The Wire That Connects Everything (And the Stock Nobody Owns)

In 1976, a small team of engineers in Hickory, North Carolina, started making coaxial cable for the nascent cable television industry. Their timing was perfect. Cable TV was about to wire America's living rooms, and somebody had to make the stuff that carried the signal from the headend to the set-top box. For the next two decades, that's exactly what this company did. It made cable. Miles and miles of it.

Then came fiber optics, and the company pivoted again. Then Wi-Fi. Then 4G. Each generational shift in how data moved from Point A to Point B required new physical infrastructure, and each time, this firm was there with the connectors, the patch panels, the cable trays, and the network management systems that made it all work.

Here's what most people miss about the AI boom: it is, at its foundation, a plumbing problem.

Every hyperscaler data center that NVIDIA talks about on earnings calls, every GPU cluster that Alphabet (NASDAQ:GOOG) is deploying to support its $462 billion cloud backlog, every inference rack that Penguin Solutions is building with CXL memory architectures, all of it needs to be physically connected. Fiber runs between racks. Copper carries power distribution signals. Structured cabling systems organize thousands of connections so that technicians can actually find what broke at 3am. The software gets all the headlines. The wiring gets none of them.

This is not a metaphor. A single hyperscale data center uses between 20 and 50 miles of fiber optic cable internally. The number of data centers under construction in the United States alone has doubled since 2023. Microsoft, Amazon, Google, Meta, and Oracle have collectively committed over $350 billion in capital expenditure for 2026 and 2027, with the vast majority flowing into data center construction and expansion. Goog's CFO said on the Q1 call that 2027 capex will "significantly increase" above an already staggering $180-190 billion range. That money has to go somewhere physical.

The companies that supply the actual hardware for these facilities occupy a strange position in the market. They're essential (you literally cannot turn on a data center without their products) but they trade at valuations that reflect legacy telecom exposure rather than AI-era demand. Wall Street loves to price software companies on future TAM expansion. It prices infrastructure hardware companies on last year's margins.

Consider what's happened in the past 18 months. Indoor cellular solutions (think private 5G inside factories and warehouses) went from a niche product category to a strategic necessity for any enterprise running IoT sensors at scale. Wi-Fi 7 deployments are ramping and require entirely new access points and switching gear. The BEAD (Broadband Equity, Access, and Deployment) program is finally pushing $42.5 billion in federal broadband funding into actual construction contracts, which means rural fiber builds are accelerating nationwide. And data center operators, who used to buy off-the-shelf connectivity gear, are increasingly demanding custom structured cabling solutions designed around high-density AI rack configurations.

All of these demand vectors converge on a surprisingly small number of companies that can manufacture at scale, certify to telecom-grade standards, and deliver globally.

One of those companies rebranded itself at the start of 2026. The name change was telling. After nearly five decades operating under a name synonymous with cable television infrastructure, the company chose a new identity that emphasized networks and visibility rather than physical cable. The old name had served it well through the coax era, through the fiber transition, through the acquisition of networking assets from a major European telecom equipment maker. But the new name was a declaration: we're not a cable company anymore. We're the connective tissue of modern networks.

The timing of the rebrand matters because it coincided with a strategic restructuring that split the business into three distinct segments, each targeting a different layer of the network stack. One segment handles the physical layer (fiber and copper connectivity). Another handles the intelligent layer (Wi-Fi, cellular, switching, security). The third handles the access layer (the equipment that service providers use to deliver broadband to homes and businesses).

This three-layer architecture mirrors exactly how data center and enterprise network spending is evolving. Physical connectivity is the foundation. Intelligent networking sits on top. Access extends the reach. The company that controls all three layers has a cross-selling advantage that pure-play competitors in any single layer cannot match.

And yet, despite sitting at the center of multiple secular spending cycles, this company trades at a valuation that suggests the market still sees a legacy telecom vendor, not an AI infrastructure beneficiary.