SANTA CLARA, California — Nvidia's market capitalization crossed $4 trillion in after-hours trading following its latest earnings report, making it the second-most-valuable company in the world behind Microsoft and cementing its status as the primary economic beneficiary of the artificial intelligence boom. But for the first time since ChatGPT ignited the AI arms race in late 2022, the conversation around Nvidia is shifting from "how high can it go?" to "what could possibly go wrong?"

The numbers remain staggering. Revenue for the most recent quarter reached $42 billion, more than triple the company's entire annual revenue from just three years ago. Data center revenue alone — driven by H200 and the newer B100 Blackwell accelerators — was $36 billion. Gross margins have expanded to 78%, a level normally associated with enterprise software companies rather than hardware manufacturers. The reason: Nvidia's CUDA software ecosystem, built over 15 years, functions as a powerful moat. "Nvidia isn't just selling chips — they are selling a platform," wrote Goldman Sachs analyst Toshiya Hari, who raised his price target to $1,200. "The switching costs are enormous because an entire generation of AI researchers, developers, and infrastructure has been built on CUDA."

The bear case, articulated most prominently by hedge fund manager Michael Burry, is that Nvidia's current growth rate is unsustainable. AMD's MI400 series is gaining traction with hyperscale customers. Custom AI chips from Google (TPUs), Amazon (Trainium), and Microsoft (Maia) are reducing dependence on Nvidia at the largest cloud providers. And geopolitical risk remains acute: roughly 20% of Nvidia's data center revenue comes from China despite export controls, and any escalation of US-China tensions could threaten that revenue stream. But for now, Jensen Huang's empire shows no sign of slowing.