AI growth runs on networks

AI growth is not only a compute story. Networking has become one of its strongest growth engines.

Recent results show momentum across the industry:

• Cisco networking revenue grew 25% year over year, while data center switching orders increased more than 40%.
• Arista Networks revenue grew 37.7%.
• NVIDIA data center networking revenue reached a record $14.8 billion, up 199%.
• Hewlett Packard Enterprise networking revenue grew 10% on a normalized basis. Data center switching orders increased nearly 20%, while routing orders grew nearly 30%.
• Broadcom AI revenue grew 106%, supported partly by demand for AI networking.

These figures measure different business categories, but they point in the same direction.

They also focus heavily on data center infrastructure. They do not reflect the critical role of campus and branch networks in AI growth.

As agentic AI scales, more AI-driven activity will reach users, devices, applications, and physical environments. Campus and branch networks will need to support higher traffic volumes, real-time decisions, stronger security, greater resilience, and more intelligent operations.

Cisco, Arista, NVIDIA, HPE Juniper, and Broadcom are approaching this opportunity from different positions across systems, networking, and silicon.

The next phase of AI infrastructure requires a complete architectural discussion.

• Compute provides the processing power.
• Data center networks connect the AI infrastructure.
• Campus and branch networks bring AI to where people and businesses operate.

Where do you see the biggest AI infrastructure constraint today: compute, networking, power, cooling, or data?