HPE Expands AI Networking Portfolio with New PTX Routers

Hewlett Packard Enterprise announced several new networking and cloud technology updates for service providers at the recent Mobile World Congress 2026 (MWC26). The company revealed new hardware, software, and financing options designed to help networks handle the growing demands created by artificial intelligence (AI).

According to HPE, the rapid adoption of AI is creating a major increase in data traffic, especially in AI-focused data centers. Large cloud providers and hyperscalers are investing heavily in these environments to deliver AI-powered services. In such systems, networks must offer extremely low latency and high reliability to ensure smooth performance.

While traditional internet traffic continues to grow, AI traffic is expanding much faster and behaves differently. AI workloads are highly sensitive to delays, which means older network designs may struggle to keep up with demand.

To address this challenge, HPE is expanding its routing portfolio built on technology from Juniper Networks. The company introduced new PTX12000 modular routers, designed to support dense 800G connectivity and scale up to 1.6 terabits per second in the future.

HPE also launched PTX10002 fixed-form routers, which are smaller and more efficient systems designed to help connect AI data centers and support modern network infrastructure.

Both routers run on Juniper Express 5 silicon, a specialized networking chip designed to improve performance, buffering capacity and power efficiency. The company said the chip delivers around 49% better energy efficiency compared with previous versions. In addition, the PTX10002 routers can offer up to 54% improved efficiency over earlier platforms.

The new silicon also includes built-in security features such as inline MACsec encryption to help protect network traffic.

HPE is also upgrading its Juniper Routing Director software platform. The tool helps service providers and large enterprises automate wide-area network operations. With the latest update, the platform is designed to work with AI copilots and custom AI models developed by customers. This integration can help network operators automate tasks, monitor systems more effectively and resolve issues faster.

HPE executives noted that organizations are increasingly building larger clusters of graphics processing units (GPUs) across different locations. Connecting these distributed systems requires powerful, reliable networks, making advanced networking infrastructure essential for the continued growth of AI services.