[你懂的]  AI Isn’t Just About GPUs: Meet the Company Building the Power Behind AI

MYR Group (MYRG).

It isn’t a chipmaker or a data-center operator. MYRG provides transmission, distribution, and electrical construction services for commercial, industrial, utility, and infrastructure projects.

In simple terms:

AI data centers need enormous amounts of electricity — and MYRG helps build the infrastructure that delivers it.

1. How does MYRG actually make money?

MYRG has two major business segments.

Transmission & Distribution (T&D) focuses on electric transmission and distribution infrastructure.

As the U.S. adds data centers, industrial facilities, EVs, and other power-intensive projects, the grid needs to expand and modernize.

The second is Commercial & Industrial (C&I), which includes electrical construction for large commercial and industrial facilities, including data-center-related projects.

So MYRG isn't selling electricity.

It is getting paid to build the electrical infrastructure needed to deliver it.

2. Is it actually profitable?

This is where MYRG gets interesting.

In 2025, the company generated approximately $3.66 billion in revenue and $118 million in net income, with EPS reaching $7.53.

More importantly, growth accelerated in 2026.

In Q2 2026, revenue reached approximately $1.08 billion, up around 20% year over year.

Net income reached roughly $50 million, compared with $26.5 million a year earlier.

That means revenue grew strongly, while profit grew even faster.

For me, that is much more important than simply seeing “AI” in a company's presentation.

The real question is:

Can the company turn rising demand into rising profits?

So far, MYRG is showing that it can.

3. The number I am watching most: Backlog

For an engineering and construction company, backlog is extremely important.

Think of backlog as projects that have already been awarded but have not yet been fully recognized as revenue.

MYRG's backlog has reached record levels.

That gives the company something many speculative AI stocks don't have:

visibility into future revenue.

This isn't simply a company announcing an AI partnership and hoping investors buy the story.

Customers are already spending money.

Projects are already being awarded.

And MYRG is already generating revenue and profits from the infrastructure cycle.

4. Why could the market eventually re-rate MYRG?

Because AI infrastructure is moving into its next phase.

First:

Not enough GPUs → buy more GPUs.

Then:

Not enough data centers → build more data centers.

And eventually:

Not enough electricity → upgrade the power infrastructure.

The chain becomes:

Power generation → transmission → substations → distribution → data centers.

MYRG sits toward the back end of that chain.

That is what makes the stock interesting to me.

AI is not its entire business.

AI is a potential accelerator for an existing business.

That's a much stronger setup than a company whose only investment thesis is an AI buzzword.

5. But there is a catch

I wouldn't call MYRG a “buy at any price” stock.

It is still an engineering and construction company.

Margins are nowhere near those of a high-margin semiconductor business.

Large infrastructure projects can also face cost overruns, delays, labor shortages, and execution risks.

And if interest rates remain elevated for longer, some infrastructure projects could be delayed.

So I wouldn't chase the stock simply because it has an AI connection.

Valuation still matters.

My Take

If I map out the AI infrastructure ecosystem:

NVDA = Compute

MU / SNDK = Memory

AVGO / MRVL = Connectivity

GLW = Fiber

And I would put:

MYRG = Power Infrastructure

into the same conversation.

MYRG may never become the next Nvidia.

But if the U.S. enters a multi-year cycle of AI data-center construction, grid expansion, and electricity infrastructure investment, the market could eventually realize that some of the biggest beneficiaries aren't the companies making the chips.

They are the companies making sure those chips actually have enough power to run.

That's why MYRG is on my watchlist.

High risk. High potential. Real revenue. Real profits. Growing backlog. And exposure to a structural AI infrastructure trend that could last for years.

The interesting question isn't always:

«“Who is building the next AI model?”»

Sometimes it's:

«“Who is building the infrastructure that keeps the AI machines running?”»

MYRG is one company worth watching.[思考]