HPE pushes self-driving networks to handle AI data demands

Hewlett Packard Enterprise is advancing a networking strategy centered on self-driving networks—systems capable of anticipating problems, automating operations, and reducing the manual complexity that has plagued enterprise IT teams. The company made this vision public through an exclusive broadcast with theCUBE Research, in which Ben Baker, director of data center networking at HPE, outlined the three critical problems facing enterprise networks today: limited visibility into applications running on the network, insufficient speed and agility in responding to changes, and poor reliability that forces operators into constant firefighting.

The fundamental issue, Baker explained, is complexity. Enterprise networks today feature virtual networks, multiple tunnel protocols, and thousands of physical and logical devices—a situation where operators are "drowning in data but starved for insights." Network outages can cause hours or days of intensive recovery work, with massive reputational and financial consequences. HPE's solution rests on two pillars: establishing self-driving networks as a foundation for the agentic enterprise, and using GreenLake Intelligence—its agentic AI framework—as an intelligence layer across a unified platform. The self-driving network not only supports networking for AI but also uses AI to operate the network itself, leveraging HPE's integration of Apstra Data Center Director, a data center network management tool built on a graph database, with HPE Morpheus, a hybrid cloud management and automation platform. The graph database is designed to capture and maintain information about relationships among network nodes, allowing HPE to understand correlations among disparate data sources.

Central to this strategy is HPE's integration of tools from Juniper Networks, which it acquired, including the Mist AIOps platform and a data center network assurance product that can surface potential problems before they cause damage. The Mist platform sifts through data to identify issues relevant to specific network components. Baker highlighted a key insight from the Mist team: linking people in technical support directly to data scientists, creating a bridge between operational emergencies and analytical expertise. HPE is also expanding Marvis, its AI-driven networking assistant, into Aruba Central and adding support for HPE Networking CX switches to the Mist platform—part of an effort to consolidate tools from its Juniper and Aruba acquisitions under one umbrella.

Baker emphasized that HPE's networking investments have positioned the company uniquely in the industry, capable of bringing together compute, storage, networking, hybrid cloud, and full-stack IT infrastructure solutions. He dismissed concerns that the Juniper acquisition would slow HPE down, noting that teams from different business units have successfully executed portfolio integrations. On the infrastructure hardware side, HPE continues to invest in liquid-cooled networking, exemplified by the QFX5250, a fully liquid-cooled switch with 64 ports supporting speeds of up to 1.6 terabits per second. Baker acknowledged that AI's demand for bandwidth is "not going away anytime soon" and reaffirmed HPE's commitment to the self-driving network path, with the ultimate goal of optimizing user experience by allowing networks to operate reliably, much like utilities such as electricity or water—systems people expect to work without constant attention.