Nvidia CEO Jensen Huang arrived in Seoul, South Korea on Friday for his second visit in seven months, declaring upon landing at Gimpo International Airport that robotics represents the country's next major growth sector and hinting at unspecified announcements ahead. "I have brought a lot of business to Korea. I have some surprises," Huang told reporters, according to Reuters and Yonhap News Agency, before a schedule that includes high-level meetings with the leadership of Samsung Electronics, SK Hynix, Hyundai Motor, LG, and Naver — a lineup that spans semiconductors, consumer electronics, automotive, and artificial intelligence platforms.
The visit underscores South Korea's uniquely critical position in the global AI supply chain. Samsung Electronics and SK Hynix together produce approximately 70% of the world's high-bandwidth memory — the specialized DRAM that enables Nvidia's data center GPUs, including Blackwell and the forthcoming Vera Rubin architecture, to operate at the speeds required for large-scale AI inference and training workloads. Any disruption or acceleration in HBM supply from Korea directly affects Nvidia's ability to ramp production and meet the extraordinary demand documented in its May earnings report, which showed fiscal Q1 revenue of $81.6 billion and data center revenue doubling to $75.2 billion.
"Because Korea is a manufacturing center of the world, we can apply the robotics technology, the physical AI technology that we invent here for the industry," Huang told reporters at the airport. He specifically highlighted the semiconductor manufacturing sector as a target for robotics and AI-driven automation, noting that "the manufacturing of semiconductors will become increasingly robotics and increasingly AI-driven in the future." The framing positions Nvidia's Isaac robotics platform and GR00T foundation models — the physical AI software stack Huang has been promoting since CES 2026 as the "ChatGPT moment for robotics" — as the natural partner for Korean industrial automation at scale.
Huang's meetings with Hyundai Motor extend the automotive angle of Nvidia's physical AI push. Hyundai has been an active partner in Nvidia's autonomous vehicle and robotics ecosystem, and any announcements from that partnership would carry implications for both the automotive software stack and the broader industrial AI deployment cycle. Meetings with Naver — South Korea's dominant internet platform and an active investor in both large language models and robotics research — add a software and AI model dimension to the trip that goes beyond pure hardware supply chain.
The visit carries direct market implications for multiple sectors. Nvidia's expanding robotics narrative — moving AI demand from cloud data centers into physical machines, factories, and vehicles — broadens the company's total addressable market beyond the $1 trillion data center capex cycle that has driven its recent performance. For South Korean manufacturers, deepening Nvidia partnerships validate their strategic position at the intersection of AI hardware supply and physical AI deployment. Analysts at Invezz noted that Huang's robotics signal reinforces the thesis that the next wave of AI spend shifts from cloud-only to AI embedded in machines — an expansion that extends Nvidia's CUDA and accelerated computing platform into industrial automation, warehouse robotics, simulation, and automotive AI simultaneously.
No formal agreements or product announcements had been confirmed at the time of writing. Markets will watch closely for any news flow from Seoul throughout the Friday session and into the weekend, as any material partnership announcement — particularly involving HBM supply commitments or robotics platform integrations — could move both Nvidia and its Korean partners when Asian markets open Monday morning.