Meet the AI Winner You Have NOT Heard Of

By Costas Bocelli ·

Yesterday, Bill showed you something important about the Artificial Intelligence boom.

The first AI fortunes were made in the chips. But as Bill showed you yesterday, the opportunity has gotten considerably wider.

And today I'm going to show you one stock that sits directly in the path of that expanding AI buildout.

I see about a +45% potential upside in the shares from here. And using the options strategy I'll show you below, you could potentially turn that move into an 80% gain or more.

Thumbnail

But to understand why I like this stock so much, you first have to understand what's happening to AI itself.

Basically, AI is escaping the confines of the "chip". I mean that, every new AI model... every company putting AI agents to work... every new application moving from experiment to everyday use...

...creates demand for an enormous amount of physical infrastructure needed to make all of that intelligence actually work.

Servers. Storage. Networking. Cooling. Data centers. And somebody has to take all those pieces and turn them into a functioning system.

That's where today's opportunity comes in. Because this week's company does considerably more than sell another component that goes inside the AI buildout.

It helps customers put the AI infrastructure itself together.

So now let's...

Meet the Company Building AI Factories

The company is Penguin Solutions (PENG).

And if that name doesn't immediately scream "Artificial Intelligence" to you, that's part of what makes this story interesting.

Penguin describes itself as an "AI Factory Platform Company." In plain English, it helps enterprises, governments and specialized cloud providers design, build, deploy and manage the enormously complicated computing infrastructure required to run AI at scale.

Think about what that means. Buying thousands of cutting-edge GPUs doesn't magically give you an AI data center.

Those GPUs have to be connected to high-speed networking and storage. The architecture has to be designed around the workloads it's expected to handle. The hardware has to be assembled, tested, installed and managed. Cooling has to work. And the entire system has to operate reliably once it goes live.

Penguin helps customers put all of those pieces together.

Its OriginAI infrastructure, for example, uses pre-defined architectures that can scale from 256 to more than 16,000 GPUs.

Across its business, Penguin says it has deployed and managed more than 99,000 GPUs representing more than 4 billion hours of GPU runtime.

That's why I think PENG fits so neatly with the idea Bill introduced yesterday. AI itself can spread at digital speed. AI infrastructure cannot.

And the faster companies race to put AI to work, the greater the need becomes for businesses that know how to turn chips, memory, networking, storage, cooling and software into an actual working AI factory.

And this isn't just a compelling AI story, because...

Demand Is Showing Up in the Numbers

In its most recently reported quarter, Penguin generated a record $479 million in sales, up 48% from a year earlier. Non-GAAP earnings per share climbed 79%, to $0.84.

Management also raised its full-year outlook, citing strong demand tied to agentic AI across its AI infrastructure and memory businesses.

Last month, Penguin announced a particularly interesting project with Lektra, a young AI infrastructure company that's building smaller data centers where energy is already available.

Rather than waiting years for another enormous centralized data center and its grid connections, Lektra distributes AI computing capacity across smaller sites. And it selected Penguin to help design, deploy and optimize that infrastructure.

The idea is to distribute AI computing infrastructure geographically, using existing carbon-free energy and helping overcome some of the centralized-grid constraints complicating the AI buildout.

That's the second-wave AI opportunity we're talking about. Nvidia and the other chipmakers helped make the AI revolution possible. Now somebody has to help the rest of the world build what AI requires.

And Penguin Solutions is right in the middle of that buildout.

So fundamentally, PENG is sitting squarely in the path of the AI buildout Bill described yesterday.

But a great story doesn't automatically make a great trade.

For that, we need to look at the stock.

PENG

As you can see, PENG is down nearly 40% from its record high. But the decline has brought the stock down to an important retracement level, where it has begun building a base of support.

With PENG trading around $55, I like the risk/reward from here. A break below $45 would undermine my technical thesis. So I'm looking at roughly $10 per share of downside before I'd want to reconsider the trade.

To the upside, however, a return to this summer's highs would give us roughly +45% profit potential.

One way to play it is simply to buy the shares directly. At $55 per share, a 100-share position would require about $5,500 in capital. But there's another way to establish a similar 100-share position while committing considerably less money.

The Options Play

Consider buying the PENG April 2027 $45 Call. At current prices, one contract costs roughly $1,900 and gives you the right to control 100 shares through April.

That's about 65% less capital than buying 100 PENG shares outright. And unlike owning the shares, the roughly $1,900 premium represents the maximum you can lose on the call.

And if PENG makes it back to its summertime highs, I estimate that call could deliver a gain of +80% or more.

That's the opportunity I see: approximately 45% potential in the stock with the possibility of magnifying that move to +80% or more using the call.

And if this approach to using options interests you, there's something else I want you to see.

My colleague Chris Rowe uses a variation of this strategy designed to systematically reduce the capital at risk in successful positions. He calls the master strategy CX90².

You can see how to put it to work here.

Riding the next wave of AI? You should be!

Costas