Nvidia just dropped a bombshell at its annual GTC conference in San Jose: The tech giant predicts its AI chip revenue could hit $1 trillion by 2027 as it shifts focus to real-time AI systems. CEO Jensen Huang, rocking his trademark leather jacket 🧥, unveiled new hardware to dominate the next phase of AI development.
From Training to Talking: The AI Shift
While Nvidia’s chips currently power most AI model training, the company now eyes the "inference" market – where AI answers questions in real time. This $17 billion Groq-powered move comes as rivals like Google develop custom chips to challenge Nvidia’s dominance.
"The inference inflection has arrived," Huang told 18,000 attendees at the hockey arena-turned-tech-stage. "Demand just keeps going up."
Investors Breathe Easier
After Nvidia became the first $5 trillion company in October 2025, some worried about slowing growth. But Huang’s bold $1 trillion forecast (up from February’s $500B estimate) suggests confidence in their new Blackwell/Rubin chips and AI infrastructure plans.
What’s Next? 🤖
Analyst Jacob Bourne notes: "This signals AI’s move from experimentation to global deployment." With the conference running through March 20, all eyes are on whether Nvidia can maintain its 80%+ market share in the AI gold rush.
