Hello, Reader.
Let’s jump right into it…
Nvidia Corp. (NVDA) released its much-anticipated second-quarter earnings after yesterday’s bell.
And while “better” seems to be an expectation for Nvidia in perpetuity, the company delivered better-than-expected results… for the most part.
For the quarter, Nvidiareported that total revenue increased 56% year-over-year to $46.7 billion. Earnings jumped 54.4% year-over-year to $1.05 per share, up from $0.68 per share in the same quarter a year ago.
However, the company’s outlook failed to impress Wall Street, and the stock slid nearly 3% immediately following its earnings release.
But Wall Street isn’t the only one colored “unimpressed” by the company.
I have been cautious for some time about Nvidia’s sky-high valuation, and where that could eventually lead the company… basically, back down to Earth.
That’s not “better” for anyone. And my wariness around the company hasn’t changed.
So, in today’s Smart Money, I’d like to explain why, despite its earnings beat and dip, I still don’t recommend Nvidia.
Instead, I’ll share a company to replace the chip king in your portfolio.
Let’s dive in…
Beware the Risks
Nvidia has been the undisputed golden child of the AI Revolution for the past two years.
But just because that is the company’s past does not mean it is a guaranteed future, especially since steep AI competition is starting to steal profits.
Nvidia used to keep $0.78 of profit on every $1 of sales. Now it’s down to $0.71… and falling fast.
That’s what happens when other members of the Magnificent Seven – like Amazon.com (AMZN), Alphabet Inc. (GOOG), Microsoft Corp. (MSFT), and Meta Platforms Inc. (META) – start to develop their own chips.
For example, Amazon Web Services (AWS) made its in-house Graviton4 chip generally available in July. This chip is one of many that come from Amazon’s Annapurna Labs in Austin.
Google designs and manufactures its own chips, called Tensor Processing Units (TPUs), for AI, data centers, and smartphones. In fact, Apple uses Google’s TPUs – while also making its own custom processors, known as “Apple Silicon.”
To round it out, Meta started developing and testing its own AI chips, known as Meta Training and Inference Accelerators (MTIA), earlier this year.
All of these chips were created to reduce dependency on Nvidia, which could eventually get cut out of the picture all together.
What’s more, Nvidia has had to write off $5.5 billion worth of chips that it can’t sell to China because of trade restrictions. Specifically, it still cannot sell its most advanced AI processors to China. That’s real money disappearing from its bottom line.
And with global tensions rising, more countries could be removed from Nvidia’s customer list.
So, I recommend avoiding the monumental risks here by steering clear of Nvidia all together.
But now you’re probably wondering, “What company could possibly outdo the AI king in my portfolio?”
Well, here’s what nobody’s talking about…
The Nvidia Replacement
While the larger focus remains on AI chips, there is one component that makes everything work. Without it, even the most powerful AI chip, including Nvidia’s, is just expensive silicon.
See, when you build an AI data center, thousands upon thousands of servers must be installed.
But here’s the kicker: Those servers are useless unless they can talk to and learn from each other. And the way they communicate is through fiber-optic cables.
New AI hyperscalers need 10X more cables than regular data centers. That’s enough fiber to circle the globe eight times – in a single facility.
To harness that growth, I’ve got a pick that fits squarely in the category of stock I love the most – overlooked and underhyped.
This company quietly built the backbone of the internet, while scores of companies boomed and busted.
And now, as AI explodes, it is a leading supplier of what every data center desperately needs. In fact, this company’s CEO recently said the company’s production of AI fiber is tripling every month.
High demand means customers are inking deals with the company to reserve product ahead of time to edge out competitors. Already, 80% of the AI fiber-optic cable this company makes over the next five years is spoken for.
And it is manufacturing most of it right here in America. This means virtually no tariffs and no trade restrictions for its U.S. customers.
Here’s the best part…
While Nvidia’s customers turn into competitors, nobody is trying to manufacture their own optical-fiber cables.
So, AI hyperscalers are all fighting to get more cables from this company, not replace them.
You can learn how to access my full research and all of this company’s details in my latest free special broadcast, and then you can add it to your portfolio immediately.
Regards,
Eric Fry
