FGN to acquire ACME Broadband to speed AI network traffic
What's the deal? FGN Inc.Dealroom has a profile for this one. Try Dealroom → has signed a definitive agreement to acquire ACME BroadbandDealroom has a profile for this one. Try Dealroom →, a managed network performance company focused on AI inference and other real-time applications. The Claymont, Delaware-based buyer announced the deal on September 17, 2026. Terms were undisclosed.
What does each side bring? FGN operates a federation of independent broadband networks and licenses an 11-patent portfolio of real-time communications and network-performance technologies. ACME provides enterprises and AI infrastructure providers with a managed performance layer that runs across their existing connectivity.
What's the endgame? The companies want to improve how performance-sensitive traffic travels between users, applications, and distributed compute infrastructure. The combined platform will measure available network paths and apply real-time and historical intelligence to route critical traffic.
Why it matters: Fiber provides capacity, but bandwidth alone does not determine responsiveness. Interactive inference, AI agents, voice AI, and industrial robotics all depend on consistent communication between users, machines, and compute.
"AI performance starts before a request reaches the GPU," said Darcy Lorincz, president of FGN Inc. "Bringing ACME Broadband into FGN gives us a clear commercial platform for optimizing that network journey."
What changes for customers? ACME will continue as the customer-facing brand for enterprises and AI infrastructure providers. FGN stays focused on independent broadband operators through the Fiber Gaming NetworkDealroom has a profile for this one. Try Dealroom →, FGN BOOSTDealroom has a profile for this one. Try Dealroom →, and the FGN Marketplace. Enterprises can add ACME's managed layer without replacing their existing providers.
The signal: The deal reflects a growing push to optimise the network path around AI workloads, not just the compute behind them, as inference moves closer to the edge.
Read more: openpr.com
Image credit: Wesley Fryer
