Nvidia Unveils 'Scale-In' Infrastructure Category to Secure Agentic AI Workloads

Nvidia is expanding its role in enterprise infrastructure by introducing 'scale-in,' a new category that connects more AI factory resources while extending security controls across its infrastructure.

The company's senior vice president of networking, Gilad Shainer, explained that scale-in brings users, models, agents, storage, and compute resources into the AI factory with security controls that extend across its infrastructure.

According to Shainer, agentic workloads strengthen the rationale for scale-in, as they require fast communication between graphics processing units while also controlling what agents can access, what actions they can take, and how their activity affects other workloads.

Nvidia is using BlueField, its data processing unit, alongside DOCA software and OpenShell agent runtime to achieve this expansion of security controls.