Nvidia, the tech giant known for its chips that power the global artificial intelligence (A.I.) boom, is facing a major hurdle. Shares of the company dropped by over 16 percent this week after a smaller Chinese competitor, DeepSeek, revealed a game-changing achievement.

DeepSeek demonstrated that it could train a state-of-the-art A.I. system using significantly fewer Nvidia chips compared to previous methods. This breakthrough shakes up the long-held belief that building advanced A.I. systems requires massive investments in data centers packed with Nvidia’s hardware.

Over the past two years, Nvidia has enjoyed incredible success, with its revenue surging by 200 percent to $126 billion and its market value skyrocketing by 700 percent. The company dominated the market, controlling an estimated 90 percent of specialized A.I. chips. But DeepSeek’s innovation has raised questions about whether Nvidia’s chips will remain as indispensable as they’ve been.

Experts suggest that while Nvidia still holds a strong position, companies investing in A.I. might rethink their spending. “The idea was always bigger chips, better A.I. But this raises the possibility that those chips might not be as crucial in the future,” said a tech analyst.

Nvidia’s CEO, Jensen Huang, recently discussed the company’s role in A.I. at CES 2025, but DeepSeek’s announcement has certainly shifted the conversation. The tech industry’s reliance on Nvidia’s chips may not be as ironclad as once thought.

As the A.I. race continues, Nvidia faces new challenges. DeepSeek’s breakthrough could mark the beginning of a shift in how the world approaches A.I. development—and who supplies the tools to make it happen.