HPE Private Cloud AI adds MCP servers, mixed H200 and RTX PRO 6000 clusters

At NVIDIA GTC Berlin, HPE updated HPE AI Factory with NVIDIA: Private Cloud AI now offers governed API and MCP access to inference, models and pipelines, and clusters can mix H200 and RTX PRO 6000 Blackwell Server Edition GPUs. HPE also introduced a liquid-cooled NVIDIA Vera CPU Rack with 256 CPUs for agentic workloads. Mixed clusters and KAI scheduling are available now, with the API platform rolling out in late 2026.

• MCP servers give AI agents governed access to models and pipelines with audit logging and RBAC
• A single cluster can combine H200 and RTX PRO 6000 Blackwell Server Edition GPUs
• The Vera CPU Rack carries 256 CPUs and supports over 22,500 sandboxes
• HPE claims 80% faster agentic CPU performance without naming a baseline