Hewlett Packard Enterprise and Nvidia have expanded their partnership to focus on secure, large-scale deployment of AI agents. The companies aim to provide solutions for regulated industries that require hybrid or on-premises infrastructure to maintain data control and compliance.
Starting in July 2026, HPE Private Cloud AI will incorporate new agentic capabilities and security features. Additional functions are scheduled for release in the fourth quarter of 2026. The platform now integrates the Nvidia Agent Toolkit to help organizations orchestrate AI workflows within controlled environments.
The collaboration introduces the HPE ProLiant Compute DL394 Gen12 server, which utilizes Nvidia Vera CPUs. This hardware is designed to process sensitive data locally, avoiding the need for public cloud infrastructure. Security is further enhanced through Nvidia Confidential Computing, which creates isolated execution environments.
To mitigate risks from autonomous AI actions, HPE is adding its Zerto continuous data protection product. This tool provides a rollback mechanism if AI systems cause errors in databases or workflows. This integrated stack aims to offer a turnkey solution for enterprises that cannot rely on hyperscalers due to data sovereignty and regulatory requirements.
