LG Electronics has obtained official certification from NVIDIA on its 600 kW Coolant Distribution Unit. This milestone validates the reliability and efficiency of LG’s liquid cooling solutions and thermal management technology under NVIDIA’s AI infrastructure standards, supporting the company’s continued expansion in the AI data center liquid cooling market.
As one of the few companies globally to have passed NVIDIA’s technical verification process, LG has successfully validated its 600 kW CDU against more than 100 demanding technical evaluation criteria. These metrics are designed to standardize liquid cooling performance, reliability, and fault tolerance capabilities in high-density AI factories.
This certification represents a strategic milestone for LG, reinforcing its position as an official NVIDIA verified partner for AI factory infrastructures. For data center operators, this distinction provides greater confidence in selecting suppliers, reducing supply chain risks and accelerating qualification for large-scale liquid cooling-based AI infrastructure deployments.
End-to-end cooling solution provider – from chip to cooler
LG’s 600 kW CDU, certified by NVIDIA, is designed to handle the thermal demands of high-power AI GPUs. By implementing a direct-to-chip liquid cooling solution, which distributes coolant directly to the components that generate the most heat (such as GPUs and CPUs), the unit creates a highly efficient “hybrid cooling” system that works in perfect synergy with existing air cooling infrastructure.
The unit is designed to achieve a temperature control accuracy of ±0.25°C*, which helps improve system thermal stability and ensures reliable operation of AI servers. It also integrates virtual sensor technology for real-time monitoring and leak detection. Combined with LG’s data center cooling control manager, the solution enables predictive control and centralized monitoring, and facilitates seamless integration with data center management systems using standard communication protocols. These capabilities optimize liquid cooling performance in critical environments.
To ensure absolute reliability, LG evaluates these solutions at its dedicated validation center in Pyeongtaek. There, the company performs system-wide transient response, temperature variation, and low-load stability testing across the entire liquid cooling chain: from the AI chip load and cold plates, to the CDU and cooler. Meeting NVIDIA’s strict fault tolerance requirements, the CDU is designed to maintain operational stability even under the most demanding conditions.
LG aims to further consolidate its position as a trusted partner in AI data center infrastructures thanks to its comprehensive Chip-to-Chiller portfolio, which offers end-to-end thermal management solutions, from the chip itself to the cooling plant. The company’s comprehensive offering includes chillers that produce cold water, CDUs that distribute coolant, and cold plates that bring liquid cooling directly to AI chips, the key components of liquid cooling systems for AI data centers.
Expanding capabilities in AI infrastructures
This certification is the starting point for broader expansion of LG’s AI infrastructure business. Going forward, LG plans to sequentially obtain NVIDIA certifications for its entire range of high-capacity CDUs, including 1 MW, 2.5 MW and 4 MW models, reinforcing its leadership in liquid cooling solutions.
As AI data centers grow, demand for reliable liquid cooling solutions soars
Backed by 60 years of engineering experience in the HVAC industry, LG CDUs are designed for high-density AI data center environments, offering precise temperature control, advanced management capabilities and high reliability for liquid cooling applications. The company also strengthens its competitiveness by in-house production of key components, such as inverter systems and magnetic bearing compressors, technologies widely applied across its chiller range.
Recently, LG has been awarded major global data center cooling projects, including facilities in North America and the Sinar Mas data center, SM+ Data Center’s flagship AI-ready facility in Jakarta, Indonesia. Leveraging its core technology capabilities, LG aims to increase its global orders and strategic alliances in the next-generation AI data center market, driving the adoption of liquid cooling solutions.
“As AI data centers continue to grow, the need for highly reliable liquid cooling solutions is more pressing than ever,” said James Lee, president of LG ES Company. “Leveraging our advanced cooling technologies along with LG’s integrated infrastructure capabilities, we are expanding our AI infrastructure business with comprehensive solutions that help enable physical AI, where AI generates value in real-world environments.”
