The NVDA print is irrelevant; guidance is the only thing that matters. The core thesis is simple: demand vastly outstrips supply. China is the main overhang, but it's largely a narrative risk, not a fundamental one for the quarter. Sell-side models have already zeroed out China revenue, creating an asymmetric setup. The real risk isn't geopolitics, which are discounted, but any hint of a slowdown in core AI demand from the hyperscalers.