News
Public · Published
NVIDIA and Meta to Expand Collaboration on Muse and Muse Family of Models Amid Robust Growth
NVIDIA and Meta announced they will expand their partnership on Muse and the Muse family of models to improve AI training and inference performance. Meta will also expand deployment of NVIDIA Vera Rubin, Vera CPU, Spectrum-X Ethernet networking and BlueField-4.
Published:
Updated:
What happened
NVIDIA and Meta announced they will expand their partnership on Muse and the Muse family of models to improve AI training and inference performance. Meta will also expand deployment of NVIDIA Vera Rubin, Vera CPU, Spectrum-X Ethernet networking and BlueField-4.
Confirmed
Global impact / market context
Deeper work on Muse ties Meta's AI agents more closely to NVIDIA hardware, while larger Vera Rubin and networking deployments support Meta's computing buildout. That can lift NVIDIA data-center revenue and help Meta lower the total cost of running AI models.
Analyst inference
The deal reinforces the pattern of large AI buyers committing to NVIDIA's full stack of chips and networking, which matters for NVIDIA's sales pipeline and for rivals trying to win the same AI infrastructure spending. It also signals Meta is still increasing AI capacity.
Analyst inference
What to watch
- The companies will consider collaborating on NVIDIA OpenShell Reference Architecture, including OpenShell for safe deployment, plus NVIDIA Sentry for monitoring structured, unstructured and graph data processing. Confirmed
- Watch whether Meta confirms firm orders and timing for Vera Rubin, Vera CPU, Spectrum-X and BlueField-4, since expansion plans stated as intentions may shift before they become committed capital spending. Proposed
- Track Meta's AI capital spending and NVIDIA data-center revenue for signs the Muse partnership translates into real hardware demand, which would support NVIDIA sales and Meta's cost per AI workload. Analyst inference