HPE expands AI networking with new unified features & hardware
HPE has broadened its AI-native networking portfolio, introducing expanded capabilities across both its Aruba Networking and Juniper Networking solution lines. This forms a major step in the integration process following HPE's recent acquisition of Juniper Networks. The company's new offerings aim to enable self-driving operations that maximise performance and scalability for AI workloads.
Expanded AIOps
HPE's latest portfolio includes unified AIOps features designed to support both HPE Aruba Networking Central and HPE Juniper Networking Mist platforms. These capabilities use a combination of agentic AI, a microservices framework, and shared hardware to deliver a consistent network management experience. The update introduces features that increase observability and automation, assisting IT teams in maintaining hybrid cloud environments.
The Juniper Networking Mist Large Experience Model (LEM) will now be available on HPE Aruba Networking Central. This system analyses substantial data volumes from applications such as Zoom and Teams, incorporating synthetic data via digital twins, to identify, resolve, and predict video issues. Meanwhile, Aruba's Agentic Mesh technology will now enhance Mist, improving anomaly detection and root-cause analysis with advanced reasoning features. Mist will also integrate new organisational and NOC views from Aruba, aligning operational insights across platforms. Newly introduced WiFi-7 access points are compatible with both Aruba Central and Mist, offering flexibility for organisations planning infrastructure investments.
Aruba Networking Central On-Premises 3.0 introduces further functionality with advanced generative and traditional AIOps, actionable alerts, automation, and a new interface designed for secure on-premises deployments.
High-performance hardware
High-speed networking is increasingly vital as AI inferencing moves towards the network edge. To address these demands, HPE has launched the Juniper Networking QFX5250 switch. Built with Broadcom Tomahawk 6 silicon, the switch offers 102.4Tbps bandwidth and is described as suitable for connecting GPUs in data centres, combining Junos software, liquid cooling and AIOps for a more efficient and manageable infrastructure.
The Juniper Networking MX301 multiservice edge router seeks to bring AI inferencing closer to data sources. Presented in a compact 1RU form factor, the MX301 delivers 1.6 Tbps performance and 400G connectivity across edge, metro, mobile backhaul and enterprise routing environments.
Partner integrations
HPE is extending its collaborations with NVIDIA and AMD to accelerate AI deployments further. Networking solutions for AI factories have been expanded to use high-speed connections between users, devices, and large AI clusters over longer distances or across multi-cloud environments. The offering supports NVIDIA Spectrum-X Ethernet and BlueField-3 DPU technologies, which intend to improve workload experiences in production scenarios.
With AMD, HPE is contributing to the Helios AI rack-scale architecture, billed as a single, turnkey system for large-scale AI training and inferencing. It features what is described as industry-first scale-up Ethernet networking, using a Juniper switch and Broadcom silicon to reach 260TB/s bandwidth and 2.9 exaflops FP4 performance within a rack system.
Unified IT operations
HPE is also refreshing its approach to hybrid cloud operations with enhancements to OpsRamp software, providing full-stack observability and predictive troubleshooting. Deeper integration with its GreenLake platform supports the consolidation of telemetry from multiple domains-compute, networking, and storage-enabling a more comprehensive hybrid command centre for IT teams. New features include the connection of third-party AI agents for no-code integrations and updates that provide guided remediation, sustainability insights and wellness dashboards.
Network financing
HPE Financial Services has introduced zero-percent financing options to encourage adoption of AI-native networking. Customers buying networking AIOps software can now do so on term-based licences, and those investing in new datacentre or enterprise routing hardware can benefit from reduced upfront costs and multi-vendor trade-in options.
"In the era of AI, customers need networks that are purpose-built with AI and for AI to handle the rapid growth of connected devices, complex environments, and increasing security threats," said Rami Rahim, executive vice president, president and general manager, Networking, HPE. "By delivering autonomous, high-performing networks, HPE is poised to disrupt the networking industry with future-ready solutions that redefine user experiences and provide robust, secure connectivity across all environments."
