Nvidia's Lukewarm Forecast Sparks AI Spending Slowdown Fears

Nvidia's Financial Performance and Outlook

• 📊 Nvidia's second quarter revenue exceeded estimates at $46.7 billion, with adjusted earnings per share at $15.
• 🎯 However, the third quarter revenue forecast of approximately $54 billion, plus or minus 2%, was seen as tepid by analysts, falling short of some higher projections.
• ⚠️ Data center revenue for the second quarter was $41.1 billion, slightly below the estimate of $41.29 billion, contributing to investor concerns.

Analyst Reactions and Market Concerns

• 📉 Analysts expressed that while the quarter might be considered "okay" by normal company standards, it fell short of Nvidia's usual "massive blowouts."
• 💡 The deceleration in sequential data center growth from double digits to mid-single digits is a key reason for the stock's negative reaction.
• 💰 A $60 billion share buyback was announced, which some analysts viewed as a sign of maturity rather than aggressive growth reinvestment.

Challenges in the China Market and Supply Chain

• 🚫 Nvidia reported no H20 sales to China-based customers in the second quarter, despite an $180 million release of previously reserved H20 inventory sold to a customer outside China.
• ❓ There is significant uncertainty and unanswered questions regarding China's market and the ability to sell future AI chips there due to export restrictions and geopolitical pressures.
• 🏭 Nvidia relies on outsourced manufacturing at TSMC, and while advanced packaging capacity is easing, new constraints like electricity availability for data centers are emerging.

Future of AI Spending and Demand

• 📈 While the Blackwell architecture revenue grew 17% sequentially, concerns remain about the overall pace of AI spending and whether current demand is sustainable.
• 🧠 The narrative is shifting from broad AI model training to enterprise-specific data training, potentially benefiting companies like Snowflake and MongoDB more directly.
• ⚠️ Hyperscalers are investing heavily in AI infrastructure, but the return on investment (ROI) for these massive expenditures is not yet clear, leading to questions about the long-term demand for AI hardware.
• ⚡ Emerging constraints include the US grid's capacity for electricity needed for gigawatt-scale data centers, posing a significant bottleneck for future growth.