NVIDIA and Meta announced plans to expand their collaboration on the Muse model family. Their planned optimizations aim to improve performance during training and inference, or running trained models—with a focus on AI agents.
Optimizing workflows and computing infrastructure
NVIDIA says the companies intend to optimize workflows together to maximize performance and reduce the total cost of ownership of infrastructure for Muse.
Meta plans to expand its deployment of NVIDIA infrastructure for the next generation of AI agents. That expansion is set to include Vera Rubin and Vera CPUs, with the goal of maximizing performance per watt. Meta also plans to deploy NVIDIA Spectrum-X Ethernet networking and BlueField-4 to support network efficiency and throughput.
Potential collaboration on OpenShell and agent monitoring
NVIDIA OpenShell Reference Architecture is another potential area of collaboration. The companies will explore opportunities involving NVIDIA OpenShell for safe agent deployment and NVIDIA Sentry for monitoring agent behavior.
NVIDIA also plans an equity investment in Meta
Alongside the technology collaboration, NVIDIA plans to invest $3 billion in Meta’s equity offering.
