HPE AI Factory with NVIDIA Expands for the Agentic Era

HPE and NVIDIA are expanding HPE AI Factory with NVIDIA to help enterprises move AI agents from proof-of-concept (PoC) stages into real production environments. This expansion includes the NVIDIA Vera CPU and NVIDIA Agent Toolkit, optimized for agent tool calling and real-time data processing.

The NVIDIA Vera CPU delivers the deterministic, low-latency performance required for agent loops, and will be available with HPE ProLiant Compute DL394 Gen12 servers in 2027. Additionally, the NVIDIA Vera Rubin platform, designed for large-scale models with over a trillion parameters, will be offered with HPE's rack-scale system, NVIDIA Vera Rubin NVL72.

The NVIDIA Agent Toolkit, combined with HPE Private Cloud AI, acts as an operating system for monitoring agent behavior, enforcing governance policies, and safely building autonomous multi-agent systems. Key features include:

• Security and Governance: Local agent registration enables pre-execution approval procedures for models and tools, while HPE Zerto Software enables detection of abnormal agent behavior and data recovery.
• Data Management: HPE Alletra Storage MP X10000 automatically applies metadata and governance policies to unstructured data, improving token throughput in AI pipelines.
• Confidential Computing: NVIDIA Confidential Computing is applied across all HPE AI Factory solutions, encrypting and protecting running models and sensitive data.