Despite having delivered 62% year-over-year revenue growth in Q3 to a record $57 billion and net income rising 65% to $31.9 billion, Nvidia trades at a forward P/E of 25—below the Nasdaq-100 average of 26 and well beneath peers such as Amazon (28) and Apple (33). With a $4.63 trillion market capitalization and gross margins consistently above 70%, the company has authorized a $62.2 billion buyback plan to reduce share count and boost EPS, highlighting management’s confidence in the sustainability of its core data-center business.

Analysts at Goldman Sachs forecast that hyperscalers could spend $527 billion on AI infrastructure in 2026, a figure that reflects both competitive deterrence and speculative outlay. On the consumer side, Deutsche Bank projects OpenAI’s cumulative burn could reach $143 billion by 2029 ahead of its anticipated IPO. While Nvidia remains a pick-and-shovel supplier insulated from direct profit erosion, prolonged pressure on customers’ cash flows may eventually temper demand for premium GPUs and weigh on Nvidia’s near-term order visibility.

Building on the ‘off-the-charts’ adoption of its Blackwell architecture, Nvidia plans to launch its Rubin class GPUs—optimized for AI video generation—in early 2026. The company’s data-center division, which already accounts for over 90% of operating profit, could face demand tapering if AI spending sentiment falters. Even a modest contraction from current 70%+ gross margins would have a magnified impact on free cash flow, underscoring the importance of monitoring customer capex trends alongside product ramp execution.