LG CNS headquarters. [LG CNS] LG CNS headquarters. [LG CNS]

LG CNS has teamed up with Naver Cloud to build next-generation liquid cooling infrastructure, moving to supplement conventional air cooling with large-scale liquid cooling technology. The system will be deployed first at the Samsong data center, which LG CNS operates under a management contract.

LG CNS announced Wednesday that it will apply direct-to-chip cooling — a form of liquid cooling — at the Samsong data center in collaboration with Naver Cloud.

LG CNS previously took over management of the Samsong data center and signed a colocation agreement with Naver Cloud for the facility.

The liquid cooling infrastructure is being built to allow Naver Cloud to stably operate Nvidia’s next-generation, ultra-high-performance GPU “Vera Rubin,” owned by the National IT Industry Promotion Agency under the Ministry of Science and ICT.

“The fact that Vera Rubin will be deployed in domestic data centers on the basis of liquid cooling infrastructure means the Samsong data center will become a key hub for AI factory implementation,” LG CNS said, adding that the facility — with a total power capacity of 80 megawatts — is expected to establish itself as one of the largest AI factories in South Korea.

A conceptual diagram of liquid cooling infrastructure. [LG CNS, AI-generated] A conceptual diagram of liquid cooling infrastructure. [LG CNS, AI-generated]

The direct-to-chip cooling system to be installed at the Samsong data center works by attaching a cooling plate carrying chilled water directly to GPU chips to absorb heat. It uses liquid, which transfers heat far more efficiently than air, to manage the thermal output of high-density GPU servers.

Conventional air cooling systems are generally limited to handling around 20 to 30 kilowatts per rack, while high-performance GPU server racks are increasingly pushing power densities above 200 kilowatts. That gap has made liquid cooling solutions such as direct-to-chip a critical technology for AI factory deployment.

To address this, LG CNS is combining its data center design, build and operations capabilities with AI infrastructure technology to deliver an integrated solution covering direct-to-chip liquid cooling, GPU server operating environments, power supply systems and AI platform operations.

Naver Cloud plans to provide high-performance AI services to public institutions and enterprises in South Korea. The two companies aim to complete construction and validation of the liquid cooling infrastructure by next year.

“Working with Naver Cloud, we will continue building core infrastructure that supports the growth of South Korea’s AI ecosystem, drawing on LG CNS’s data center design, build and operations capabilities and next-generation cooling technology,” said Jo Heon-hyeok, executive vice president of LG CNS’s data center business division.

Meanwhile, LG CNS has also signed a colocation service agreement with Naver Cloud for the Samsong data center, under which Naver Cloud will lease space in the facility to install and operate servers, networking equipment and other hardware. The contract runs from July through May 2035.

sjpark@heraldcorp.com

This content was produced with the assistance of AI translation services.