Faker meetup at PC bang kicks off visit; Chey Tae-won, Koo Kwang-mo and Lee Hae-jin join dinner near Hongik University; NCsoft and Krafton also on agenda; every move of 3-night, 4-day trip draws attention
![Nvidia CEO Jensen Huang (from left) is scheduled to meet SK Group Chairman Chey Tae-won, LG Group Chairman Koo Kwang-mo and Naver board chairman Lee Hae-jin for a dinner in the Hongik University area of Seoul on Friday evening. [Getty Images; SK; LG]](https://wimg.heraldcorp.com/news/cms/2026/06/05/news-p.v1.20260605.2452da27c7814384b82e4ddfbaa13fdf_P1.png)
Nvidia CEO Jensen Huang arrived in South Korea on Friday through the Seoul Gimpo Business Aviation Center in Gangseo-gu, kicking off his second round of AI partnership meetings with Korean companies.
His itinerary spans a samgyeopsal dinner with the country's top conglomerate chiefs, followed by a relay of meetings with game industry CEOs, AI and robotics startups, and students at Seoul National University.
Huang has also signaled a ceremonial first pitch at a professional baseball game and an appearance on a television entertainment program, ensuring that virtually every moment of his three-night, four-day stay will generate buzz across industry, culture and sports.
Particular attention has fallen on his first official engagement: a meeting with professional gamer Lee Sang-hyeok, known by his handle "Faker," at a PC bang in Mapo-gu, Seoul. Huang is well known for having toured Yongsan Electronics Market in the mid-1990s — before Nvidia became a global powerhouse — to promote the company's graphics cards.
Huang has said that "without PC bangs and esports, today's Nvidia would not exist," reflecting his longstanding affection for Korean PC bang culture. His meeting with Faker, the defining figure of Korean esports, is expected to reaffirm that connection.
▶ Dinner with SK and LG chiefs near Hongik University — Huang flew directly to South Korea after completing his schedule at GTC Taipei 2026 and Computex 2026 in Taipei, Taiwan, according to business circles Friday.
After meeting Faker, Huang is set to join a dinner at "Hyeongnim Jeoyo," a samgyeopsal restaurant near Hongik University. SK Group Chairman Chey Tae-won, LG Group Chairman Koo Kwang-mo and Naver board chairman Lee Hae-jin are expected to attend. Hyundai Motor Group Executive Chair Euisun Chung was said to be in last-minute discussions over whether to join.
The gathering comes roughly eight months after Huang's so-called "gganbu dinner" with Samsung Electronics Chairman Lee Jae-yong and Euisun Chung at a fried chicken restaurant in Samseong-dong, Gangnam-gu, last October.
The prospect of Huang and Korea's top business leaders sharing samgyeopsal and soju alongside ordinary diners has already generated expectations that surpass last year's gganbu dinner.
Chey, who held a separate meeting with Huang in Taiwan, is expected to continue discussions on the AI partnership between SK Hynix and Nvidia when the two meet again in Seoul.
▶ Koo Kwang-mo to broaden ties in AI, robots and HVAC — Business circles are paying particular attention to Koo, who will sit down publicly with Huang for the first time. Koo and senior LG Group executives plan to hold a business meeting with Huang at LG Twin Towers in Yeouido on June 8 to advance cooperation in AI, robotics and air-conditioning systems.
Koo has consistently told his organization that "AI transformation is not a task for any single unit — it is something the CEO and business leaders must personally steer," pressing for swift execution of AI transformation across the group.
As part of that push, LG is rapidly expanding its touchpoints with Nvidia. LG AI Research, a unit Koo holds in particular regard, met with Nvidia on May 22 and pledged wide-ranging cooperation, including developing specialized models by combining LG's proprietary AI model "Exaone" with Nvidia's open-source AI model "Nemotron."
▶ Expanding the cloud and gaming alliance with game companies — For Naver board chairman Lee Hae-jin, the agenda with Huang is expected to cover sovereign AI, global AI factories, physical AI, cloud and digital twins. Specific cooperation outcomes are to be announced June 8, when Huang visits Naver's second headquarters, 1784, in Bundang-gu, Seongnam, Gyeonggi Province.
According to Naver, Naver Cloud plans to use Nvidia's open large language model "Nemotron 3 Ultra" to advance its HyperCLOVA X. The two companies also plan to collaborate on optimizing large language models and on foundational technology research.
At his GTC Taipei 2026 keynote, Huang introduced Naver Cloud as an "AI-native cloud partner," underscoring the closeness of the relationship. Earlier, in March, Naver Cloud unveiled a "Seoul World Model" — a digital replica of Seoul's actual road environment and spatial layout built using Nvidia's Cosmos platform. Naver Cloud intends to leverage its full-stack technology and dedicated cloud capabilities to support the construction of sovereign AI models tailored to individual regions.
The AI gaming alliance with Korean game companies is another highlight of Huang's visit. On Sunday, Huang is scheduled to meet Krafton Chairman Chang Byung-gyu and NCsoft CEO Kim Taek-jin in succession — both companies are major customers of Nvidia GPUs.
The meetings are expected to focus on gaming collaboration built around Nvidia's recently unveiled AI PC brand "RTX Spark." The two sides are also likely to explore using virtual game environments as test beds for integrating Nvidia AI technologies such as physical AI and humanoid robots.
Krafton's senior executives visited Nvidia's headquarters in California in April last year to discuss next-generation technology cooperation including robotics. The upcoming meeting is expected to yield more concrete plans.
NCsoft is expected to expand its engagement with Nvidia in physical AI, including object simulation. With subsidiary NC AI at the center of its physical AI push as a future business, NCsoft is likely to focus on cooperation with Nvidia's robotics and digital-twin platform.
By Kim Hyun-il and Park Hye-rim
joze@heraldcorp.com
rim@heraldcorp.com
