LG Expands AI Data Center Cooling Production in U.S. and South Korea

LG Electronics is expanding its global manufacturing footprint for data-center cooling equipment, adding a new chiller plant in Virginia and increasing production at two facilities in South Korea as artificial intelligence drives heavier investment in power-intensive computing infrastructure.

The South Korean electronics group plans to invest a combined KRW 150 billion in the new U.S. facility and expanded manufacturing operations in Pyeongtaek and Changwon.

The Virginia plant will be built in Windsor on a 170,000-square-meter site and will include about 30,000 square meters of manufacturing space. Production of air-cooled chillers is scheduled to begin during the first half of 2027.

The project gives LG a domestic manufacturing base for serving the North American data-center market, where construction of AI and hyperscale computing facilities has accelerated sharply. Virginia is already one of the world's largest concentrations of data-center capacity, making the state an important market for cooling, electricity and other infrastructure suppliers.

LG is simultaneously expanding its existing air-cooled chiller lines in Pyeongtaek and Changwon, introducing conveyor-based manufacturing systems intended to support higher-volume production.

Cooling has become an increasingly critical constraint for the data-center industry as AI chips and high-density server racks generate substantially more heat than traditional computing equipment. LG is positioning its business around what it calls a "Chip-to-Chiller" strategy, combining conventional chillers with coolant distribution units, cold plates and direct-to-chip liquid cooling.

The expansion also comes as LG deepens its relationship with Nvidia. The company said in late September that it had joined the Nvidia Partner Network as a preferred partner for power and cooling solutions. Its 2.6-megawatt coolant distribution unit has also qualified under Nvidia's DSX AI Factory infrastructure program.

LG cited estimates from Omdia projecting global data-center capacity will rise from 226 gigawatts in 2026 to 420 GW by 2030, with North America accounting for roughly 60% of the additional capacity.

The manufacturing investment underscores how the AI infrastructure boom is extending beyond semiconductor companies and utilities into cooling and thermal-management equipment. With hyperscale facilities requiring both more electricity and more sophisticated heat removal, suppliers capable of integrating cooling systems directly into data-center design are emerging as another beneficiary of rising AI infrastructure spending.