Nvidia Faces Billions in Revenue Loss Amid H20 Chip Licensing Hurdles

30.05.2025

As Nvidia released its first-quarter earnings for fiscal year 2026, the company revealed the significant financial toll of recent U.S. chip-export restrictions. The firm reported a $4.5 billion charge directly tied to licensing requirements that limited its ability to sell its H20 AI chips to Chinese companies. Additionally, Nvidia was unable to recognize $2.5 billion in potential H20 chip revenue during the quarter.

The export restrictions, initially introduced in April, have had a swift and heavy impact. Nvidia had anticipated about $5.5 billion in related losses for Q1 and confirmed that projection in its latest report. Looking ahead, the company expects the revenue hit to grow, forecasting an $8 billion shortfall in the second quarter due to the ongoing limitations — a considerable portion of its expected $45 billion quarterly revenue.

During the Q1 earnings call, CEO Jensen Huang emphasized China’s importance in the global AI landscape, noting that the export ban has effectively shut Nvidia out of the $50 billion Chinese market. He stated that the H20 chip ban has ended the company’s Hopper data center business in China and that further modifications to the chip design to comply with regulations are no longer viable.

Despite ongoing tensions and regulatory hurdles, Huang affirmed Nvidia’s commitment to exploring paths to compete in China’s AI sector. He criticized the broader geopolitical strategy, arguing that limiting U.S. chip exports strengthens foreign competitors while diminishing America’s influence in global AI development.

Huang also addressed the broader implications of the export controls, warning that AI development in China will continue regardless of U.S. restrictions. The concern, he noted, lies not in whether China will develop AI, but in whether that development will rely on American technologies. He cautioned that shielding Chinese companies from U.S. competition could backfire, weakening America’s strategic position in the global AI race.