LG Electronics has formally joined the NVIDIA Partner Network (NPN) as a Power and Cooling Solution Preferred Partner, strengthening its position in the infrastructure ecosystem supporting next-generation AI factories. The partnership comes as rapidly expanding AI workloads push data-centre operators to rethink how they manage unprecedented power consumption and heat.
AI Data Centres Face a Growing Cooling Challenge
Generative AI and large-scale model training require dense clusters of high-performance GPUs. Unlike conventional data centres, these AI factories can concentrate enormous computing power within individual racks, substantially increasing both electricity demand and thermal output.
Traditional air-cooling systems are increasingly challenged by such densities, making liquid cooling an important part of the next generation of data-centre infrastructure.
Direct-to-chip liquid cooling addresses this challenge by circulating coolant directly to heat-generating components such as GPUs and CPUs. Coolant Distribution Units (CDUs) regulate and distribute the coolant while integrating with the wider facility cooling system.
NVIDIA Partnership Strengthens LG’s Position
NVIDIA’s Partner Network brings together technology and infrastructure providers whose products undergo technical and reliability assessments for AI data-centre applications. Within NVIDIA’s DSX AI Factory Platform, the DSX Ready designation indicates compliance with applicable reference-design requirements for infrastructure integration, including power and cooling.
On September 28, 2026, LG announced its entry into the NPN as a Preferred Partner, following NVIDIA qualifications for several of its CDU solutions.
LG’s portfolio includes:
· 2.6MW CDU: Qualified as NVIDIA DSX Ready.
· 600kW CDU: Completed NVIDIA infrastructure qualification.
· 1MW CDU: Also completed the applicable qualification process.
· 4.0MW CDU: LG plans to seek NVIDIA qualification by the end of 2026.
LG says its 2.6MW CDU currently represents the largest cooling capacity among DSX Ready CDU partners listed on NVIDIA’s marketplace.
Liquid Cooling Becomes Central to AI Infrastructure
The significance of these systems extends beyond simply removing heat. High-density AI facilities require cooling architectures capable of maintaining performance, reliability and operational continuity as computing loads increase.
LG’s CDUs are designed to work alongside existing air-cooling infrastructure, allowing operators to develop hybrid cooling architectures. Such systems can improve thermal management while supporting greater energy efficiency and stability in high-density environments.
The company’s expansion of CDU capacity also reflects the changing scale of AI infrastructure. As hyperscalers and AI developers deploy increasingly powerful computing clusters, cooling systems must evolve alongside processors and servers.
Targeting the Global AI Infrastructure Market
LG is using its NVIDIA qualifications to strengthen its credentials with hyperscalers, data-centre operators and AI infrastructure developers worldwide. The company is showcasing its qualified CDU technologies at Data Centre World Asia 2026 in Singapore on September 29–30.
James Lee, president of LG’s ES Company, described the NPN Preferred Partner status as a pivotal step in the company’s AI ecosystem strategy.
Cooling Capacity Could Define the Next AI Race
The AI infrastructure race is increasingly becoming a race not only for faster processors but also for the ability to power and cool them efficiently. LG’s NVIDIA partnership places its cooling technologies within a recognised AI-factory infrastructure ecosystem. Its planned move toward a 4MW CDU indicates how rapidly thermal-management requirements are scaling as AI computing moves toward ever-denser deployments.
(With agency inputs)
