In partnership with
Start investing right from your phone
Jumping into the stock market might seem intimidating with all its ups and downs, but it’s actually easier than you think. Today’s online brokerages make it simple to buy and trade stocks, ETFs, and options right from your phone or laptop. Many even connect you with experts who can guide you along the way, so you don’t have to figure it all out alone. Get started by opening an account from Money’s list of the Best Online Stock Brokers and start investing with confidence today.
Nvidia's $57B Quarter: Why the AI Bubble Fears Just Got Crushed
Listen, if you've been hearing whispers about an "AI bubble" ready to pop, Nvidia just walked into the room and said: not today.
On November 19, 2025, Nvidia dropped earnings that didn't just beat expectations—they obliterated them. We're talking $57 billion in quarterly revenue, up 62% year-over-year, crushing Wall Street's $54.9 billion estimate by a cool $2 billion. But here's the kicker: they guided Q4 revenue to $65 billion, roughly $3-4 billion ahead of what analysts expected.
That's not a company seeing demand slow down. That's a company that can't make chips fast enough.
The Data Center Dominance
Let's break down where this money is actually coming from. Nvidia's data center business—the part selling AI chips to companies building the future—generated $51.2 billion. That's 90% of their total revenue and up 66% from last year.
The star of the show? Blackwell. CEO Jensen Huang put it simply: "Blackwell sales are off the charts, and cloud GPUs are sold out." Not "doing well." Not "exceeding expectations." Sold. Out.
And here's what matters: Nvidia has $500 billion in cumulative orders for Blackwell and their next-gen Rubin architecture stretching through the end of 2026. That's not a backlog—that's a multi-year revenue visibility that most companies would kill for. CFO Colette Kress noted these orders span "every market"—cloud providers, sovereign nations, enterprises, supercomputing centers. Everyone wants in.
Three Massive Shifts (Or: Why This Isn't Slowing Down)
Huang framed something crucial on the earnings call. He's not just selling chips for today's AI models. He's positioning Nvidia at the center of three simultaneous platform shifts:
The move from general-purpose to accelerated computing – Think of this as the entire computing industry rewiring itself around GPUs instead of traditional CPUs
Generative AI replacing old machine learning – Every search engine, recommendation system, and content moderation tool getting rebuilt from scratch
Agentic and physical AI – AI that actually does things in the real world, not just responds to prompts
This is Huang's counter to the bubble narrative: we're not in a hype cycle, we're in the early innings of a fundamental infrastructure transition. And Nvidia has a 20-year head start.
The Margins Tell the Real Story
Want to know if a company has genuine pricing power or is just riding a wave? Look at the margins.
Nvidia's gross margins came in at 73.6% (non-GAAP). That's world-class profitability at unprecedented scale. But here's what's wild: they guided Q4 margins to 75%. Margins are expanding as they scale, not compressing. That's the opposite of what happens when commoditization sets in.
Translation: Nvidia isn't just selling picks and shovels in a gold rush. They're selling nuclear reactors, and nobody else knows how to build them.
The China Problem (And Opportunity)
Not everything is perfect. Nvidia's guidance assumes zero China revenue—their H20 chips cleared for export generated just $50 million in Q3, well below expectations. CFO Kress noted that "sizable purchase orders never materialized due to geopolitical issues."
But she also flagged $2-5 billion in potential upside if those tensions ease. That's not built into current guidance, which means the $65 billion Q4 target is conservative.
What This Actually Means
The stock jumped 4-5% after-hours, and here's why investors should pay attention: Nvidia just validated that hyperscalers (Microsoft, Amazon, Google, Meta) aren't pulling back on AI infrastructure spending. They're accelerating.
When Nvidia says they expect $3-4 trillion in total AI infrastructure spending by 2030, they're not making it up. They're literally taking the orders.
The AI skeptics keep waiting for the moment when companies realize they've overspent. Nvidia's Q3 just proved that moment isn't here—and might not be coming anytime soon.
The Bottom Line: With gross margins expanding, a $500 billion order book, and guidance that keeps beating expectations, Nvidia isn't showing cracks in the AI story. They're showing us what happens when you're the sole provider of the infrastructure everyone needs to build the future. The question isn't whether AI is a bubble. The question is whether anyone can catch Nvidia before 2030.