Nvidia just reported the best quarter in its history—$68.1 billion in revenue, up 73% year-over-year—and Wall Street barely blinked. The stock nudged up in after-hours trading, then cooled off. Why? Because the real story isn't what Nvidia earned. It's what they're not earning from China.

The company guided to $78 billion in revenue for the next quarter, crushing analyst expectations of $72.5 billion. Sounds great, right? Here's the catch: CFO Colette Kress said explicitly on the earnings call that this guidance assumes zero revenue from China's data center business.

Let me translate that: Nvidia's H200 chips—the ones designed to comply with U.S. export restrictions—were approved for sale to China by the Trump administration last month. Reports say Chinese tech giants like ByteDance, Alibaba, and Tencent placed orders for 400,000 chips. But none of it has shipped. Beijing hasn't cleared imports, and Nvidia has generated exactly zero dollars from those deals.

That $78 billion guidance? It's the floor, not the ceiling. If China approves shipments, Nvidia could blow past that number. If the geopolitical winds shift again and those chips never leave the warehouse, well, Nvidia already told you they weren't counting on it anyway.

This is the investment uncertainty nobody wants to talk about. You're holding a stock trading at nearly 30 times forward sales, betting on a company that just posted a 63.1% net profit margin—yes, they keep 63 cents of every dollar—but whose next quarter's upside is hostage to a trade dispute between Washington and Beijing.

Let's talk about the numbers that are real. Data Center revenue hit $62.3 billion, up 75% year-over-year. That's hyperscalers like Amazon, Google, and Microsoft buying Blackwell chips as fast as Nvidia can make them. Networking revenue—thanks to Spectrum-X and NVLink—surged 263% to $11 billion. The company returned $41.1 billion to shareholders during the fiscal year through buybacks and dividends.

But here's the thing: Nvidia is now the most profitable company in the world. They topped Apple's net income for the first time, pulling in $43 billion in profit this quarter. When you're that dominant, the only risk is concentration. Six companies—Apple, Amazon, Google, Meta, Microsoft, and Nvidia—now generate $202 billion in quarterly profit. That's the entire market's upside packed into a handful of stocks.

If you own Nvidia, you're not just betting on AI demand. You're betting that U.S.-China trade relations don't blow up, that hyperscaler capex stays at $700 billion annually, and that no competitor closes the moat. Those are reasonable bets. But they're still bets.

The China revenue wildcard is what separates this from every other earnings beat story. Most investors don't realize the $78 billion guidance already prices in zero China revenue. If Beijing clears those shipments, it's massive upside. If they don't, Nvidia's already told you it doesn't matter.

Pay attention to what management didn't say. They didn't commit to a timeline on China. They didn't quantify what the revenue opportunity could be if approvals come through. That's the tell. When a company sandbags guidance this hard, they're either being conservative—or they know something the market doesn't.

For now, the AI infrastructure buildout is real. Demand is real. Margins are obscene. But if your thesis depends on China coming back online, you're speculating, not investing. And if they can guide to $78 billion without China, maybe you don't need to worry about it at all.