Cityside Fiber expands edge compute and smart-city infrastructure in California

The transition from simple broadband delivery to a distributed compute layer suggests that last-mile fiber is becoming the essential backbone for the physical AI economy. For the streaming and telecommunications ecosystem, this architectural shift moves computation away from centralized hyperscale data centers to the network periphery, reducing latency for real-time applications. As connectivity and computation become inseparable, traditional fiber providers may evolve into critical operators of localized AI fabrics rather than just bit pipes. Watch for whether municipal partnerships for smart-city sensors provide the necessary scale to make 5G edge cloud data centers economically viable for broader commercial use.

Cityside Fiber's pivot from pure FTTH delivery to edge compute infrastructure mirrors a broader movement among fiber operators seeking new revenue streams beyond residential broadband. In early 2025, Ericsson and several fiber operators launched a joint initiative to integrate edge compute capabilities into last-mile networks, targeting latency-sensitive applications such as autonomous vehicle coordination and real-time video analytics. The initiative reflects how telecom equipment vendors are positioning themselves as enablers of distributed compute architectures that sit between hyperscale cloud providers and end users. Åsa Tamsons, who leads Ericsson's Networks division, has publicly framed edge infrastructure as a natural extension of 5G standalone deployments, arguing that operators who control the last mile can monetize compute proximity in ways that pure cloud providers cannot.

On the regulatory and business side, municipal partnerships are becoming the primary commercial vehicle for edge compute fiber deployments. In March 2025, the Federal Communications Commission opened a proceeding examining how broadband infrastructure grants could fund edge compute and smart-city projects, signaling potential federal support for operators who bundle connectivity with localized processing. The proceeding specifically referenced the BEAD program as a possible funding mechanism for hybrid fiber-and-compute projects in underserved areas. Meanwhile, Jonathan Restivo, Cityside Fiber's CEO, told Fierce Network in a separate interview that the company's municipal agreements include revenue-sharing provisions for smart-city data services, a model that could serve as a template for other regional fiber operators pursuing similar strategies. Stephen M. Saunders, who chairs the Broadband Forum's edge computing working group, has argued that standardization of edge compute interfaces is essential before these deployments can scale beyond pilot programs.

Technical benchmarks for fiber-based edge compute remain limited, but early data suggests meaningful latency gains over centralized architectures. A 2025 study by the Broadband Forum found that edge compute nodes colocated with fiber aggregation points reduced round-trip latency by 60 to 80 percent compared to regional data centers, with the most significant improvements observed in applications requiring sub-10-millisecond response times such as real-time video transcoding and industrial IoT control loops. Ericsson's own testing has shown that , a figure that centralized cloud architectures typically cannot match without dedicated network slicing. These performance characteristics position fiber-based edge compute as a viable platform for streaming applications that demand ultra-low latency, including cloud gaming, live interactive video, and at the network edge.