NVIDIA and HPE expanded the HPE AI Factory portfolio to support enterprise agentic AI deployments. New additions include the Vera CPU, Agent Toolkit, confidential computing capabilities, advanced networking technologies, and stronger governance tools. The updates aim to improve security, performance, and scalability for production AI environments.
NVIDIA and HPE Expand AI Factories
Enterprises are increasingly moving agentic AI projects from experimental stages into production environments. As adoption grows, infrastructure demands are becoming more complex.
During HPE Discover in Las Vegas, NVIDIA and HPE announced a major expansion of the HPE AI Factory with NVIDIA.
The updates introduce new hardware, software, security tools, and networking technologies designed to support large-scale AI deployments.
The expansion focuses on helping organizations build and operate AI agents more efficiently. It also strengthens security, governance, and performance across enterprise AI environments.
The announcement highlights how businesses are preparing for the next generation of AI factories built specifically for agentic workloads.
Also Read: NVIDIA and SK Hynix Announce Multiyear AI Memory Partnership Expansion
Vera CPU Powers Enterprise AI Agents
One of the biggest additions announced by NVIDIA and HPE is the NVIDIA Vera CPU, which will be available with HPE Private Cloud AI in 2027 through the HPE ProLiant Compute DL394 Gen12 server.
Designed specifically for AI agents, the processor handles orchestration tasks, tool execution, and real-time data processing while delivering low-latency performance.
The New York Stock Exchange, together with Redpanda and NVIDIA and HPE, is already exploring the new platform. Vera also forms part of the NVIDIA Vera Rubin architecture, which is entering full production through the NVIDIA Vera Rubin NVL72 rack-scale system.
HPE is further expanding its capabilities with the HPE Compute XD700, which supports up to 128 Rubin GPUs per rack. The NVIDIA Agent Toolkit has been launched for use with the HPE Private Cloud AI solution.
This NVIDIA Agent Toolkit combines the capabilities of the NVIDIA Nemotron open models, NVIDIA OpenShell secure runtime, and NVIDIA NemoClaw blueprints to help companies oversee their AI agents, enforce governance policies, and securely operate multi-agent autonomous systems. With the latest HPE Zerto Software, companies can defend against any unintended agent actions.
Confidential Computing Strengthens AI Security Systems
Confidential Computing by NVIDIA is now available across all HPE AI Factory offerings through HPE services. It protects sensitive data and AI models while executing with encryption and cryptography.
Such a solution can be used to build both on-premises and sovereign AI. The HPE ProLiant Compute DL380a has been certified by NVIDIA’s Confidential Computing initiative.
The additional security is ensured by NVIDIA BlueField DPUs and NVIDIA DOCA, ensuring zero trust policy enforcement, runtime threat protection, and encrypted networks—without sacrificing performance.
NVIDIA’s new portfolio includes the NVIDIA RTX PRO 6000 Blackwell Server Edition GPU, NVIDIA Spectrum-X Ethernet networking, NVIDIA BlueField-3 DPUs, and NVIDIA ConnectX-8 SuperNICs.
The improvement in networking options, coupled with new software partners, further solidifies the ecosystem, enabling businesses to bring their AI projects from concept to reality.
Also Read: Nvidia Hits $5.4 Trillion Market Cap as Jensen Huang Joins Trump on China Trip
