- AOI shipped more than 1 million 1.8 GHz cable amplifiers in about 16 months
- Amplifiers enabled by AI are already in the field, helping operators locate and triage faults
- The company is testing 3 GHz-plus coax that could rival 25 Gbps symmetrical service
For about three decades, the telecommunications industry has predicted the imminent demise of coaxial cable. Fiber was going to replace it. Then fiber was definitely going to replace it. Then, rather inconveniently, somebody invented yet another generation of DOCSIS.
And here we are again.
In an interview with FNTV this week, Steve Pederson, SVP and GM of broadband business at Applied Optoelectronics (AOI), described a company experiencing extraordinary demand for technology designed to extend the useful life of existing cable networks.
AOI has shipped more than 1 million 1.8 GHz amplifiers in approximately 16 months, a pace Pederson described as unlike anything he had previously experienced in his career.
But the bigger story isn't simply the survival of coax. It's the emergence of intelligent physical infrastructure, and the extraordinary economics of making something you've already paid for considerably more useful.
The economics of not digging everything up
Fiber is an excellent technology. Nobody would argue that. But replacing an existing hybrid fiber-coaxial (HFC) network is a very different economic proposition from building a new one.
AOI serves both markets. Pederson explained that passive optical networking (PON) makes sense for greenfield construction, while upgrading existing HFC infrastructure to DOCSIS 4.0 can deliver comparable services at a fraction of the cost of wholesale replacement.
"The cost to upgrade that infrastructure versus a wholesale change-out and going to PON is a fraction of the price," he told me.
Most major operators pursue both strategies. This isn't a contest over which technology is theoretically superior. It's about getting the greatest return from billions already invested in physical infrastructure.
I collect old cars, including a 1972 Chevy Malibu I bought three weeks after I first arrived in America in 1993. It's probably the most reliable car I own. After all these years, it doesn't owe me anything.
I suspect a few cable operators feel similarly about their coax.
The network gets a brain
The most interesting revelation in my conversation with Pederson concerned automation.
Physical access networks have traditionally been among the least intelligent components of telecommunications infrastructure. But AOI's latest amplifiers incorporate microprocessors, collect operational telemetry and support intelligent configuration.
That intelligence is increasingly connected to cloud-based systems capable of analyzing performance across large networks.
Pederson explained that the technology is already being deployed.
"We're starting to have deployments of our overall solutions out there in the field that include AI," he said, describing systems capable of troubleshooting and geolocating physical problems so technicians can go directly to the fault rather than searching through entire cascades of amplifiers.
This is no longer theoretical. Charter's Spectrum has selected AOI's QuantumLink software to remotely manage its installed base of intelligent amplifiers.
The next step is automated triage and mitigation.
If somebody drives a car (preferably not a '72 Malibu) into a roadside cabinet, software cannot magically reconstruct the equipment. But intelligent systems may be able to isolate the fault, mitigate its impact and maintain some level of service until technicians arrive.
That's an important distinction between autonomous network marketing and practical automation.
And unlike many grand promises about AI, this is already beginning to happen.
3 GHz and beyond
AOI is also exploring the next generation of coaxial technology, extending beyond today's 1.8 GHz spectrum toward what Pederson described as "three gigahertz plus."
The ambition is striking: technology capable of competing with 25 Gbps symmetrical services.
"I think in the next evolution of this, the industry will kind of look and say, 'Okay, at what point or how far can we stretch the bandwidth?'" Pederson explained.
AOI demonstrated a 3.0 GHz proof of concept at SCTE TechExpo, although significant engineering and economic questions remain before commercial deployment.
The critical issue is how far spectrum can be extended before the cost of doing so outweighs the value of retaining the existing network.
There is, after all, a limit to how much money you should spend keeping an old car running. Even a Chevy Malibu.
Two infrastructure revolutions
AOI is simultaneously expanding its data center business, adding manufacturing capacity in the United States and Asia to support demand from both markets.
Pederson remains bullish on AI infrastructure despite my concerns about data center financing, debt and the possibility of a correction.
But AOI occupies an interesting position between two very different infrastructure investment cycles.
One involves spending enormous sums building new computing facilities in anticipation of future demand. The other involves extracting more economic value from infrastructure that already exists.
And increasingly, those two worlds are connected.
The commercial value of AI cannot remain confined to giant data centers. It must reach businesses, factories, machines and people through physical networks.
AOI is helping build those networks while making them more intelligent.
Coax, it seems, has not only declined to attend its own funeral. It's decided to learn a few new tricks.
Learn more about AOI on Fierce Network
- AOI bets on AI-driven network maintenance as cable operators look beyond DOCSIS 4.0 — Elizabeth Coyne, Fierce Network, Oct. 2, 2026.
- Charter selects AOI software to control its cable network amplifiers — Linda Hardesty, Fierce Network, June 15, 2026.
Stephen M. Saunders MBE is a communications analyst and USPTO-registered inventor examining how digital infrastructure — 5G, cloud and AI — is reshaping industry, power and society, as well as underpinning the emerging, ubiquitous global digital economy. As anchor of FNTV and a longtime industry insider, he focuses less on growth narratives and more on execution, risk and how hyperscale technology is distorting markets, governance and society at scale.