Nvidia bets on AI inference as chip revenue opportunity hits $1 trillion

Nvidia is increasingly focusing on AI inference—the stage where trained models generate answers—marking a shift from its traditional dominance in AI training. CEO Jensen Huang stated that “the inference inflection has arrived,” highlighting rapidly growing demand as AI adoption expands into real-world applications. He projected a $1 trillion revenue opportunity by 2027, doubling Nvidia’s previous estimate of $500 billion through 2026 for its Blackwell and Rubin AI chips. This reflects expectations that AI infrastructure spending will continue to surge as companies scale deployment. However, unlike training, inference is a more competitive space. Nvidia faces pressure from central processing units (CPUs) and custom-built chips developed by major players like Google. At the same time, investors have raised concerns about whether Nvidia can sustain its rapid growth and whether its strategy of reinvesting heavily in the AI ecosystem will pay off. Huang’s comments helped ease some of those concerns, reinforcing confidence that demand remains strong and that Nvidia can maintain its leadership position. The announcement briefly boosted Nvidia’s stock, signaling continued optimism as the AI industry moves beyond experimentation into large-scale deployment.