LG Electronics has secured official NVIDIA certification for its 600kW coolant distribution unit, validating its thermal management technology for AI infrastructure.
This milestone supports the company’s expansion in the AI data center cooling market. Consequently, the certification positions the company as an official partner for AI Factory infrastructure.
The validation process required the system to meet more than 100 technical evaluation criteria. Moreover, these metrics standardize cooling performance, reliability, and failover capabilities for high-density AI Factories. For global data center operators, this designation reduces procurement risks and accelerates qualification for large-scale deployments.
Notably, the demand for efficient cooling has risen alongside the growth of artificial intelligence technologies. This certification provides greater confidence to global hyperscalers during supplier selection. Meanwhile, the company aims to address these needs through its full chip-to-chiller portfolio.
Technical Evaluation and Validation
The evaluation tested the system’s ability to manage the thermal demands of high-heat AI GPUs. Specifically, the system uses a Direct-to-Chip liquid cooling solution that delivers coolant directly to GPUs and CPUs. This creates a hybrid cooling system that operates alongside existing air-cooling infrastructure.
Features of the coolant distribution unit
The certified coolant distribution unit is designed to achieve a temperature control precision of ±0.25°C. This precision enhances system thermal stability and supports reliable server operation. Furthermore, the unit integrates virtual sensor technology for real-time monitoring and leak detection.
Paired with the Data Center Cooling Control Manager, the solution enables predictive control and centralized monitoring. It also supports integration with data center management systems through standard communication protocols. As a result, operators can manage complex computers and server environments more effectively.
Testing and Reliability Standards
To ensure reliability, the company evaluates these solutions at its dedicated validation site in Pyeongtaek. Specifically, here, engineers test system-wide transient responses, temperature changes, and low-load stability across the entire cooling chain. This chain extends from the AI chip load and cold plates to the CDU and chiller.
Consequently, by meeting strict failover requirements, the system maintains operational stability under demanding conditions. The company’s full portfolio includes chillers that produce chilled water, CDUs that distribute coolant, and cold plates. These components are essential for liquid cooling systems in modern data centers.
Future Infrastructure Expansion
This certification serves as the starting point for broader AI infrastructure business expansion. Therefore, moving forward, the company plans to secure certifications for its larger coolant distribution unit models, including 1MW, 2.5MW, and 4MW versions. This expansion aligns with growing global investments in the digital economy.
The company has recently secured major global data center cooling projects, including hyperscaler facilities in North America. In addition, it secured a project at the Sinar Mas data center, which is an AI-ready facility in Jakarta, Indonesia.
“As AI data centers continue to scale, the need for highly reliable liquid cooling solutions has never been greater,”
James Lee, president of the LG ES Company
