Nvidia announced on October 27 that it plans to invest $3 billion in Meta’s equity offering as the companies expand their work on the Muse artificial intelligence agent and its models. The announcement from Santa Clara, California, links new funding with efforts to improve model training and inference, the process of running trained models.
The agreement follows Meta’s financing plans announced Monday. Yahoo Finance reported that the company outlined a fundraising programme of up to $95 billion to support infrastructure and other AI initiatives. Nvidia’s commitment forms part of that wider financing effort, rather than a separate acquisition of the social media company.
According to Nvidia’s announcement, Meta intends to deploy more Vera Rubin hardware and Vera central processors. It also plans to use Spectrum-X Ethernet networking and BlueField-4 to improve network efficiency and throughput. The companies aim to improve performance and reduce the overall cost of operating Muse workloads.
They will also consider cooperation on Nvidia’s OpenShell architecture for secure agent deployment and Sentry for monitoring agent behaviour. This security work remains a possible extension of the partnership; the announcement does not describe it as an already completed integration.
Muse was introduced in September. In its launch announcement, Meta described software that can send emails, book travel and work across connected services through a dedicated virtual machine. The company said users choose its access permissions and must approve sensitive actions such as sending an email or making a purchase.
Nvidia chief executive Jensen Huang said agents could change computing by carrying out tasks for users, while Mark Zuckerberg presented Muse as a way to reach billions of people. Those ambitions remain company objectives. Nvidia classified the planned investment and expected technological benefits as forward-looking statements subject to risks and applicable closing conditions.
Comments