Zayo is constructing more than 8,000 miles of new long-haul fibre across the United States in a collaboration with Nvidia, aimed at closing what the company calls a widening gap between surging AI compute demand and available network capacity.
The expansion will deliver six new long-haul routes through emerging AI corridors, alongside overbuilds of existing infrastructure in ten high-demand markets. Zayo said the project is designed to connect AI factories, GPU clusters and hyperscaler deployments as they expand beyond traditional data centre hubs.
Steve Smith, chief executive of Zayo, said the company had invested heavily in modelling where AI-driven demand will emerge and is building ahead of that demand rather than focusing solely on existing routes. He added that the partnership with Nvidia aims to extend high-capacity connectivity to neoclouds and enterprise AI users, not only the largest hyperscalers.
Dylan Patel, chief executive of SemiAnalysis, noted that connectivity is becoming as significant a constraint on AI scaling as compute itself. He highlighted the need for long-haul fibre to link distributed infrastructure in corridors where capacity remains scarce.
ALSO READ: MTN UGANDA PROFITS SURGE 37.7% IN FIRST HALF OF 2026
The announcement comes after Zayo’s acquisition of Crown Castle’s Fiber Solutions business, which added 90,000 metro route miles and 40,000 on-net enterprise locations to its network. The deal has strengthened Zayo’s metro density, particularly for AI inference workloads.
Over the past 18 months, Zayo said its AI-focused build and overbuild projects have covered more than 15,000 route miles across North America. The company also recently launched its AI Infrastructure Blueprint, a framework designed to connect training, inference and interconnection environments across the AI ecosystem.
Vladimir Troy, vice president of engineering for AI infrastructure at Nvidia, said network capacity is becoming as critical to AI progress as compute.
