[DigitalToday reporter Jinju Hong] U.S. power-generation equipment markets are booming as electricity demand surges due to the expansion of artificial intelligence (AI) data centres. GE Vernova, one of the world’s largest gas turbine makers, said orders for large gas turbines are effectively fully booked through 2029. The race to build AI infrastructure is spreading beyond semiconductors to competition over securing power equipment.

On June 27 local time, CNBC reported that major hyperscalers including Amazon, Google, Microsoft (MS) and Oracle are expanding adoption of GE Vernova’s industrial gas turbines to secure stable power needed for AI data centres.

Behind this is the massive power demand of AI data centres. Generative AI services must run large GPU servers 24 hours a day, so they need far more stable electricity than existing data centres. With grid connections delayed across the United States, big tech companies are also moving to secure their own power-generation facilities rather than waiting for grid expansion.

To meet such demand, GE Vernova is expanding production capacity at its Greenville plant in South Carolina. After hiring 200 new employees last year, it plans to hire another 300 by the end of this year to accelerate production of large gas turbines.

Pablo Koziner (파블로 코지너), GE Vernova’s chief commercial and operating officer (CCO), explained that industrial gas turbines are one of the most practical solutions to meet the AI era’s large-scale power demand. "With AI and electrification expanding, we need large-scale and stable power," he said. "The gas turbines we produce at our plant are responsible for a significant portion of that demand."

Large technology companies’ projects are also continuing. Microsoft purchased seven GE Vernova gas turbines to supply power to a Texas data centre. Total generation capacity is 2.7 gigawatts (GW), enough to supply electricity to about 3 million households.

According to market research firm Cleanview, GE Vernova gas turbines have also been installed at xAI’s 'Colossus 1' data centre in Tennessee. Gas turbine deployment of about 1 GW is also under way for OpenAI’s Texas 'Stargate' project, it is said.

AI companies’ interest is also intense. CNBC, citing a source familiar with the matter, reported that most executives of major hyperscalers have visited GE Vernova’s Greenville plant in person. Companies including OpenAI are directly examining power facilities and industrial infrastructure, in an analysis that the AI race is expanding beyond servers and semiconductors to competition over securing power equipment.

Gas turbine prices are also rising quickly. Large gas turbines produced by GE Vernova are about 9.4 metres tall and weigh 280 tonnes, and a single turbine can supply power to about 500,000 households. The industry estimates the price per unit at more than $250 million, and investment bank Melius analysed that prices have risen about 300 percent over the past three years.

Orders have already been secured for years. GE Vernova said about 20 percent of its current backlog in its gas power generation business comes from data centres and AI projects. Koziner explained, "All production capacity through 2029 has been reserved, and contracts continue to be signed for 2030 and 2031."

The market also reacted positively. GE Vernova shares have risen about 60 percent over the past six months. The result reflects expectations that the expansion of AI infrastructure will lead to improved performance by power equipment companies.

Still, the expansion of AI data centres is not proceeding smoothly in all respects. Community opposition and environmental burdens surrounding construction of large power facilities and data centres are still cited as issues that must be resolved.

In response, GE Vernova is also focusing on improving generation efficiency. Koziner explained that the efficiency of its current gas turbines has roughly doubled compared with models from 20 years ago, and said it is continuing technology development to improve generation efficiency and reduce carbon emissions.

As infrastructure competition in the AI era expands beyond semiconductors to competition over securing power, power equipment companies are emerging as a new beneficiary industry. At the same time, how to balance stable power supply and environmental burdens is emerging as an important task for the growth of the AI industry.