AI Demand: A Trillion Dollars and Some Servers

by

Jensen Huang, CEO of Nvidia, recently declared that demand for his company’s chips – Blackwell and Rubin, to be precise – is hovering somewhere around the $1 trillion mark through 2027. A trillion dollars. It’s a number so large it almost ceases to be meaningful, like trying to count all the grains of sand on a particularly ambitious beach. (And let’s be honest, who has the time? There’s tea to be drunk, and the universe isn’t going to observe itself, you know.) This is a substantial uptick from the merely impressive $500 billion he mentioned a year ago. The implications, for those of us tethered to the somewhat linear progression of financial markets, are… noteworthy.

Valuations in the AI infrastructure space have, shall we say, experienced a period of adjustment. A polite way of saying “come down a bit.” But Huang’s message remains stubbornly optimistic. AI isn’t going anywhere. It’s like a particularly persistent houseguest. And three companies, in particular, seem poised to benefit. Nvidia, naturally, is the obvious beneficiary. But Dell Technologies and Amazon also received a mention during Huang’s keynote. Let’s unpack that, shall we? (Unpacking is a metaphor, of course. We’re not actually dismantling anything. Though, given the complexity of modern finance, perhaps we should be.)

Nvidia: The Chip Maestro

Huang’s insights aren’t based on wishful thinking. He’s privy to purchase orders from the top AI companies, researchers, and even sovereign nations. (One imagines slightly stern-faced officials demanding ever-more-powerful chips to, well, do whatever it is sovereign nations do with that sort of thing.) This suggests significant growth potential for Nvidia. They supply the essential building blocks for AI data centers – not just GPUs, but also networking components and, crucially, the CUDA software platform. CUDA, you see, is Nvidia’s secret sauce. It allows customers to tailor the GPU to specific tasks, like training large language models. (Think of it as a highly sophisticated Lego set for artificial intelligence. Except, instead of plastic bricks, it’s silicon and algorithms.)

This advantage translates into high margins. Last year, revenue soared 65% to $216 billion, generating a profit of $120 billion. Impressive, even by the standards of companies that routinely deal with numbers of that magnitude. However, growth isn’t guaranteed. Competition from rival chipmakers is a constant threat. (The chip industry is a bit like a galactic empire, constantly vying for dominance.) While Nvidia is currently in a strong position, any pressure on revenue in its data center business could limit future gains. Still, hedge funds seem to agree with the bullish outlook. The stock currently trades at 22 times this year’s earnings and 17 times next year’s, suggesting there’s still room for upward movement.

Dell Technologies: The Server Specialist

All those Nvidia GPUs need somewhere to live, don’t they? Enter Dell Technologies, the world’s leading supplier of servers. More GPUs mean more server racks, which translates into more sales for Dell. It’s a surprisingly straightforward equation, really. (Though, in the world of finance, simplicity is often a sign that you’ve missed something.) The stock trades at a relatively modest 12 times forward earnings, making it an interesting proposition.

Dell’s business is divided into infrastructure solutions (servers, storage, networking) and client solutions (PCs). The PC side has been… let’s say “under the weather” in recent years. But the infrastructure segment is booming, surging 40% year-over-year to $61 billion in revenue. Huang specifically highlighted the partnership between Dell and Palantir Technologies during his keynote. Nvidia chips power Dell’s AI Factory and form the backbone of Palantir’s AI operating system for sovereign and enterprise clients. This collaboration demonstrates Dell’s vital role in the broader AI ecosystem. Dell’s AI business is exploding, with AI-optimized server revenue up 342% year-over-year in the fourth quarter, reaching $9 billion. Analysts predict earnings growth of 15% annually in the coming years.

Loading widget...

Amazon: The Cloud Colossus

Amazon, at this point, seems almost unstoppable. It serves hundreds of millions of customers through its online retail store, but it has also developed high-growth, high-margin revenue streams in advertising and cloud computing. (It’s a bit like a digital octopus, tentacles reaching into every corner of the market.) On the AI infrastructure front, the growth engine is Amazon Web Services (AWS), the leading enterprise cloud platform. AWS growth accelerated last year, with revenue up 24% year-over-year in the fourth quarter.

Interestingly, AWS deliberately left some revenue on the table in 2025, as demand for AI services exceeded its data center capacity. (A rather unusual situation, akin to a bakery refusing to sell bread because it’s too popular.) As Amazon invests in expanding compute capacity, this constraint should ease, potentially supporting stronger growth than investors currently anticipate. Huang mentioned that OpenAI’s recent partnership with AWS could drive “enormous consumption” of cloud computing. OpenAI selected AWS as its exclusive cloud provider for its Frontier enterprise platform, helping companies build, deploy, and manage AI agents. This could drive sustained usage-based cloud spending, serving as a catalyst for Amazon’s growth since AWS generates roughly half of the company’s profits. Analysts expect earnings to grow 18% annually in the coming years. Assuming Amazon performs in line with those estimates, the stock should be a rewarding investment for patient investors. It’s currently trading at the lowest multiple of operating cash flow in over a decade.

Loading widget...

Read More

2026-03-24 07:02