Nvidia (NVDA) and Meta (META) have expanded their multiyear AI partnership, with Meta committing to buy millions of Nvidia’s Blackwell and Rubin GPUs, along with CPUs and networking systems, to power its next wave of AI models.
The chips will be deployed across Meta’s data centers for both training and running large-scale AI systems, including recommendation engines and advanced research models. Meta will also tap Nvidia’s Cloud Partner network, which includes firms like CoreWeave, to scale capacity.
Nvidia CEO Jensen Huang said the collaboration brings the “full Nvidia platform” to Meta, highlighting deep integration across hardware and software.
Meta is also rolling out Grace CPU-only servers, with next-generation Vera systems planned for 2027, potentially increasing pressure on traditional server chipmakers like Intel (INTC) and AMD (AMD).
The deal comes as investors debate whether heavy AI spending is sustainable. Despite concerns about custom chips from companies like Amazon (AMZN) and Alphabet (GOOG), analysts say Nvidia’s general-purpose GPUs remain dominant due to their flexibility across AI workloads.
Nvidia shares edged higher following the announcement, while Meta stock traded slightly lower.
Disclosure: This article does not represent investment advice. The content and materials featured on this page are for educational purposes only.
Related: Nvidia Jensen Huang: AI Buildout Has 7–8 Years to Run as Demand Stays “Sky High”
Nvidia’s Jensen Huang: AI Replacing Software Is “Most Illogical Thing in the World”
Nvidia’s $100B OpenAI Deal Was Never a Promise, Jensen Huang Says
