Meta gets $3 billion from Nvidia and plans to spend more on Nvidia chips

Nvidia plans to put $3 billion into Meta's planned equity offering, according to reporting by the Wall Street Journal, and the two companies said in the same announcement that they will widen an existing partnership around Meta's AI product Muse. Meta intends to move more of its next generation of AI agents onto Nvidia infrastructure, naming Vera Rubin systems and Vera CPUs as part of that plan.
The investment sits inside a much larger capital raise. Meta is seeking up to $95 billion through stock offerings to fund its AI infrastructure buildout, and Nvidia's $3 billion is a slice of that. CNBC reports that AMD and Broadcom have also signaled nonbinding interest in buying up to $5 billion in Meta shares combined, a figure that covers both companies rather than each one separately.
The structure is one investors have debated for months. Nvidia funds customers such as OpenAI, which then spend on data centers filled with Nvidia chips. Critics describe this as circular financing and argue it can obscure how much genuine end demand exists for the hardware. The Meta deal follows that shape: money from the chip supplier flows to a buyer that has publicly committed to buying more of its systems.
For developers, the practical signal is where Meta intends to run agent workloads. Vera Rubin is Nvidia's next-generation data-center platform, and Vera is the CPU line paired with it; the naming points to Meta standardizing more agent infrastructure on Nvidia rather than spreading it across vendors. Meta's mention of Muse ties the compute commitment to a specific product line, though the announcement gives no timelines, dollar figures for the hardware purchases, or performance targets.
The nonbinding nature of the AMD and Broadcom interest is worth noting: it signals appetite without a committed purchase, so the final mix of Meta's raise and its supplier relationships could look different from this announcement. Nvidia's $3 billion, by contrast, is described as part of the offering itself.
Why it matters: Teams building on Meta's AI products or agents should expect more of that work to sit on Nvidia hardware, which shapes tooling, CUDA-based workflows and capacity planning. The circular-financing critique also means procurement watchers will read this deal as much as a demand signal as a capital one; in my reading, that ambiguity is likely to persist until Meta discloses actual infrastructure spending and the offering closes.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
ia is buying into its customer Meta, which plans to run more of its AI agents on Nvidia hardware. The article Meta gets $3 billion from Nvidia and plans to spend more on Nvidia chips appeared first on The Decoder .