Nvidia`s CEO raises concerns about energy gap in US-China AI race

Nvidia CEO Jensen Huang warned that China’s rapid building capacity and expanding energy supplies give it advantages as global demand for artificial intelligence accelerates.

Huang said large U.S. data centers take about three years to complete. He contrasted that pace with construction in China, saying the country can finish major facilities in a fraction of that time. He told Center for Strategic and International Studies President John Hamre in late November that China can build a hospital in a weekend.

Huang said China’s speed is only part of the challenge. He noted that the country has roughly twice the national energy capacity of the United States, despite having a smaller economy. He said China’s energy supply continues to rise sharply, while U.S. growth remains flat.

Huang said Nvidia still leads China by several generations in advanced AI chips. He cautioned that China’s manufacturing strength should not be underestimated and said it would be a mistake to assume China cannot scale production.

He also pointed to U.S. efforts to expand domestic manufacturing. He said President Donald Trump’s push to reshore jobs and increase AI investment supports Nvidia’s long-term prospects.

Huang recently drew attention with remarks suggesting China could win the AI race. He later revised that view, saying China was only “nanoseconds behind America” in a message posted on his company’s X account.

Nvidia is part of a broader wave of companies planning major data center growth in the United States. Industry executives expect more than $100 billion in new construction next year.

Raul Martynek, CEO of DataBank, said data centers typically cost between $10 million and $15 million per megawatt. He said a smaller facility requires around 40 megawatts. Martynek expects five to seven gigawatts of new U.S. capacity in the coming year to meet rising AI demand. He said the investment could range from $50 billion to $105 billion.