NVIDIA Launches OpenShell Reference Architecture for Safe AI Agents
NVIDIA launched the OpenShell Reference Architecture at GTC Berlin, a full-stack framework for building, containing, observing and enforcing safe AI agents across enterprise infrastructure and edge systems. The architecture combines Vera Rubin for inference, Vera CPUs for agent harnesses, and BlueField-4 DPUs running NVIDIA Sentry for in-silicon policy enforcement, with claimed 30x higher throughput per megawatt on Vera Rubin NVL72 systems. A broad ecosystem — including Anthropic, OpenAI, Microsoft, AWS, Google Cloud, CrowdStrike, SAP, ServiceNow and major banks BBVA and Santander — is working with the framework, and it is available now.
Key figures
- Vera cpu performance claim
- 2x performance of x86 processors
- Sentry agent capacity boost
- up to 20%
- Vera rubin nvl72 throughput claim
- 30x higher throughput per megawatt
AI analysis
The rest of the AI analysis, red flags and sentiment are part of Signal8 Pro.
AI-generated analysis of a public disclosure. Not investment advice; verify against the original document.