The Hidden Race in Artificial Intelligence

By Stephen Hester, Analyst, Wide Moat Research

What are the key companies involved in the artificial-intelligence (“AI”) revolution?

I asked two friends that question last week at a long-overdue happy hour. It’s a local place in Austin, Texas called Benders. I know, the name says it all.

Nvidia (NVDA) came up immediately. Then they talked about Open AI’s ChatGPT. My tech-savvy friend mentioned Tesla’s (TSLA) Grok and his new favorite Perplexity AI. Now, they are both in their 40s and use this technology more than most. One even has a job applying AI systems to a tech company.

They also manage their own portfolios and trade stocks, so they follow financial news closely. But they were clueless when it came to 95% of the companies powering the AI renaissance. I’m willing to bet the people you know are the same.

When most investors think of “investing in AI,” they mostly think of companies like Nvidia or Tesla. A few might be familiar with the other Big Tech firms’ AI ambitions… but that’s about it.

What few understand is that there is an entire ecosystem of lesser-known, but still high-quality, companies that provide the infrastructure for the AI models consumers use. And I would argue that this collection of supporting players will be far more valuable than any single AI model.

And for serious investors, these are the names worth knowing…

Like Buying a Car

You see, most people look at AI like buying a car. They make decisions based on brand names, what’s popular or cool, and how it feels to drive. They don’t understand the difference between a twin-turbo V6 or a supercharged V8, much less multi-point versus sequential port fuel injection. And they really don’t know that BMW, Audi, Jaguar, Land Rover, Chrysler, Dodge, Jeep, Aston Martin, Bentley, and even Rolls-Royce all use an almost identical transmission sourced from ZF Friedrichshafen in Germany.

The reason is simple: As a regular consumer, you can get away with not knowing.

But you and I aren’t seeking “consumer” portfolio performance. To achieve Wall Street returns, we need to understand the core components and who makes them. I believe that the infrastructure behind the front-facing AI systems like ChatGPT will end up many times more valuable, and make a lot more investors wealthy. And, at least for now, the public has no idea what’s behind the scenes.

Let’s stick with the car metaphor. The “engine” is the graphics processing unit (“GPU”), like the ones produced by Nvidia, and if you want a high-performing V12 like a Ferrari has, you need Nvidia’s top-shelf products like the H100 and A100. Sure, you can get a nice four-cylinder from Intel (INTC), but you won’t be winning the AI race.

Supermicro (SMCI) builds the chassis. That’s the metal frame that holds everything together. Their high-density, AI-ready server systems are tough to beat. Dell (DELL) and Hewlett Packard (HPE) are also trying hard to gain traction in this space.

You don’t even need a driver’s license to know cars require fuel. For AI systems, gasoline is silicon and memory. Taiwan Semiconductor Manufacturing (TSM) produces Nvidia’s and AMD’s most advanced chips. Without TSM, the car’s engine is dead.

ASML (ASML) is similar in that it makes specialized lithography machines used by firms like TSM in their manufacturing process. It’s proprietary technology that’s almost impossible to replace at least for now.

Applied Materials (AMAT) builds the tools that are used by TSM to assemble their GPUs. They are another irreplaceable link in the AI chain. Several of these companies were recently included in my put selling service, Intelligent Options Advisor. Every single one of them has been a winner, and some of these trades were the best in our multi-year history.

In your car, the transmission transmits energy between components so that it can move. That’s where Marvell Technology (MRVL), Broadcom (AVGO), and Arista Networks (ANET) come in. These firms make products that help data move between GPUs, servers, and anywhere else it needs to go. Arista Networks is debt free, has the most exposure to AI, and has the highest growth potential, so it’s my pick in this category.

Vertiv (VRT) and Schneider Electric (SBGSF) handle electrical distribution in AI centers. Think of this as the cooling system in a car. Ever run low on coolant, or worse, had a hose come loose? Then you know exactly how critical these are. They don’t have the upside of some of the other names, but they are also less risky.

The last thing I’ll mention doesn’t have to do with the car itself, but no less important. Companies like Digital Realty (DLR) and Equinix (EQIX) are the highways of AI. This is the physical infrastructure needed so that cars can run reliably, quickly, and safely across vast networks.

Conclusion

OpenAI and its peers are getting all the headlines. And I understand why. But the AI boom I suggest paying closer attention to are the firms behind the scenes making it all possible.

Valuations are infinitely better. Many of the AI infrastructure companies I mentioned trade for just 15 to 25 times forward earnings and are already highly profitable with strong balance sheets. Taken together, AI infrastructure is cheaper than the S&P 500 Index based on price-to-earnings ratios. Which do you think will grow faster in the coming years?

I know my answer.

Regards,

Stephen Hester
Analyst, Wide Moat Research