On Wednesday afternoon, Jensen Huang looked directly at the market's favorite worry and shrugged it off. “There's been a lot of talk about the AI bubble. From our perspective, we're seeing something very different,” Nvidia's CEO told analysts during its quarterly earnings call, touting quarterly revenue of $57 billion, guidance of $65 billion and sold-out cloud GPUs. By the time he hung up, the stock had risen more than 6% in after-hours trading, with analyst emails pouring into inboxes with subjects like “AI Bubble Who?” with notes that they had raised their earnings forecast (again), and even the people who were paid to worry about surplus livelihoods seemed strangely relieved to let Juan win the argument – at least for one more quarter.
Thursday came and my mood took a turn for the worse. Nvidia shares rose sharply to open the day and were down nearly 3% midday, with the S&P 500 and Dow Jones Industrial Average falling along with it. Nvidia may have produced numbers that make any talk of an AI bubble seem frivolous, but the market is still acting as if something under the surface isn't settled at all.
How did the numbers come about?
On paper, Nvidia delivered exactly what nervous AI investors needed: a monster quarter. Revenue jumped 62% year over year. Sales of data centers, the heart of the artificial intelligence boom, were about $51.2 billion. Gross profit increased to 75%. Management raised $4 billion, beating Wall Street consensus expectations. No analyst seemed surprised by another strong quarter from a company that continues to produce nothing. Wedbush analyst Dan Ives called it another “monster quarter from the godfather of artificial intelligence” (aka Huang), raising his target to $230 and saying Blackwell and Rubin's $500 billion in revenue through 2026 now looks conservative. BNP Paribas raised its profit forecasts and said fears of an “AI infrastructure bubble” appeared overblown.
Inside the call the subtext was louder. Nvidia has confirmed that its supply is effectively locked in for the next two years. CFO Colette Kress pointed to a sequential 63% jump in delivery commitments—an absurd number for a typical hardware cycle, but not unreasonable in an industry where every hyperscaler is vying for space, capacity and delivery windows. Analysts who cover the broader hardware stack have made it clear: TSMC remains crowded, component prices remain tight, and companies selling cables, optics, memory and power infrastructure continue to follow Nvidia.
The sentiment among salespeople was the same: Whatever you think about artificial intelligence, the growth itself is not imaginary.
David Wagner, portfolio manager and head of equities at Aptus Capital Advisors, said Nvidia's earnings provide “more to the market than [they do] for stocks,” because the last few weeks have been one long swing – the technology disruption, the hype cycle sagging under its own weight. Nvidia's quarter didn't just beat the numbers; it told him that the artificial intelligence investment boom was still going strong. “Let's be realistic,” he said. “There are good bubbles and bad bubbles. So maybe there's a pocket here? Absolutely. Is it sometime in the near future? Probably not.” In his opinion, this is a macro story, not a micro story, and “this capex train will continue to go down.”
“There is so much anxiety in this market,” he said. “Nvidia continues to prove the naysayers wrong.”
“It's hard to say there's a bubble,” said Motley Fool senior investment analyst David Meyer. But he also said no one yet knows the true return on investment in AI at Nvidia's scale. “Anyone who buys these chips should get a return on their investment, right? And we won't know. Right now, no one is telling us how much extra revenue they're getting from AI.” Still, he said, the demand is real and investment dollars are as real as they come.
The problem is that the AI bubble means two completely different things, and Nvidia may have only killed one of them. The narrative that Huang rejected—the idea that demand is peaking or that earnings are illusory—doesn't hold up in a quarter like this. Another theory that makes investors wince at every macro swing is whether construction can sustain the pace: energy pressures, a hyperscaler arms race, funding cycles, a handful of companies holding back all trade.
By Thursday afternoon, all the excitement surrounding Nvidia had faded somewhat, and the market had given up much of its early gains and gone into the red. Leveraged ETFs that had been lining up to take advantage of Nvidia's “champagne pop” moment, as Wedbush put it, suddenly found themselves focusing on volatility instead. Analysts at Direxion, which manages some of these funds, acknowledged as much, describing Nvidia's earnings as “inflated to a level that suggests the broader market is riding on their success” before turning around and reminding clients that its products could face cuts in both directions if volatility remains high.
So even after a monster, amazing, sensational—or whatever word analysts want to use for “huge”—quarter, concerns about AI remain. The same notes and analysts who praised Nvidia also noted that valuations for the overall AI complex are still high, the circular economy of AI is a big concern, the Fed remains hawkish, and the rest of the economy is still showing late-cycle swings. Nvidia may have bought more time to trade artificial intelligence, but that clearly hasn't cured the overall uneasiness of a jittery market.
What does tomorrow depend on?
Nvidia's numbers are solid today, but that doesn't mean everyone stops worrying about tomorrow.
It's easy to forget how much Nvidia's future still depends on things it doesn't control. The clearest thing is electricity. Today's grid is not built for what Nvidia, hyperscalers, and graphics cloud providers promise. Yvette Schmitter, co-founder and CEO of IT consulting firm Fusion Collective, noted that Nvidia is “selling futures contracts backed by electricity that won't flow through transmission lines for another 3 to 7 years.” She's not wrong. Between permits, substations and transmission scrambles, the infrastructure underpinning this boom is moving at the speed of politics rather than the speed of semiconductors.
Kevin Cook, senior equity strategist at Zacks Investment Research, said: “You can't keep scaling models without scaling electricity; This is the bottleneck.” He added: “The demand curve for computing is vertical, but the grid is not.” However, he said Nvidia gives investors the clearest view of AI demand globally and that “this package will add hundreds of basis points to GDP.”
Huang is not wrong when he says he sees something completely different from the AI bubble. The demand he describes is real—in the present tense. Hyperscalers continue to increase data center capital expenditures. Network and ancillary products grow rather than delay results. China, once a major engine of growth, remains closed to cutting-edge chips, and without it Nvidia's sales in all other countries have roughly doubled. And the bubble talk lives in that gap: everyone can see that Nvidia's numbers are real, and everyone can see how fragile it is to have such strong index-level performance dependent on the same small circle of companies.
Even in the purest version of this world, where demand for AI continues to grow and its adoption expands, its shortcomings are obvious. Tier 2 GPU lessors are taking out loans to keep up with hyperscalers. Utilities are trying to greenlight large industrial loads. Regulators are considering focusing on both cloud markets and chip pipelines. Energy-intensive deployment schedules are concentrated in networks that are not designed for it. None of these issues contradict Nvidia's numbers. They sharpen them. Demand is so strong, so compressed and so loaded that the weakest links, not Nvidia, will determine how long this cycle can continue at its current pace.
And this is why the market both loves and fears Nvidia. This is the star of the story, the strongest hand in the trade and a signal that the entire index is trading. But it's also a reminder that this era is built on infrastructure, cash flows and physical constraints that don't care about hype cycles. Juan can brush off the bubble questions because his quarter gives him the right to do so. Investors can agree with him because the numbers allow them to. None of them can guarantee that the world around these numbers will hold up.
Nvidia's decision didn't refute the bubble argument. It just pushed him further away. That's enough for now.
