HPE President and CEO Antonio Neri and NVIDIA Founder and CEO Jensen Huang expanded their joint AI factory portfolio at HPE Discover Las Vegas 2026 on June 16, 2026. The new capabilities move agentic AI from proof-of-concept into production with enhanced security, governance, and control. This expansion introduces the NVIDIA Vera CPU and the NVIDIA Agent Toolkit for HPE Private Cloud AI, creating a full-stack infrastructure for autonomous enterprise agents.
The collaboration focuses on transforming companies into “agentic enterprises,” where AI agents adapt, evolve, and collaborate to automate business processes. High-performance data processing is central to this shift, supported by the new HPE ProLiant Compute DL394 Gen12 server. As African IoT sector growth accelerates, these enterprise-grade architectures provide a framework for managing complex distributed AI workloads across sovereign networks.
HPE is also introducing NVIDIA Confidential Computing across its AI factory solutions, including at-scale and sovereign models. This technology uses cryptographic attestation to protect models and private data during execution, satisfying regional regulatory standards. By establishing a chain of trust at every stage of hardware and software verification, the system allows enterprises to maintain control over their most sensitive AI assets.
Specific hardware for agentic orchestration
The hardware foundation of this rollout includes the HPE ProLiant Compute DL394 Gen12, which will be available with the NVIDIA Vera CPU. This processor is the first built specifically for AI agents, designed to handle tool calls, orchestration, and real-time data processing. For larger clusters, the HPE Compute XD700 chassis now supports the NVIDIA HGX Rubin NVL8, which can scale up to 128 Rubin GPUs per rack.
Data efficiency has also been upgraded to reduce the latency that often stalls production AI. The HPE Alletra Storage MP X10000 provides inline intelligence by applying metadata and governance policies to unstructured data in real-time. According to HPE, this can cut token response times by up to 20 times, while improvements to HPE Private Cloud AI boost token throughput by up to 20%.
For manufacturing and industrial sites, pivoting to Manufacturing Execution Systems requires this type of high-velocity data handling. The integration of HPE Data Fabric Software broadens data availability for agentic workflows by extending Model Context Protocol (MCP) support to Apache Airflow. This ensures agents have a governed, persistent inventory of enterprise data to pull from.
Guardrails for autonomous agent actions
Managing the risks of autonomous agents is a key priority for the new governance tools. HPE Zerto Software now includes capabilities to detect rogue agent actions, using continuous data protection to rewind the system to a clean state if an agent behaves outside its defined parameters. This provides a safety net for enterprises automating high-stakes business logic.
IT departments can now use secure local agent registration to approve specific AI models, skills, and tools. This ensures that all deployed agents adhere to centralized security policies before they are integrated into the workforce. The New York Stock Exchange, in collaboration with Redpanda and HPE, is currently exploring the use of the NVIDIA Vera CPU with the ProLiant DL394 Gen12 server to evaluate its data-processing performance.
Global reaches of the Unleash AI program
To support regional compliance, HPE has opened a new AI Factory Lab in Grenoble, France. This facility allows European customers to validate performance on EU-located infrastructure, specifically addressing data sovereignty and regulatory requirements. The lab serves as a testing ground for companies that must keep their AI workloads within specific jurisdictions.
The software ecosystem supporting these deployments is growing through the Unleash AI Partner Program. Nearly a dozen new AI software partners have joined the program, including Aizen, BridgeTEK, deepset, Deliverance, Faclon Labs, Gallop, Rocket, Supervity, Thales, Trustwise, and Vortiqx. These partners provide the specialized tools necessary for sovereign AI and at-scale factory deployments.
HPE Sovereign AI Factory and HPE AI Factory at-scale are now available with a full-stack NVIDIA integration. This includes the NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, NVIDIA Spectrum-X Ethernet networking, and NVIDIA BlueField-3 DPUs. These components collectively allow enterprises to protect agents and data across the AI factory while maintaining operational efficiency.
