NVIDIA CEO Jensen Huang stated that demand for AI computing has surged significantly in 2025 as AI advances from basic question answering to complex reasoning. Speaking on CNBC’s Squawk Box, Huang explained the ultimate 6 months as a period of “big” increase in computing requirements, using NVIDIA’s shares up more than 2% on Wednesday and increasing the Nasdaq Composite.
“The requirement for computing has gone up extensively,” Huang stated “The AIs are smart enough that everybody wants to use it. We now have two exponentials occurring on the same time.”
The Exponential Era of AI
Huang described that artificial intelligence (AI) models are consuming exponential amounts of computing power even as concurrently producing exponential requirements because of their effectiveness. This feedback loop, he stated, is fueling a new segment of AI-driven industrial expansion.
“Requirement for Blackwell is really, really high,” he stated, referring to NVIDIA’s cutting-edge technology of GPUs. “I think we’re at the starting of a brand new buildout, the starting of a new industrial revolution.”
Powering the AI Revolution
The quick growth of AI infrastructure has increased concerns about power consumption and its effects on availability. NVIDIA lately declared a $100 billion investment in OpenAI’s data center growth, with a purpose to use its GPUs to power up to 10 gigawatts of compute capacity—more or less equivalent to the annual electricity use of 8 million U.S. Households or New York City’s peak summer demand.
Huang warned that US risks falling in the back of China if it fails to boost power generation for AI workloads. “China is way in advance on energy,” he stated, noting Beijing’s rapid progress in power systems to support AI.
Building Energy for the Future
To meet developing energy needs, Huang advised that AI companies consider building self-sufficient data centers powered by using natural gas and, eventually, nuclear energy. “We should put money into just about every possible way of generating energy,” he stated. “Data center self-generated power may want to move lots quicker than putting it at the grid, and we have to do that.”
Huang’s comments underscore the intertwined future of AI and energy. As requirement for high-performance computing increased, energy infrastructure could become the finding out factor in which countries lead the next era of technological development.
