US fiber infrastructure expansion must reach 373 million miles by 2029

The immediate implication is that the physical layer is now the primary bottleneck for AI scaling, forcing a shift from software-centric discussions to massive hardware deployment. For the streaming ecosystem, this expansion provides the necessary low-latency foundation for distributed micro data centers and edge compute, which are critical for next-generation video analytics and real-time applications. As GPU densities rise, the industry must adopt modular fiber management to prevent operational complexity from stalling infrastructure growth. Watch for whether public funding programs like BEAD successfully accelerate these middle-mile upgrades to support AI-enabled applications in rural markets.

The Fiber Broadband Association's 373-million-mile target arrives amid a wave of federal and private investment aimed at closing the gap between current fiber coverage and AI-era demand. In June 2026, Ericsson launched its AI in RAN commercial software subscription, claiming up to 20% higher downlink throughput across more than 15 live deployments, underscoring how network operators are already layering AI workloads onto existing infrastructure and creating immediate backhaul pressure that fiber expansion must relieve. Verizon disclosed at the same time that its 60,000-site vRAN deployment is now applying agentic AI to configuration changes and service assurance, a scale of automation that depends on deterministic low-latency transport between distributed sites and regional compute hubs.

On the business and partnership side, Nokia is assembling the cloud and data architecture that will sit atop expanded fiber networks. At DTW Ignite in June 2026, Nokia announced work with AWS and Databricks to build the data, cloud, and control layers for autonomous networks, positioning its Autonomous Network Fabric as the orchestration plane between telco edge workloads and centralized AI models. The Databricks proof-of-concept demonstrated code-once data-processing workflows that run across proprietary and open-source stacks, while the AWS integration brings Amazon Bedrock and SageMaker tools into Nokia's fabric for cloud-scale inference. Nokia reported that operators using its autonomous networks portfolio are already achieving automation rates above 90 percent, service delivery times under four hours, and up to 85 percent reduction in slice rollout time, metrics that presuppose the fiber density the Fiber Broadband Association says is still years away.

From a technical standpoint, the divergence between Ericsson and Nokia on AI-RAN architecture highlights how fiber capacity constraints shape vendor roadmaps. Ericsson and Nokia are now diverging sharply on AI-RAN strategy, with Nokia building its entire Layer 1 RAN to run on Nvidia CUDA software following Nvidia's $1 billion investment in the Finnish company, while Ericsson anchors its approach in standalone 5G cores and programmable RAN. Ericsson's CTO Erik Ekudden has noted that , with uplink growth already outpacing downlink by 50 percent in roughly a third of operator networks. That uplink surge directly reinforces the Fiber Broadband Association's argument that current mileage is insufficient for the bidirectional bandwidth AI workloads demand, particularly at the edge where streaming and inference converge.