LG Group Chairman Koo Kwang-mo will visit Nvidia’s headquarters in Silicon Valley next week to meet with Nvidia CEO Jensen Huang. This marks their reunion roughly two months after Huang visited LG Twin Towers in Seoul’s Yeongdeungpo district on June 8.
According to business circles on the 6th, Chairman Koo is expected to focus discussions with CEO Huang on fleshing out collaborative projects the two companies are pursuing, including physical artificial intelligence (AI), AI data centers, and smart manufacturing. Ahead of this top-level meeting, approximately 30 LG Group executives made a preliminary visit to Nvidia’s headquarters on June 22, and working groups from both companies have reportedly been shuttling between South Korea and the U.S. for working-level discussions since then.
The backdrop to this meeting lies in the invitation and response exchanged during their first encounter in June. At that time, CEO Huang stated, “We will build the future of humanoid robotics together with LG,” adding, “The two companies are collaborating on future data center architecture (AI factories).” Chairman Koo also remarked that “more collaboration will be needed to accelerate the AI era” and mentioned receiving an invitation to California from CEO Huang.
Given that another top-level meeting has been arranged within such a short period, business circles are placing weight on the possibility of concrete collaboration plans being announced. Kim Chang-tae, CFO of LG Electronics, raised expectations during a second-quarter earnings conference call on the 30th of last month, stating that “a strategic partnership at the AI platform ecosystem level with Nvidia could be sufficiently considered.”
At the forefront of the two companies’ collaboration is the robotics business. LG Electronics has agreed to co-develop a reference robot based on Nvidia’s open robotics foundation model “Isaac GR00T.” This is interpreted as a convergence of interests: LG Electronics seeks to expand its business from home appliances to robotics, while Nvidia needs to secure customers for its robotics foundation models and robotics semiconductors. If the reference robot becomes widely adopted, there is potential for core components held by LG Group—such as cameras, actuators, and sensors—to become industry standards.
Collaboration is also advancing rapidly in the smart manufacturing sector. The core objective is to combine 770 terabytes (TB) of manufacturing and production data accumulated by LG Electronics over the past decade at its production sites with Nvidia’s digital twin platform “Omniverse” to realize automated factories.
AI data center cooling solutions represent another noteworthy area of cooperation. LG Electronics announced on the 27th of last month that its 600-kilowatt (kW) class coolant distribution unit (CDU) had completed Nvidia’s AI infrastructure verification. A CDU is equipment that supplies coolant to graphics processing units (GPUs) and central processing units (CPUs) inside servers to control heat generation, and is considered a core component of AI data centers.
Alongside this, the issue of securing Nvidia’s advanced GPUs—essential for the development of LG Group’s proprietary AI model “EXAONE” and the expansion of its AI business—is highly likely to be on the discussion table. LG Group is reportedly pursuing a plan to acquire 10,000 Blackwell GPUs from Nvidia, and there is speculation that additional quantities and delivery schedules could be discussed in detail at this meeting.
Industry observers view this reunion between Chairman Koo and CEO Huang as a potential watershed moment that goes beyond a mere ceremonial meeting, accelerating a strategic alliance spanning the entire AI ecosystem between the two companies.