LG CNS Introduces Next-Generation Liquid Cooling at Samsong Data Center, Targeting the AI Factory Market

Collaboration with Naver Cloud: Direct Chip Cooling Technology Implemented
Superior Heat Transfer Efficiency Compared to Air Cooling
Stable Operation of NVIDIA GPUs

LG CNS, in collaboration with Naver Cloud, will establish next-generation liquid cooling infrastructure. As the power consumption and heat generation of high-performance graphics processing units (GPUs) continue to rise, the company plans to proactively implement liquid cooling technology in artificial intelligence (AI) data centers, as it offers superior performance compared to traditional air cooling methods.

LG CNS announced on the 26th that it will apply direct chip cooling (DTC), one of the liquid cooling technologies, at the Samsong data center in partnership with Naver Cloud. Previously, LG CNS signed a colocation agreement with Naver Cloud for the Samsong data center after having been entrusted with its operation. Colocation refers to multiple companies leasing space and facilities within a single data center.

LG CNS announced on the 26th that it will apply direct chip cooling (DTC), one of the liquid cooling technologies, at the Samsong Data Center in collaboration with Naver Cloud. Liquid cooling infrastructure concept diagram (AI-generated image). LG CNS

This liquid cooling infrastructure will be constructed so that Naver Cloud can stably operate NVIDIA's next-generation GPU "Vera Rubin," which is owned by the National IT Industry Promotion Agency. The Vera Rubin GPU is expected to provide approximately 3.3 times better inference performance compared to the existing Blackwell Ultra. The Samsong data center, with a total power capacity of 80 megawatts (MW), is anticipated to become one of the largest AI factories in Korea, and the use of liquid cooling infrastructure is expected to ensure stable operations.

The DTC system to be applied at the Samsong data center is a liquid cooling method where a cold plate, through which coolant flows, is attached to the GPU chip to absorb heat. By utilizing a liquid with better heat transfer efficiency than air, the system controls the heat generation of GPU servers. The cooling capacity of conventional air-cooled systems is limited to about 20–30 kilowatts (kW) per rack, but high-performance GPU server racks are now reaching power densities of over 200 kW. As a result, LG CNS emphasizes that liquid cooling technologies such as DTC have become essential for realizing AI factories.

LG CNS will combine its data center design, build, and operate (DBO) capabilities with AI infrastructure technology to provide integrated solutions, including DTC liquid cooling, GPU server management environments, power supply systems, and AI platform operation. Naver Cloud, leveraging its AI platform management expertise and experience with large-scale GPU services, will offer high-performance AI services to domestic public institutions and enterprises. The two companies plan to complete both the construction and proof-of-concept of the liquid cooling infrastructure by next year.

Cho Heonhyeok, Executive Director in charge of the Data Center Business at LG CNS, stated, "Data center cooling technology is becoming a key competitive factor for AI service quality and operational efficiency. By working together with Naver Cloud, we will continue to establish the essential infrastructure that supports the growth of the Korean AI ecosystem, building on LG CNS's strengths in data center DBO capabilities and next-generation cooling technologies."

Meanwhile, according to the market research firm MarketsandMarkets, the global liquid cooling market for data centers is expected to grow from USD 4.07 billion (approximately KRW 5.6 trillion) in 2026 to USD 27.65 billion (about KRW 38.2 trillion) by 2033.

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