Nvidia$NVDAB dropped at close, then flipped +4.7% after hours within the same session. That's not noise — that's the market repricing in real time on real numbers.

Nvidia's fiscal Q2 numbers, released after Wednesday's close, showed data center revenue hit a company-record $89 billion, up 116.6% year-over-year, with hyperscale demand from names like Amazon, Google, and Meta driving the bulk of it. Amazon alone confirmed a deal to buy 2 million Nvidia GPUs paired with Nvidia's new Vera CPU — a scale of commitment that tells you hyperscalers aren't slowing capex, they're accelerating it, with next year's combined spend from the top five projected to hit $1.3 trillion.

The stock's initial dip at close and the sharp reversal afterward is a good real-time lesson in how earnings reactions actually work: the headline number isn't what moves price, the context around it is. A slight guidance nuance or margin question can send a stock red in the first ten minutes, and then genuinely strong underlying data (record data center revenue, hyperscaler commitments, forward capex guidance) pulls it back green once analysts finish digesting the full report on the call.

If you trade tokenized stocks or TradFi perpetuals on Binance, this is exactly the kind of volatility window that separates traders who wait for confirmation from traders who react to the first candle.

Questions for you:
— Did you catch this move live, or see it after the fact?
— Do you trade earnings volatility on names like NVDA, or stay out of it entirely?
— Is $1.3T in hyperscaler capex next year a green light for the AI trade, or a sign the spending is getting too aggressive?$NVDA.US #TradingCommunity