NVIDIA Invests $3 Billion in Meta as the Two Companies Deepen Their Muse Partnership
NVIDIA and Meta announced an expanded partnership on Muse and the Muse model family on Tuesday, pairing joint work on inference and training performance with…
By Icarus3 min read
The joint work covers inference and training performance for the agent, which runs on NVIDIA infrastructure.
NVIDIA and Meta announced on Tuesday that they will expand their partnership on Muse and the Muse family of models, pairing joint work on inference and training performance with a $3 billion NVIDIA investment in Meta’s equity offering. Reuters reported that Meta raised the offering to $97.5 billion earlier in the day, the largest fundraising in the company’s history, as it accelerates spending on AI infrastructure.
The two companies said the expanded work will focus on optimizations across workflows to maximize performance and improve total cost of ownership for Muse’s models and agents, which run on NVIDIA infrastructure. Meta also said it will expand its deployment of NVIDIA’s Vera Rubin and Vera CPUs for next-generation AI agents, and deploy NVIDIA Spectrum-X Ethernet networking and BlueField-4 to support network efficiency and throughput.
OpenShell and Sentry Under Consideration
The companies also said they will consider additional collaboration on NVIDIA’s OpenShell Reference Architecture, including OpenShell for safe deployment of agents and Sentry for monitoring agent behavior. Neither item is a committed product or launch; the announcement frames both as opportunities under consideration.
NVIDIA chief executive Jensen Huang said the partnership aims to optimize computing infrastructure end to end for capable, efficient, and secure AI agents. “AI agents represent a new computing paradigm — software that understands our goals, reasons, uses tools and gets work done on our behalf,” Huang said. “NVIDIA and Meta are expanding our partnership to optimize the computing infrastructure end to end for capable, efficient and secure AI agents, so Muse can empower billions of people to achieve more and pursue greater ambitions.”
Meta chief executive Mark Zuckerberg said Muse is the company’s route to personal superintelligence for everyone. “Muse is how we bring personal superintelligence to everyone. We’re excited to keep working with NVIDIA to deliver the best personal agents and frontier models for billions of people,” Zuckerberg said.
Why the Partnership Lands Now
The announcement follows a month of rapid growth for Muse. Meta said on its third-quarter earnings call on October 21 that Muse has passed 10 million users, with shopping emerging as a leading activity and about a quarter of users who set up the agent now using it daily. NVIDIA’s announcement points to that demand, citing frontier-level benchmark performance and surging consumer downloads since Muse launched on September 8.
The infrastructure behind that growth is already NVIDIA-based. Meta’s Prometheus supercluster in New Albany, Ohio, which trained the Muse models, runs on NVIDIA GPU platforms, and Meta has said a third platform based on NVIDIA’s Vera Rubin NVL72 is on the way. The expanded partnership extends that relationship from hardware supply into joint optimization of the full stack behind the agent.
The investment itself is part of Meta’s equity offering rather than a standalone purchase of Muse assets, and NVIDIA’s announcement notes that the investment remains subject to closing conditions. What is concrete today is the direction: Meta is raising the largest sum in its history to build AI infrastructure, and the company building much of that infrastructure is now also a $3 billion investor in the offering funding it.
