黄仁勋否认AI泡沫!英伟达三季度净利大涨65%,失望无法向中国交付更多产品

Core Insights - Nvidia reported Q3 FY26 revenue of $57.01 billion, exceeding market expectations of $54.92 billion, with a quarter-over-quarter growth of 22% and a year-over-year growth of 62% [2][3] - The net income for the quarter was $31.91 billion, a significant increase of 65% year-over-year, with adjusted earnings per share at $1.30, surpassing the expected $1.25 [2][3] - The gross margin stood at 73.4%, slightly down by 1.2 percentage points year-over-year [2][3] Revenue Breakdown - Data center business accounted for nearly 90% of total revenue, reaching a record $51.2 billion, with a quarter-over-quarter growth of 25% and a year-over-year growth of 66% [3][4] - Gaming revenue was $4.3 billion, showing a slight decline of 1% quarter-over-quarter but a growth of 30% year-over-year [4] - Professional visualization revenue was $760 million, up 26% quarter-over-quarter and 56% year-over-year [4] - Automotive revenue reached $592 million, with a 1% increase quarter-over-quarter and a 32% increase year-over-year [4] Future Outlook - For Q4 FY26, Nvidia expects revenue to be around $65 billion, with a margin of fluctuation of 2% [4] - Projected gross margins are expected to be 74.8% under GAAP and 75.0% under non-GAAP, with a fluctuation of 50 basis points [4] Market Context - Concerns regarding an "AI bubble" have led to significant sell-offs in tech stocks, including Nvidia, with notable investors like SoftBank and Peter Thiel reducing their stakes [5] - Nvidia's CEO Jensen Huang emphasized that the company views the current market dynamics differently, highlighting three major platform transformations: the shift from CPU to GPU computing, the critical point of AI, and the rise of intelligent AI systems [5][6] Infrastructure Investment - Nvidia's CFO Colette Kress projected that revenue from the Blackwell and Rubin platforms will reach $500 billion from early 2023 to December 2026 [6] - The company aims to be a preferred partner in the $3 trillion to $4 trillion AI infrastructure market by the end of the decade, with demand for AI infrastructure exceeding expectations [6] - The Rubin platform is set to begin production in 2026, promising significant performance improvements over the Blackwell platform [6] Geopolitical Challenges - Nvidia faced challenges in fulfilling large orders from the Chinese market due to geopolitical issues and increased competition [6][7] - The company remains committed to engaging with both the U.S. and Chinese governments to maintain competitiveness in the AI computing sector [7]