The 600kW unit is engineered to handle the thermal load of high-performance AI GPUs through a Direct-to-Chip liquid cooling system. By delivering coolant directly to heat-generating components, the system achieves a temperature control precision of ±0.25 degrees Celsius. The hardware also features integrated virtual sensors for real-time monitoring and leak detection, designed to operate in tandem with existing air-cooling infrastructure.
LG Secures NVIDIA Certification for AI Data Center Cooling Hardware
LG Electronics has cleared the technical verification process for its 600kW Coolant Distribution Unit, meeting over 100 rigorous criteria set by NVIDIA. This certification marks the company as an official partner for AI Factory infrastructure, providing global operators with a vetted solution for high-density thermal management.
To ensure reliability, LG subjects its cooling chain to stress tests at its Pyeongtaek validation site, measuring transient responses and low-load stability. This process is part of a broader strategy to scale LG's influence in the AI sector, with plans to certify 1MW, 2.5MW, and 4MW models in the future. The company is currently leveraging its 60-year HVAC legacy to supply major data center projects across North America and Southeast Asia, including the SM+ Data Center in Jakarta. By producing core components like magnetic bearing compressors in-house, LG aims to offer a unified 'Chip-to-Chiller' portfolio that spans from the cooling plant down to individual server processors.
Comments (0)
No comments yet. Be the first!