The AI Energy Economy — Part 2.1: Companies That Connect the Grid

The AI Energy Economy — Part 2.1: Companies That Connect the Grid

In Part 2 of this series, we focused on the pick-and-shovel companies building the physical backbone of AI electrification: turbines, transformers, substations, transmission lines, grid-scale batteries, and the construction required to deploy them. These firms profit whenever utilities and hyperscalers invest in more electrical capacity, regardless of which energy source ultimately produces the power.

But as AI demand scales, another reality becomes clear: electricity systems don’t just need more hardware — they need coherence.

Power must be generated, transmitted, conditioned, distributed, stabilized, and controlled, all at the same time. At AI scale, the seams between those layers become just as important as the layers themselves. A failure, bottleneck, or inefficiency at any one of those junctions can limit the usefulness of power everywhere else. That is where the companies in this addendum come in.

Part 2.1 focuses on companies that connect the electrification stack. These firms become increasingly valuable as the system grows larger, denser, and more complex — and as AI pushes grids closer to their physical and operational limits.

GE Vernova (GEV): The Bridge Company of AI Electrification

GE Vernova doesn’t belong in just one part of this series. It doesn’t own power plants and doesn’t sell electricity. Instead, it operates across nearly every physical layer of the power system, supplying the equipment and systems that allow electricity to be produced, moved, and managed at scale.

GE Vernova builds gas turbines used for firm and fast-ramping power; nuclear reactor technology and services (via GE Hitachi); grid equipment such as transformers and substations; high-voltage transmission components and power electronics; and grid-management software used to balance increasingly complex power flows.

This makes GEV adjacent to Part 2 (grid hardware) in this series, Part 3 (grid intelligence and controls), and indirectly relevant to Part 4, because the value of its equipment rises as power becomes scarcer and more strategically priced.

In the AI era, this matters enormously. AI demand doesn’t stress just one part of the grid — it stresses all of them at once. New data centers increase demand for generation, transmission capacity, voltage control, and real-time grid management simultaneously. GE Vernova benefits whenever utilities add generation, expand transmission corridors, experience grid congestion, and require more sophisticated control systems to keep everything stable.

Investment view:

GE Vernova is not cheap, but its valuation is reasonable relative to its order backlog, improving margins, and strategic positioning across multiple layers of the power system. It is one of the few companies that benefits from AI-driven electrification almost regardless of where the bottleneck emerges.

Bottom line:

GEV is a Buy as a cornerstone, system-wide electrification play.

Prysmian Group (PRYMY): The Arteries of the AI Grid

Prysmian focuses on moving electricity itself. It is the global leader in high-voltage and HVDC cabling — the specialized cables required for long-distance transmission, inter-regional grid connections, offshore and underground power routes, and virtually every new HVDC project worldwide.

As AI data centers scale, power increasingly has to travel farther — from nuclear plants, wind farms, and gas hubs to data-center clusters that often sit far from generation. In many cases, the limiting factor isn’t how much electricity can be generated, but whether that electricity can physically get to where it is needed.

That is Prysmian’s domain.

Cables are capital-intensive, long-lead-time products with limited substitution. They are custom-engineered, difficult to manufacture at scale, and slow to permit and install. As a result, they often become the pacing item in grid expansion projects, which gives Prysmian pricing power, strong backlog visibility, and strategic importance in national and regional grid upgrades.

Investment view:

Prysmian benefits from the same AI-driven electrification forces as the rest of Part 2, but with even stronger scarcity dynamics. There are very few global competitors capable of executing HVDC cable projects at this scale.

Bottom line:

Prysmian is a Buy for investors who want exposure to the physical constraints of AI electrification.

nVent Electric (NVT): Where Power Becomes Usable

As electricity volumes grow, so does the complexity of distributing and protecting that power inside data centers, substations, and industrial facilities. That complexity is nVent’s opportunity.

nVent sits downstream of transmission and generation, capturing value as projects move from concept to execution. The company specializes in electrical enclosures, busbars and power distribution systems, thermal management and protection hardware, and modular infrastructure that allows electrical systems to scale safely and quickly.

These components rarely make headlines, but they are essential. At AI scale, power density rises sharply — meaning more electricity flowing through smaller spaces. That creates more heat, higher fault risk, and greater consequences if something goes wrong. You cannot simply “plug in” a gigawatt-scale data center without massive internal electrical infrastructure designed to distribute, protect, and manage that power reliably.

Investment view:

nVent offers cleaner margins, less cyclicality, and more direct exposure to data-center build-outs than many traditional industrial companies. It is often overlooked precisely because it operates behind the scenes rather than at the headline level.

Bottom line:

nVent is a Buy as a quietly compounding AI-infrastructure play.

Amphenol (APH): The Connection Point Inside AI Infrastructure

As electricity and data move closer to their point of use, another layer of infrastructure becomes critical: the physical connections inside equipment itself. That is where Amphenol operates.

Amphenol designs and manufactures connectors — the specialized components that allow power and data to move safely and reliably between machines, racks, sensors, cooling systems, and control equipment. In an AI data center, these connectors sit everywhere electricity and information change direction, change format, or enter a new piece of hardware.

This matters because AI dramatically increases power density. More electricity is flowing into smaller spaces, at higher temperatures, with far less tolerance for failure. A loose, overheated, or unreliable connection doesn’t just cause downtime — it can take entire systems offline.

Amphenol’s products are engineered for exactly this environment. They are built to handle high power, high heat, and continuous operation while maintaining precise electrical and data performance. As AI data centers grow larger and more complex, the number and importance of these connection points increases.

Amphenol sits at the final step — making sure that all the power, cooling, and automation infrastructure supplied by companies like GE Vernova, Prysmian, and nVent actually works once it reaches the equipment level.

Investment view:

Amphenol is a high-quality industrial compounder with strong exposure to AI data centers and automation. Its business is less directly tied to grid build-outs than other names in Part 2.1, but it benefits from rising power density and system complexity inside data centers.

Bottom line:

Amphenol is a Hold — a durable, well-run company that complements the AI-energy theme, even if it lacks the same torque as grid-level infrastructure providers.

Why These Companies Matter

The companies highlighted in Part 2.1 illustrate a central reality of the AI energy transition: AI stresses the entire electricity system at once, not just one layer of it.

In this series:

Part 1 explains who generates electricity.

Part 2 explains who builds the grid.

Part 3 explains who automates and stabilizes it.

Part 4 explains how power is sold, priced, and monetized.

GE Vernova, Prysmian, nVent, and Amphenol don’t fit neatly into any single one of those categories. They are not power producers, and they are not utilities. Instead, they operate between the major layers of the electricity system, enabling those layers to function together as AI demand scales.

Each company addresses a different point in the journey electricity takes from generation to use. GE Vernova connects entire power systems, supplying the equipment and controls that allow generation, transmission, and grid operations to scale together. Prysmian connects regions, making it possible to move large amounts of electricity over long distances to where AI demand is concentrated. nVent connects facilities, ensuring that power entering substations and data centers can be distributed, protected, and managed safely inside those buildings. Amphenol connects components, providing the final physical links that allow power, data, cooling systems, and controls to function reliably inside machines and racks.

In practice, that means these companies participate in both Part 2 and Part 3 of the series. They help expand the grid by supplying essential hardware, and they help make that expanded grid usable by enabling power flow, protection, reliability, and coordination inside increasingly dense and complex environments.

As AI pushes electricity systems to become larger, more concentrated, and more unforgiving of failure, the value of these connective layers increases. Complexity is no longer a side effect of electrification — it is the defining feature of AI-driven electrification. And these companies are positioned precisely where that complexity turns into long-term demand.

Conclusion: The Value of the Connective Layer

As AI pushes electricity demand to levels never seen before, the most valuable companies may not be those producing the electrons, but those making the system work under stress.

Part 2.1 fills in that missing layer. Together, the companies of Part 2 and Part 2.1 form the physical and operational backbone that allows AI, electrification, and power markets to scale at the same time.

In an AI-driven energy system, connection is value — and these companies are positioned squarely at that intersection.

• The AI Energy Economy Series:

  • The AI Energy Economy — Part 1: The Nuclear & Utility Winners of the AI Power Boom
  • The AI Energy Economy — Part 2: The Pick-and-Shovel Suppliers Powering AI Electrification
  • The AI Energy Economy — Part 2.1: Companies That Connect the Grid
  • The AI Energy Economy — Part 3: Industrial Automation, Cooling & Controls
  • The AI Energy Economy — Part 4: Merchant Power, Nuclear Scarcity, and AI Contracts