LG Electronics Inc (KS:066570) saw its Seoul‑listed shares climb 6.2% to 215,000 won on Tuesday, outpacing the KOSPI’s 1.2% gain, after Nvidia announced qualification of LG’s 2.6‑megawatt (MW) coolant distribution unit (CDU) for the chipmaker’s DSX AI Factory platform. The qualification confirms that the 2.6 MW CDU satisfies Nvidia’s DSX reference‑design requirements and has been added to the Nvidia DSX AI Factory Marketplace, where it is categorized as a “Large‑Scale” solution.
The 2.6 MW unit is engineered to dissipate heat from large AI server clusters through a direct‑to‑chip liquid‑cooling system that circulates coolant via cold plates attached to CPUs and GPUs before transferring the heat to the data‑center’s broader cooling infrastructure. This addition expands LG’s lineup of Nvidia‑qualified cooling equipment, which already includes infrastructure‑qualified 600‑kilowatt (kW) and 1 MW CDUs. The 2.6 MW model also met additional criteria to become “DSX Ready.”
LG highlighted that AI accelerators are becoming increasingly power‑intensive, pushing rack power density into the hundreds of kilowatts and driving demand for centralized megawatt‑class cooling systems capable of serving multiple high‑density racks, rather than conventional rack‑level solutions. To capture this market, LG is promoting its “Chip‑to‑Chiller” portfolio that spans chillers, computer‑room air handlers, and liquid‑cooling systems, covering the full cooling chain from heat absorption at AI chips to heat rejection at the data‑center facility.
Earlier in July, LG secured Nvidia validation for a 600 kW CDU, with Nvidia reporting that the product passed more than 100 technical evaluation criteria covering cooling performance, reliability, and fail‑over capabilities. On the Monday preceding the qualification announcement, LG also launched a new air‑cooled centrifugal chiller designed for AI data centers, further broadening its cooling product suite.
These developments underscore LG’s strategic push into the rapidly expanding liquid‑cooling market for high‑density AI data centers, leveraging Nvidia’s platform validation to enhance market credibility and address the growing thermal management needs of power‑dense AI workloads.