SK Group Chairman Chey Tae-won attended NVIDIA GTC 2026 in San Jose, signaling a stronger commitment to expanding SK hynix’s role within the global AI ecosystem and deepening partnerships with leading tech companies.
According to SK hynix, Chey visited the global AI conference, which opened on March 16, marking his first appearance at the event. His participation highlights the growing strategic importance of AI infrastructure and memory technologies for the company.
Chey attended NVIDIA CEO Jensen Huang’s keynote, where a wide range of innovations were unveiled, spanning GPU-accelerated computing, AI factories, open-source AI models, agentic systems, and physical AI. The keynote also outlined the evolution of the AI ecosystem—described as a five-layer structure covering energy, chips, infrastructure, models, and applications—and introduced next-generation AI accelerators, including the Vera Rubin platform and its successor architecture, Feynman.
Industry observers view Chey’s visit as more than a symbolic appearance, interpreting it as a strategic move to expand collaboration with global big tech companies, particularly around high-bandwidth memory (HBM), a critical component in AI infrastructure.
Chey also reunited with Jensen Huang during the event, following a previous meeting in Silicon Valley in February where the two reportedly discussed cooperation on AI semiconductors, including SK hynix’s sixth-generation HBM4. The latest meeting is seen as reinforcing and advancing that partnership.
The long-standing collaboration between SK hynix and NVIDIA dates back to early efforts to develop and supply high-performance HBM for GPU computing, even before the rapid expansion of the AI market. SK hynix’s HBM4 is expected to be integrated into NVIDIA’s next-generation AI platforms, further strengthening the partnership.
During the event, Chey and Huang visited the SK hynix exhibition booth together, where they reviewed the company’s latest AI memory technologies. The showcase featured HBM4 designed for the Vera Rubin accelerator, SOCAMM2 low-power server DRAM modules, and GPU systems equipped with HBM3E.
Additional solutions on display included customized HBM, enterprise SSDs for data centers, and LPDDR5X memory for on-device AI applications, drawing strong interest from visitors.
In a symbolic gesture of collaboration, Huang signed one of the featured exhibits, writing “JENSEN ♡ SK hynix” on the Vera Rubin 200 display.
Chey also toured NVIDIA’s exhibition area to explore next-generation GPU roadmaps, including Blackwell, Rubin, and Rubin Ultra, as well as various AI applications such as autonomous driving and robotics. He further visited booths of key ecosystem partners, including companies providing power and cooling solutions for AI data centers.
“Chairman Chey is actively engaging with key players in the global AI ecosystem to strengthen SK hynix’s leadership in AI,” the company said. “His participation at GTC 2026 will further solidify the company’s position in next-generation AI memory, including HBM4.”
