Nvidia CEO Jensen Huang has made a bold prediction: the next generation of artificial intelligence will require 100 times more computing power than today’s models. As AI shifts toward reasoning-based intelligence—where systems process information step by step to reach the best possible conclusions—the demand for cutting-edge hardware will skyrocket.
“The amount of computation necessary to do that reasoning process is 100 times more than what we used to do,” Huang told CNBC’s Jon Fortt in an exclusive interview following Nvidia’s latest earnings report.
He highlighted advanced models such as DeepSeek’s R1, OpenAI’s GPT-4, and xAI’s Grok 3 as early examples of this emerging AI evolution, where machines don’t just predict but actively reason, analyze, and optimize responses.
Nvidia’s AI Boom: A Revenue Powerhouse
Despite mounting challenges, Nvidia’s financial performance remains nothing short of extraordinary. The company posted a record-breaking $39.33 billion in revenue for Q4—a staggering 78% year-over-year increase.
Nvidia’s AI-driven data center sales alone surged 93% to $35.6 billion, dominating more than 90% of the company’s total earnings.
Yet, despite this financial success, Nvidia’s stock has struggled to fully recover from a 17% drop on Jan. 27—the company’s sharpest decline since 2020. Investors were rattled by concerns that AI startups, like China-based DeepSeek, could achieve similar AI breakthroughs with significantly lower infrastructure costs.
Huang’s Response: AI Demand is Only Growing
Huang dismissed these fears, emphasizing that DeepSeek’s achievements actually reinforce the massive demand for Nvidia’s AI chips.
“DeepSeek was fantastic,” he said. “It open-sourced a reasoning model that’s absolutely world-class.”
Despite tighter U.S. export restrictions preventing Nvidia from selling its most advanced chips to China, Huang remains confident. While China’s revenue contribution has dropped by half, Nvidia continues to thrive, with tech giants worldwide pouring billions into AI infrastructure.
However, competition is heating up. Chinese tech giant Huawei is aggressively expanding its AI hardware capabilities, presenting a new challenge for Nvidia’s market dominance.
AI Without Limits: Software Will Always Find a Way
One thing Huang is certain about? AI innovation won’t be stopped by hardware restrictions.
“Ultimately, software finds a way,” he asserted. “Developers will optimize AI for any system—whether it’s supercomputers, PCs, phones, or gaming consoles.”
As Nvidia looks ahead, Huang is banking on its latest GB200 AI chip, which is now available in the U.S. This powerhouse chip can generate AI content 60 times faster than the restricted versions allowed for sale in China, giving Nvidia a significant competitive edge.
The Future of AI: A Computing Arms Race
With AI advancing at an unprecedented pace, the demand for high-performance computing is set to explode. Will Nvidia continue to dominate this new AI era, or will emerging competitors reshape the landscape?
One thing is clear: the next phase of AI won’t just be about speed—it will be about intelligence, adaptability, and the raw power to fuel a new generation of machine reasoning.
The AI revolution is here. Are we ready for what comes next?
