HPE launches AI‑native telecom kit spanning core to edge

HPE has introduced new networking, compute, and cloud operations products for telecoms and other service providers, focusing on infrastructure for AI workloads and 5G networks from the data centre core to the edge.

The package includes new Juniper routing platforms, updates to Juniper's wide area network management software, new HPE ProLiant telco servers, and a new cloud operations software stack. HPE also unveiled a new financing offer through HPE Financial Services.

The announcement is part of HPE's broader push into service provider infrastructure following its integration of Juniper Networks. HPE is positioning the strategy around what it calls AI-native networking, security, and cloud operations, alongside compute systems designed for telecom environments.

Rami Rahim, Executive Vice President, President and General Manager, Networking, HPE, said demand patterns linked to AI are changing the requirements for carrier networks and data centres.

"AI infrastructure is a critical growth driver for service providers. More than ever, HPE is committed to helping these customers lead in the AI era by building intelligent, next-generation networks that can support complex operations, rising data traffic, and the transformative capabilities of AI."

Routing upgrades

On the networking side, HPE added two new Juniper PTX router lines aimed at scaling traffic linked to AI and cloud services. The systems use 800G interfaces and target network fabrics that connect data centres and large compute clusters.

The Juniper PTX12000 line is a modular platform running on what HPE calls 1.6T-ready hardware, designed to expand capacity without repeated network redesigns as traffic grows.

The PTX12000 range includes an eight-slot PTX12008 and a 12-slot PTX12012. HPE said the platforms can scale to 345.6T and 518.4T respectively, and are built to deliver predictable, low-latency performance as port densities rise.

HPE also introduced the Juniper PTX10002 line of fixed form-factor routers in a 2RU chassis, aimed at AI network fabrics where space and power constraints matter. HPE said the routers support 28.8T or 14.4T of throughput and include multi-rate 100G, 400G, and 800G options.

Alongside the routers, HPE highlighted updates to Juniper Routing Director, its routing management product. HPE said Routing Director is now "agentic-AI ready," allowing organisations to connect it to their own AI copilots for WAN automation and post-deployment operations.

Telco compute

HPE's compute updates focus on systems targeted at 5G radio access networks and core network workloads that increasingly run on standard servers. The company introduced the HPE ProLiant Compute EL9000 chassis and the EL140 Gen12 server.

HPE said the combination delivers twice the fronthaul network bandwidth of previous generations, while core counts increase by 20%. It said the update enables operators to handle more traffic per server with fewer systems in the field.

HPE said each server provides 72 CPU cores and 24 network ports in a 2U form factor, with the chassis supporting up to two servers. The platform uses Intel Xeon 6 SoCs with vRAN boost and includes built-in security features aimed at telecom requirements.

HPE is also extending availability of the Juniper Cloud Native Router on its ProLiant systems. HPE said it now runs on the 1U ProLiant Compute DL110 and the new 2U ProLiant EL140 Gen12. HPE said this consolidates RAN and routing functions in a single server, reducing the need for separate routing hardware at cell sites.

Cloud operations

Another element is HPE Cloud Ops Software, described as a unified control plane for cloud operations across virtual machines and containers. HPE said it includes monitoring and observability features, AIOps functions, cyber resiliency, DevOps automation, and FinOps tools.

HPE said service providers are reassessing cloud operations because of higher virtualisation costs, sovereign data requirements, and the need to run both VMs and containers across 5G core environments. It framed the software stack as a way to manage multi-vendor and multicloud operations through a single platform.

Ray Mota, CEO and Principal Analyst, ACG Research, linked the product set to shifting traffic patterns associated with AI and new requirements around latency and uplink capacity.

"As AI reshapes traffic patterns and drives new uplink, latency, and capacity demands, HPE's leadership in high-performance routing, switching, AI-native automation, telco cloud architectures, and security ensures operators are equipped for the next era of AI-driven connectivity," said Ray Mota, CEO and Principal Analyst, ACG Research.

Financing offer

HPE Financial Services added a programme called 90/9 Advantage, which HPE said applies across its networking, compute, storage, and software portfolio. Under the scheme, customers make no payments for the first 90 days, followed by monthly lease payments of 1% for the next nine months.

HPE plans to demonstrate the new products at Mobile World Congress 2026, including live demonstrations of network automation, edge compute, and private 5G. Rahim said the company is emphasising simpler management across networking and compute as telecom operators continue network modernisation programmes.