Nvidia blows past revenue targets and forecasts trillions in AI infrastructure spending by end of decade

Nvidia blew past expectations in Q3, reporting revenue up 62% to $57.0bn, datacenter sales of $51.2bn (+66% y/y) and EPS of $1.30, and guided Q4 revenue of $63.7–66.3bn versus $62.4bn consensus; the stock climbed in after-hours trading. Management said demand for its Blackwell GPUs is “off the charts,” cloud GPU capacity is sold out, and they see $500bn of near-term spend on advanced chips and $3–4 trillion in annual AI infrastructure spending by the end of the decade, while expecting gross margins in the mid-70s; data-center chip sales to China were effectively nil due to export restrictions. The report reinforces Nvidia’s dominance and supports a multi-year capex cycle across cloud and AI ecosystems, but the company’s large strategic investments in AI startups and close commercial ties raise questions about circular demand, regulatory/geopolitical constraints and the sustainability of revenue versus the massive infrastructure costs.

Analysis

Nvidia reported third-quarter revenue of $57.0 billion, a 62% year-over-year increase, with datacenter sales of $51.2 billion up 66% and GAAP net income of $31.9 billion or $1.30 per share versus $1.25 expected; management guided fourth-quarter revenue to $63.7–$66.3 billion versus a $62.4 billion consensus, and the stock traded up roughly 3% in the session and as much as 5.7% in after-hours. Management characterized demand as “off the charts,” saying Blackwell sales are extremely strong and cloud GPUs are sold out, and cited visibility into $500 billion of advanced-chip spending over the next 14 months and $3–4 trillion in annual AI infrastructure spending by the end of the decade. Nvidia expects to sustain gross margins in the mid-70s despite rising input costs and is reinforcing ecosystem lock-in via large strategic investments (Nvidia’s announced up-to-$10 billion in Anthropic, prior $6.6 billion in OpenAI, $6 billion in xAI and a commitment tied to a $100 billion OpenAI arrangement). Key risks highlighted in the call include potential circularity from Nvidia’s investments that could overstate demand, geopolitical/export controls that left China datacenter chip sales “essentially nil,” and broader market concern about whether AI-driven revenues will keep pace with the massive capex and energy requirements of next-generation models.

Key Decisions for Investors

• Consider a constructive overweight on NVDA given the beat, strong datacenter demand and above-consensus Q4 guide, while sizing positions to reflect the stock's leadership and concentrated market-cap exposure
• Monitor operational indicators closely—Blackwell sell-through, cloud GPU backlogs, bookings/visibility into the cited $500 billion near-term spend and preservation of mid-70s gross margins—and de-risk if signs of demand softening emerge
• Limit tail risk from geopolitical and regulatory developments by hedging exposure to China export constraints and by watching announcements around the modified H20 product and related export policy changes
• Scrutinize disclosures on Nvidia's strategic investments and compute-commitment deals (Anthropic, OpenAI, Microsoft/Azure) for signs of circular demand or revenue recognition issues and adjust convictions if third-party commercial demand is weaker than implied