NVDA

<p>By 2030, global AI capital expenditure is expected to reach USD 3-4 trillion; this year, NVIDIA's sovereign AI-related revenue could reach USD 20 billion, more than doubling from last year; based on the Blackwell architecture, a 10-fold return on investment can be achieved per token, which is a 50-fold improvement in token efficiency compared to Hopper</p>

- Nvidia's Q2 2026 revenue reached $46.7 billion, with a 56% year-over-year increase in data center revenue, despite a decline in H20 chip sales.  
- CEO Jensen Huang highlighted significant growth opportunities in AI infrastructure, projecting $3 trillion to $4 trillion in spending by the decade's end, driven by rising demand for AI capabilities.  
- The company anticipates $20 billion in sovereign AI revenue this year, reflecting a doubling from the previous year, as global investments in AI infrastructure continue to surge.

Wallstreetcn

NVIDIA

NVIDIA conference call: "The opportunity to bring Blackwell to the Chinese market" is real, and this year's gross margin is still expected to reach the mid-70% level