# HPE Expands AI Factory for Agentic AI Deployments > Source: > Published: 2026-06-16 17:26:37.712261+00:00 # HPE Expands AI Factory for Agentic AI Deployments At HPE Discover 2026, HPE announced enhancements to its **HPE AI Factory** portfolio in partnership with **NVIDIA**, aimed at operationalizing agentic AI in production. According to HPE's press release, **HPE** introduced the **HPE ProLiant Compute DL394 Gen12** powered by the NVIDIA Vera CPU. StorageReview reports the expanded AI Factory adds support for the NVIDIA Agent Toolkit-including NVIDIA Nemotron, NVIDIA NemoClaw, and the NVIDIA OpenShell runtime-plus features for agent governance, policy enforcement, and behavioral monitoring. StorageReview also reports HPE is extending **HPE Zerto** continuous data protection to monitor agent actions and adding confidential computing capabilities for secure deployments. Antonio Neri, president and CEO of **HPE**, and Jensen Huang, founder and CEO of **NVIDIA**, provided quoted remarks in coverage about the need for secure, full-stack architectures for agentic AI. ### What happened **HPE** announced a set of enhancements to its **HPE AI Factory** portfolio at HPE Discover 2026, expanding the company's turnkey AI infrastructure co-engineered with **NVIDIA**, according to coverage in StorageReview and an HPE press release. Per HPE's press release, **HPE** introduced the **HPE ProLiant Compute DL394 Gen12** powered by the NVIDIA Vera CPU. StorageReview reports the AI Factory updates add support for the NVIDIA Agent Toolkit, which the outlet describes as including NVIDIA Nemotron, NVIDIA NemoClaw, and the NVIDIA OpenShell runtime. StorageReview also reports the portfolio additions include enhanced governance and monitoring features, extensions to **HPE Zerto** for continuous data protection of agent actions, and confidential computing capabilities. ### Technical details Editorial analysis - technical context: The announced **HPE ProLiant Compute DL394 Gen12** paired with the NVIDIA Vera CPU targets workloads the vendors characterize as requiring strong single-core performance and high memory bandwidth for sequential reasoning and reinforcement-learning style workloads, per HPE's product announcement. The reported inclusion of the NVIDIA Agent Toolkit and runtime components points to a stack that combines model runtimes, agent orchestration, and policy enforcement at deployment time. ### Context and significance Public reporting frames these updates as part of a broader vendor trend to deliver integrated hardware-plus-software stacks for agentic AI, combining CPU-optimized servers, GPU/accelerator support, runtime toolkits, and enterprise-grade security features. For practitioners, bundled solutions can reduce integration lift when moving from pilots to production, particularly for organizations that require stronger control, auditing, and confidential-computing protections for agent behavior. ### What to watch For practitioners: Track interoperability between the NVIDIA Agent Toolkit runtimes and existing MLOps and orchestration systems, the availability and licensing of the reported NVIDIA Nemotron models, and how confidential computing integrations are exposed to customer workloads. Also watch measured performance and cost comparisons of NVIDIA Vera CPU-based nodes versus GPU-centric alternatives for agentic workloads. ### Quoted context StorageReview and HPE's press release include on-record quotes: "As AI becomes more autonomous, organizations need a new architecture to run it securely, govern it responsibly, and scale it economically," said Antonio Neri, president and CEO of HPE. "Every layer of the computing stack is being reinvented for the age of AI agents," said Jensen Huang, founder and CEO of NVIDIA. "Together with HPE, we are building AI factories for this new era of computing, powered by NVIDIA Vera CPUs, accelerated infrastructure, and secure AI software, to help enterprises transform their data into intelligent action." ## Scoring Rationale This is a notable infrastructure announcement for enterprises building agentic AI, combining CPU hardware and runtime tooling. It matters to practitioners planning production deployments but is not a frontier-model breakthrough. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)