NVIDIA Holds Separate Meeting with Korean Companies in Taiwan, Strengthens AI Ecosystem Cooperation [COMPUTEX 2026]

[Financial News, Taipei, Taiwan = Jeong Won-il, Choi Hye-rim] NVIDIA CEO Jensen Huang has arranged a separate meeting exclusively for Korean business leaders in Taiwan, stepping up his outreach to Korean companies. Industry observers say it is unusual for NVIDIA to hold a separate event for companies from a specific country at an overseas venue. Some also view the move as part of an effort to deepen cooperation with South Korea, a key pillar of the AI semiconductor supply chain, while expanding partnerships across the broader AI industry, including automotive, robotics, cloud and information technology.
■ Jensen Huang meets with Korean companies on site
According to industry sources on the 1st, Huang hosted a dinner called "Korean Partner Night" at a local restaurant in Taipei, Taiwan, with executives from major Korean companies.
The separate gathering was held in Taiwan, where COMPUTEX Taipei 2026, Asia's largest AI exhibition, is taking place, but it was reserved for Korean companies rather than Taiwanese firms. Local voices in Taiwan are also understood to have expressed frustration and disappointment over the preferential treatment shown to Korea.
Attendees included Kim Jae-jun, vice president of Samsung Electronics' memory business division; Noh-Jung Kwak, CEO of SK hynix; Jeong Su-heon, executive vice president and CEO of LG Sciencepark; and Yuwon Kim, CEO of NAVER Cloud. They are believed to have discussed cooperation across the AI industry, including advanced memory semiconductors such as High Bandwidth Memory (HBM), as well as robotics, automobiles, cloud and IT services.
As NVIDIA is expanding its influence beyond being a simple supplier of Graphics Processing Units (GPU) for AI implementation and into a core infrastructure company spanning AI data centers, autonomous driving and humanoid robots, its cooperation with Korean companies is also broadening well beyond semiconductors, analysts said.
In fact, during his keynote speech at the Taipei Music Center on the morning of the same day, Huang said, "Our customers and partners do not want to buy computers; they want to build AI factories." He added, "So NVIDIA has begun transforming itself once again so that its technology can be deployed across the entire AI infrastructure." At the keynote, NVIDIA unveiled not only AI accelerators such as Vera Rubin, but also the "NVIDIA Alpamayo 2" inference model for robotaxis and "NVIDIA Cosmos," a foundation model for physical AI. The announcements were seen as part of NVIDIA's push to expand its portfolio as an AI infrastructure company.
■ Chey Tae-won listens to Huang's keynote in person
Chey Tae-won, chairman of SK Group, and Noh-Jung Kwak, CEO of SK hynix, were among the top executives spotted at the keynote venue that morning. Chey and Huang already drew attention in February when they met in Silicon Valley for what became known as a "chicken and beer" gathering. They met again about a month later in March at GTC 2026 in San Jose, underscoring their close ties.
Their appearance at GTC Taipei was seen as an extension of that relationship, giving them a chance to hear the keynote in person and realign on the AI infrastructure roadmap the two companies have been building together.
In particular, direct exchanges between top executives and global tech leaders are viewed as a strategic move to expand partnerships and strengthen market leadership in the rapidly evolving AI ecosystem.
Chey, who listened to Huang's keynote alongside Kwak, remained focused on the presentation throughout. Industry watchers say his attendance reflects SK hynix's effort to further strengthen cooperation with NVIDIA around HBM and secure a stronger position in the next-generation AI memory market.
one1@fnnews.com Reporter