NVIDIA CEO Jensen Huang will meet again with Samsung Electronics Chairman Lee Jae-yong and SK Group Chairman Chey Tae-won in New York on the 28th local time. This comes approximately two months after the three met in San Francisco in July.
According to business sources on the 27th, Lee Jae-yong and Chey Tae-won plan to attend the Korea Society's annual dinner to be held at 6 PM on the 28th in New York. Huang will attend the event as this year's Van Fleet Award recipient.
In explaining the reasons for Huang's award, the Korea Society specifically mentioned his cooperation with Korean companies such as Samsung Electronics, SK Group, and Hyundai Motor Group. The organization stated that Huang, who leads the AI and semiconductor industry, has established long-term partnerships with Korean companies, making South Korea a core partner in the global AI infrastructure sector.
The most closely watched topic in the industry for this meeting is High Bandwidth Memory (HBM). As AI processor performance has greatly improved, HBM is capable of delivering massive amounts of data to graphics processors at high speeds, and has now become a core component determining the performance of NVIDIA's next-generation products. The sixth-generation HBM4 has already been applied to NVIDIA's Vera Rubin, which entered full production this year.
Samsung Electronics and SK Hynix are not stopping at supplying HBM4 to NVIDIA. Both companies are building frameworks for joint research and development and optimization of subsequent products, including the seventh-generation HBM4E and eighth-generation HBM5. The AI chip industry model is shifting: no longer is memory selected after the GPU design is finalized, but rather GPU and HBM are co-designed from the earliest design stages. Today, how early a memory manufacturer can obtain NVIDIA's future GPU product roadmap directly determines its competitiveness.
Samsung Electronics is accelerating its cooperation with NVIDIA in this area. Sources say that Samsung Electronics Device Solutions Division Vice Chairman Jun Young-hyun met with Huang during his visit to South Korea in June, and the two discussed long-term cooperation on next-generation HBM such as HBM4E and HBM5. Samsung began mass production shipments of HBM4 in February and sent 12-layer HBM4E samples to customers in May. The company plans to expand its customized HBM business, designing memory performance, power consumption, and interfaces according to customer GPU architectures.
The New York meeting may also cover foundry services and advanced packaging — areas where Samsung also holds advantages. It is reported that Jun Young-hyun and Huang also discussed next-generation foundry cooperation during their June talks. Samsung is entering the AI chip market by being able to simultaneously provide memory, foundry, and advanced packaging services through a single company. This means that the relationship between Samsung and NVIDIA may expand from a simple HBM supply-demand relationship to production across the entire AI chip industry chain.
Comments