The expanded portfolio is designed to deliver secure, autonomous, and scalable networking capable of supporting demanding AI workloads.
Hewlett Packard Enterprise has expanded its AI-native networking portfolio, strengthening its position in AI infrastructure with an integrated suite of technologies derived from HPE Aruba Networking and the recently acquired HPE Juniper Networking. The announcement marks a major milestone in the rapid integration of Juniper Networks, coming just five months after the acquisition was completed. The expanded portfolio is designed to deliver secure, autonomous, and scalable networking capable of supporting demanding AI workloads.
The latest update introduces unified AIOps capabilities and common hardware that provide a consistent, self-driving operational experience across both networking platforms. New capabilities within HPE OpsRamp Software, alongside updated switching and routing solutions from HPE Juniper Networking, position the network as a foundational layer for AI and cloud performance. These advancements are built on an agentic AI framework compatible with HPE GreenLake Intelligence, helping organizations simplify and automate operations across hybrid environments.
As part of the integration effort, HPE is bringing key features from both Aruba and Juniper into a unified experience. HPE Juniper Networking’s Large Experience Model will now be accessible within HPE Aruba Networking Central, while Aruba’s Agentic Mesh technology will become available for Mist users. Additional updates include cross-platform organizational insights, new WiFi-7 access points, and enhanced on-premises capabilities through Aruba Networking Central On-Premises 3.0, which offers improved automation, analytics, and user interface upgrades.
HPE also expanded its AI infrastructure hardware portfolio with new switching and routing solutions tailored for high-performance AI workloads. The HPE Juniper Networking QFX5250 switch, powered by Broadcom Tomahawk 6 silicon, delivers 102.4 Tbps of bandwidth, enabling fast GPU-to-GPU connectivity in data centers. A new multiservice edge router—the MX301—brings high-speed AI inferencing capabilities closer to where data is generated, supporting performance-intensive edge environments.
Ahead of HPE Discover Barcelona 2025, the company also deepened collaborations with NVIDIA and AMD. This includes expanded AI factory networking solutions that enhance connectivity from edge environments to large AI clusters, as well as support for AMD’s new Helios rack-scale AI architecture. Helios introduces industry-first scale-up Ethernet networking, combined with an HPE Juniper Networking switch designed to support trillion-parameter training workloads.
HPE further advanced its hybrid cloud and AIOps strategy through updates to HPE OpsRamp Software and deeper integration with GreenLake. These enhancements bring together telemetry from compute, storage, and networking systems to give IT teams a unified command center for managing hybrid environments. New features include data center observability through Apstra integration, automated root-cause analysis, and support for Model Context Protocol to connect AI agents across different platforms.
To support customer adoption, HPE Financial Services has introduced zero-percent financing options for networking AIOps software and new leasing programs that offer cost savings for organizations upgrading to AI-native networking. These options aim to reduce barriers for enterprises investing in modern, AI-ready infrastructure.
Product availability will roll out over the coming months, with the QFX5250 switch scheduled for early 2026, the MX301 router set for release in December 2025, and various OpsRamp and GreenLake integrations completing between late 2025 and mid-2026.
