred stock offering. The company is also considering additional fundraising arrangements.
What each company gets out of the deal
For Nvidia, the investment comes alongside an expanded technology partnership with Meta. Meta plans to deploy more of Nvidia's computing infrastructure, including its latest Vera Rubin AI systems, Vera central processors, Spectrum-X networking equipment, and BlueField data processing units.
The two companies will also collaborate on Meta's Muse family of AI models, working to improve how they are trained and operated using Nvidia's hardware.
Muse is particularly important to Meta's ambitions to build more advanced AI assistants and autonomous agents capable of carrying out tasks on behalf of users.
AMD is pursuing a similar strategy, positioning its processors as an alternative to Nvidia's dominant AI hardware. The company confirmed that its first Helios rack-scale systems have been installed at Meta and are undergoing testing ahead of a wider deployment.
Helios is AMD's complete computing system, combining multiple processors, networking equipment and other components into server racks designed specifically for demanding AI workloads.
Meta also uses AMD's EPYC server processors and Instinct AI chips for training models and powering recommendation systems across its platforms.
The two companies previously agreed on a potential deployment of up to six gigawatts of AMD AI computing capacity.
A growing trend in the AI industry
The investments highlight an increasingly common arrangement in the technology industry, where companies supplying AI infrastructure are also investing in the businesses purchasing it. Nvidia and AMD have strong incentives to ensure that major customers such as Meta continue expanding their AI operations.
For Meta, bringing its suppliers closer financially could strengthen long-term relationships while helping secure access to the computing infrastructure needed to compete with Google, Microsoft and OpenAI.
However, the arrangement also creates an interesting financial relationship. If Meta's AI ambitions struggle, Nvidia and AMD could face losses both through reduced chip demand and declining value in their investments.
