HPE and NVIDIA add Vera CPU racks, governed MCP services and agent security to the HPE AI Factory portfolio
The AMW Read
An incremental update to an established OEM–NVIDIA partnership; it matters at the segment level because it defines CPU-dense, governed agent infrastructure as an enterprise and sovereign on-prem product category.
HPE and NVIDIA add Vera CPU racks, governed MCP services and agent security to the HPE AI Factory portfolio
At NVIDIA GTC Berlin 2026 on October 21, HPE added new capabilities to its HPE AI Factory with NVIDIA portfolio. HPE Private Cloud AI now gets HPE AI Essentials, which gives developers, ISVs and AI agents governed APIs and versioned Model Context Protocol servers. Access is controlled by role, tenant and project, with audit logging and machine identity, and it works in both connected and air-gapped deployments. Clusters that mix NVIDIA H200 and RTX PRO 6000 Blackwell Server Edition GPUs are now generally available, as is NVIDIA's open-source KAI Scheduler. HPE also introduced the NVIDIA Vera CPU Rack by HPE. It is a liquid-cooled system with 256 Vera CPUs that supports 22,500 sandboxes, and HPE says it delivers 80% faster agentic CPU performance. Schneider Electric plans to use Private Cloud AI for industrial AI agents at the edge.
The announcement treats agent workloads as their own infrastructure category, separate from GPU inference. Agents need CPU-dense sandboxes for tool execution and reinforcement-learning evaluation. They also need isolation and governance at the runtime layer. HPE's support for the NVIDIA OpenShell reference architecture combines a secure runtime on Vera, NVIDIA Sentry on BlueField-4 for hardware-enforced policy, and Vera Rubin NVL72 for inference. Together these make NVIDIA's stack reach from the chip up to agent controls. OEMs like HPE package that stack for sovereign and regulated buyers, a market that hyperscalers and neo-clouds don't fully cover.
For builders, MCP is now becoming an enterprise-governed interface rather than a developer convenience. Agent vendors selling into regulated on-prem accounts should expect role-scoped, audited MCP endpoints to be a procurement requirement. Note the timing: the AI Essentials API platform rolls out in late 2026, and self-service agent registration and MCP discovery are planned for 2027.