US vs China: Nvidia’s Jensen Huang on Data Centre Blockers

Nvidia CEO Jensen Huang has warned that the US risks falling behind China on key enablers for AI infrastructure, particularly construction speed and national energy capacity. 

Speaking with John Hamre, President of the Center for Strategic and International Studies, Jensen discribed the pace at which each country can bring large-scale digital infrastructure online and questioned whether the US has the power resources to meet escalating compute demands.

His comments come as data centre operators forecast unprecedented expansion across the US, with billions in investment expected to support rapid growth in AI workloads.

Construction speed and infrastructure scale

Jensen told John that the gap in construction timelines is becoming a defining factor for AI infrastructure. 

“If you want to build a data centre here in the United States from breaking ground to standing up a AI supercomputer is probably about three years,” said Jensen. “[China] can build a hospital in a weekend.”

He also raised concerns about comparable national energy availability, arguing that China’s capacity dwarfs that of the US at a moment when data centre operators are racing to secure long term power commitments. 

China has “twice as much energy as we have as a nation and our economy is larger than theirs. Makes no sense to me”, he said.

Jensen noted that China’s energy capacity continues to rise sharply, while growth in the US remains relatively flat. For operators planning multi-gigawatt campuses, predictable access to power is becoming just as crucial as securing semiconductors.

Maintaining an edge in chips

Despite the infrastructure gap he described, Jensen emphasised that the US maintains a clear lead in AI chip technology. He said Nvidia remains “generations ahead” of China in the design and manufacturing processes that underpin modern AI systems.

However, he cautioned against complacency. Jensen added that “anybody who thinks China can’t manufacture is missing a big idea”.

His comments reflect a broader concern across the semiconductor and data centre ecosystem: even with a technological lead, the US must also be able to deploy infrastructure at scale and speed to support its rapidly growing AI economy.

Jensen also pointed to US political momentum around reshoring and AI investment. He said President Donald Trump’s economic priorities could support domestic production and future infrastructure development, offering a path to strengthen national capabilities.

Investment surges to meet AI demand

Nvidia is one of several major technology companies investing heavily in new data centres across the US as AI workloads accelerate. Industry analysts expect the scale of new build activity to grow sharply in the next year.

Raul Martynek, CEO of DataBank, said rising demand for AI compute is reshaping investment strategies across the sector. “In the US, we think there will be 5 to 7 gigawatts brought online in the coming year to support this seemingly insatiable AI demand,” he said, according to Fortune.

He estimates the cost of a data centre at between US$10m and US$15m per megawatt. With a smaller facility typically requiring around 40MW, capital expenditure is climbing, particularly for operators planning multi-site expansions. The projected 5 to 7GW of new capacity equates to roughly US$50bn to US$105bn in new build activity.

These figures underline the scale of the challenge facing data centre developers: balancing the rapid pace of AI adoption with the construction capacity, energy availability and supply chain resilience needed to deliver some of the largest digital infrastructure projects ever undertaken in the US.