Every preview of Dell Technologies’ fiscal second quarter, which the company reports after the close on Tuesday with a call at 3:30 p.m. Central, is built around one question: does the AI server number beat. That is close to the wrong question, and Dell answered it three months ago. Management guided this quarter to roughly $15.5 billion of AI server revenue, which is below the $16.1 billion it recognized in the first quarter, and it issued that guide while carrying an order book it cannot convert and telling investors it is short of memory, standard processors and hard drives for the back half of the year. Almost nobody is treating a guided sequential decline as the story. It is the story. Dell’s AI business has stopped being limited by how much customers want and started being limited by what its suppliers will sell it, and at what price.
The Numbers Dell Already Handed Over
The first quarter was the best in the company’s modern history. Revenue reached $43.8 billion, up 88% year over year, and AI server revenue hit $16.1 billion, a 757% increase. CNBC reported that Dell posted its fastest sales growth since returning to public markets in 2018, and the market responded in kind: the stock closed 32.76% higher the following day, its best single session ever. Management booked $24.4 billion of fresh AI orders in the quarter and lifted the full-year AI server target to $60 billion from $50 billion.
Then it guided the second quarter to total revenue of $44.0 billion to $45.0 billion, up about 49% at the midpoint, with non-GAAP earnings of $4.70 to $4.90 per share. And roughly $15.5 billion of AI servers. Futurum’s breakdown of that guidance treats the raised full-year figure as the headline, which is fair. But the quarterly path inside it is the part worth reading twice. A company that just grew a product line 757% told the street the next three months would be slightly smaller.
Demand Is Not the Variable Anymore
Dell has been unusually candid about the reason, which makes the silence around it stranger. Vice chairman and chief operating officer Jeff Clarke described an inflationary environment across memory and other components and said the company has been adjusting prices frequently in response. Dell has flagged commodity constraints specifically in DRAM and NAND, and has told investors to expect supply pressure in the second half of the fiscal year across memory, standard compute and storage.
That is a supplier problem wearing a demand-story costume. An AI server is mostly other people’s silicon: Nvidia’s accelerators, somebody’s DRAM, somebody’s NAND. Dell integrates, validates, finances, ships and services it. When the scarce input reprices, the integrator absorbs it or passes it on, and Dell has been doing both.
A guided sequential decline in the middle of the largest infrastructure buildout in the industry’s history is not a demand signal. It is a procurement result.
We covered the other end of this pipe last week, when Nvidia moved to raise AI server prices by more than 15% and pointed at the memory makers as the reason. Read the two together and Dell’s position gets uncomfortable. Its largest supplier is raising prices because its input costs are rising. Dell’s own memory costs are rising directly. Dell sells into hyperscalers and neoclouds with procurement teams that negotiate hard and have alternatives.
Mid-Single Digits Is the Whole Argument
Here is the figure that should govern how anyone reads Tuesday’s release. Dell has said its AI server business is tracking to a mid-single-digit operating margin. Not gross margin. Operating margin, on the fastest-growing, most-discussed product line in enterprise hardware.
Put that against the rest of the company. Dell’s client solutions group, the PC business that analysts have spent a decade writing off, has historically run at operating margins in a similar band or better, on revenue that requires no $60 billion supply commitment and no exposure to DRAM spot pricing. The AI server line is bigger, louder and growing at triple digits. Per dollar of revenue, it is not obviously better business.
That is the gap between how Dell is being valued and how Dell actually earns. The market repriced this company on a revenue line. The revenue line converts at mid single digits and its cost base is set by suppliers with more pricing power than Dell has. Three months ago, when we previewed the first quarter against the same margin question, the backlog was the reassuring number. It is less reassuring now that Dell has said out loud it cannot get the parts to work it down.
What We Actually Think
Dell should stop leading with the AI revenue number, and investors should stop rewarding it. A company guiding a product line down sequentially while its input costs inflate and its suppliers raise prices is not in a demand story, and continuing to headline the $60 billion target invites a repricing later that a clearer disclosure would avoid now. The honest presentation is AI operating profit, quarter by quarter, alongside the memory cost assumption underneath it. Dell has the number. It has told us the margin band. It simply does not report the line, and the absence is doing work.
The Motley Fool made the narrower version of this point on Monday, framing infrastructure operating margin as the metric to watch. We would go further. Watch the sequential AI server revenue figure against that $15.5 billion guide, and treat a beat as a question rather than an answer. If Dell shipped more than it guided into a market it says is memory constrained, it either paid up for the parts or priced aggressively to move them. Both show up in margin, not in the line everyone is watching.
The backlog was never the risk. The bill of materials is.