Focus on the BIG picture.
Sunday, Dec 07, 2025
Outside the box.
Nvidia CEO Says U.S. Data-Center Builds Take Years while China ‘Builds a Hospital in a Weekend’
Jensen Huang warns U.S. regulations slow AI infrastructure rollout compared with China’s rapid build pace
Nvidia’s Chief Executive Officer, Jensen Huang, recently highlighted a stark contrast between the timelines for constructing data centres in the United States and China, suggesting that while U.S. projects take roughly three years, Chinese authorities seem to complete major builds — such as hospitals — in a matter of days.
The remarks come as part of a broader defence of rapid global AI infrastructure build-out amid regulatory and energy-cost challenges in the United States.
Huang said that data-center construction in the U.S. is slowed by regulatory complexity, state-by-state permitting, environmental reviews, and high electricity prices.
In contrast, he described China as having the capacity to mobilise resources and approvals swiftly, enabling exceptionally rapid construction — an argument he used to warn that the U.S. risks falling behind in the global AI race.
He also reiterated his firm’s expectation that worldwide spending on data centre infrastructure will surge, reaching up to one trillion dollars by 2028. Under that outlook, the infrastructure required to support generative artificial intelligence — from GPUs to power and cooling systems — must keep pace with demand.
Huang described modern data-centres as “AI factories,” urging governments and industry to treat them as foundational infrastructure akin to electricity grids or communications networks.
The comments arrive amid mounting concern over long-term U.S. competitiveness in AI, especially after the imposition of export restrictions on advanced chips.
Huang repeated that supply-chain disruptions and energy costs could undermine America’s ability to lead in AI, despite high domestic investment intentions.
While critics caution that such comparisons oversimplify regulatory, labor and environmental realities, Huang’s remarks reflect a growing urgency among industry leaders to accelerate both policy and infrastructure reforms.
As nations race to deploy AI-ready data centres, differences in regulatory frameworks, energy policies and state-led planning models may shape who ultimately leads the next wave of technological innovation.
The remarks come as part of a broader defence of rapid global AI infrastructure build-out amid regulatory and energy-cost challenges in the United States.
Huang said that data-center construction in the U.S. is slowed by regulatory complexity, state-by-state permitting, environmental reviews, and high electricity prices.
In contrast, he described China as having the capacity to mobilise resources and approvals swiftly, enabling exceptionally rapid construction — an argument he used to warn that the U.S. risks falling behind in the global AI race.
He also reiterated his firm’s expectation that worldwide spending on data centre infrastructure will surge, reaching up to one trillion dollars by 2028. Under that outlook, the infrastructure required to support generative artificial intelligence — from GPUs to power and cooling systems — must keep pace with demand.
Huang described modern data-centres as “AI factories,” urging governments and industry to treat them as foundational infrastructure akin to electricity grids or communications networks.
The comments arrive amid mounting concern over long-term U.S. competitiveness in AI, especially after the imposition of export restrictions on advanced chips.
Huang repeated that supply-chain disruptions and energy costs could undermine America’s ability to lead in AI, despite high domestic investment intentions.
While critics caution that such comparisons oversimplify regulatory, labor and environmental realities, Huang’s remarks reflect a growing urgency among industry leaders to accelerate both policy and infrastructure reforms.
As nations race to deploy AI-ready data centres, differences in regulatory frameworks, energy policies and state-led planning models may shape who ultimately leads the next wave of technological innovation.
