Nvidia has projected a $1 trillion revenue opportunity for its artificial intelligence (AI) chips by 2027, doubling its previous forecast, as the company intensifies efforts to dominate the rapidly expanding AI inference market. CEO Jensen Huang announced the ambitious target at Nvidia’s annual GTC developer conference in San Jose, California, where he unveiled a new central processor and an AI system leveraging licensed technology from chip startup Groq.
The move aims to solidify Nvidia’s position in inference computing—the real-time execution of AI queries—where competition from rivals like Google and custom processors is growing. While Nvidia’s graphics processors have long dominated AI model training, Huang emphasized that the industry is now shifting toward large-scale deployment. "The inference inflection has arrived," he told a crowd of 18,000 at the hockey arena-hosted event, quipping, "I just want to remind you, this is a tech conference."
Analysts view the updated forecast as a response to investor concerns about Nvidia’s growth trajectory after its valuation surpassed $5 trillion in October 2025. Emarketer analyst Jacob Bourne noted, "Huang mapping out a $1 trillion opportunity through 2027 underscores the durable demand for Nvidia’s AI infrastructure." The company plans to reinvest profits into AI ecosystem development, betting on sustained demand as industries transition from experimental AI projects to full-scale implementation.
