Dell’s AI Server Revenue Jumps 757% to $16 Billion: The Picks-and-Shovels Winner of the AI Boom

While the debate rages over whether AI companies will ever justify their valuations, one group is already booking enormous, verifiable revenue from the boom: the companies that sell the servers. Dell Technologies reported that its AI-optimized server revenue reached $16.1 billion in a single quarter, up 757% from a year earlier, and raised its full-year AI server outlook to $60 billion. Rival Hewlett Packard Enterprise told a similar story days later, its chief executive describing "the strongest AI-server backlog we have ever seen." In the gold rush of the AI build-out, Dell and HPE are selling the shovels — and the shovels are selling out.

For a reader trying to find solid footing in the AI-bubble argument, these are among the hardest numbers available: proof that the spending is real and that at least one layer of the AI economy is converting it into actual profit today.

How Big Are the Numbers?

Dell's fiscal first-quarter results, for the period ended May 1, 2026, showed clearly where AI capital lands. Total revenue hit a record $43.8 billion, up 88% year over year, and AI-optimized servers — the systems built around Nvidia and AMD accelerators — drove the surge with the 757% jump to $16.1 billion. The company raised its full-year AI server target to $60 billion, up from a $50 billion projection just three months earlier and implying roughly 144% growth, and lifted its overall revenue outlook to about $167 billion. Investors responded with one of the best single-day stock moves in Dell's history.

HPE's results, reported shortly after, confirmed the pattern. Its Cloud and AI segment revenue rose to $7.7 billion, and crucially, profitability improved, with the segment's operating margin nearly doubling year over year. CEO Antonio Neri said the AI-server backlog exceeded $4.6 billion and that orders had more than doubled, producing a record company backlog; HPE raised its full-year outlook and pulled its long-term financial goals forward by two years. Both companies expect strong server demand well into 2027.

Why Is "Picks and Shovels" the Right Frame?

The phrase comes from the 19th-century gold rushes, where the people who reliably got rich were not the prospectors gambling on a strike but the merchants selling them picks, shovels, and supplies. The AI version is the same: while it remains genuinely uncertain which AI models or applications will win, the build-out requires a staggering amount of physical hardware regardless of who comes out on top — and Dell, HPE, and peers such as Supermicro are the ones supplying it.

That is what makes these results matter in the bubble debate. The skeptics' strongest argument is that hundreds of billions are being spent while clear returns remain elusive. Dell's and HPE's numbers are a partial rebuttal: the spending is unmistakably real, it is flowing into hardware now, and at least one link in the chain is turning it into growing revenue. When a hyperscaler commits billions to a data center, a large slice becomes a Dell or HPE purchase order — concrete demand on an income statement, not a promise about future profit.

Does the Revenue Actually Become Profit?

The bull case has an important asterisk, and it is about margins. AI-optimized servers are, to a large degree, a vehicle for Nvidia's chips. A rack-scale AI system's cost is dominated by the GPUs inside it, which Dell and HPE buy from Nvidia and pass through to customers with a relatively thin markup. The value they add — integrating those GPUs with networking, power, and increasingly complex liquid cooling, then deploying and servicing the systems at scale — is real, but it commands far lower gross margins than the software businesses investors are used to celebrating.

In practice, a company can post eye-popping AI server revenue growth while the profit reaching the bottom line grows far more modestly. That is why HPE's nearly doubled segment operating margin matters as much as its revenue: it hints that the shovel sellers may be gaining a little pricing power as demand outstrips supply, rather than just moving more boxes at razor-thin spreads. Whether that margin gain holds is one of the more important questions on the hardware side of the AI trade.

Why Does Demand Keep Outrunning Supply?

Both companies are sitting on record backlogs because they cannot build systems fast enough, and the bottlenecks sit upstream. The Nvidia accelerators that anchor these servers are supply-constrained, as is the high-bandwidth memory that feeds them and specialized parts such as the substrates that package the chips — several reportedly sold out into 2027. A backlog here is not weak sales; it is orders arriving faster than the supply chain can fulfill them.

That dynamic also gives the picks-and-shovels businesses unusually good visibility. A backlog measured in billions, booked well ahead of delivery, gives Dell and HPE a clearer line of sight into future revenue than most hardware companies ever get — a big part of why both felt confident enough to raise guidance and accelerate their targets.

What Does It Mean?

For anyone seeking solid ground in the AI debate, Dell's and HPE's results are among the firmest data points available. They do not settle whether the AI labs and applications will ultimately earn back the hundreds of billions being spent, but they prove the spending is real and the hardware layer is profiting from it today. The shovel sellers are the part of the AI economy where the money has already, demonstrably, arrived.

The caveat keeps the enthusiasm honest: this is a lower-margin business riding on Nvidia's silicon, tied to a capital-spending wave the bears warn cannot last forever. If hyperscaler capex ever slows, today's reassuring backlogs would thin out fast. But for now, in an industry full of promises about the future, Dell and HPE offer something rarer — a present-tense profit, booked in billions, from selling the tools everyone in the gold rush needs. This article is not investment advice.

Frequently Asked Questions

How much did Dell's AI server revenue grow?

Dell reported AI-optimized server revenue of $16.1 billion for its fiscal first quarter ended May 1, 2026, up 757% year over year. It raised its full-year AI server revenue target to $60 billion, up from a $50 billion projection three months earlier, and total quarterly revenue hit a record $43.8 billion.

How did HPE's AI server business do?

HPE's Cloud and AI segment revenue reached $7.7 billion, with operating margin nearly doubling year over year. CEO Antonio Neri cited an AI-server backlog above $4.6 billion and said orders more than doubled, producing a record company backlog. HPE raised its full-year outlook and accelerated its long-term goals by two years.

What does "picks and shovels" mean in the AI boom?

It refers to profiting from a boom by selling the tools everyone needs rather than betting on who wins. In AI, Dell, HPE, and Supermicro sell the servers that all the AI spending flows into, so they earn real revenue regardless of which AI models or applications ultimately succeed.

Why are AI server margins lower than software margins?

Most of an AI server's cost is the Nvidia GPUs inside it, which Dell and HPE buy and pass through to customers with a thin markup. Their profit comes from integration, cooling, deployment, and service — valuable work, but far lower-margin than software. So soaring AI server revenue does not translate one-to-one into profit. This article is not investment advice.