The meeting between SK Hynix and Nvidia leaders focused on strengthening AI hardware collaboration, potentially boosting high-bandwidth memory chip development.
Leaders' Meeting and Collaboration Plans
At the CES conference in Las Vegas, SK Group Chairman Chey Tae Won announced increasing SK Hynix chip development to meet Nvidia's demand. Nvidia co-founder Jensen Huang and Chey discussed boosting high-bandwidth memory chips, essential for Nvidia's AI accelerators.
Korea's Advantage in AI Manufacturing
According to Huang, AI systems can revolutionize industries worth $50 trillion. Chey noted South Korea as a key partner in this transformation due to its manufacturing strengths. SK Hynix and Nvidia agreed to further collaborate on Nvidia's Cosmos platform, aimed at developing physical AI technologies like robots and autonomous vehicles.
Advancements in High-Bandwidth Memory
Nvidia's generative AI systems require fifth-generation HBM3E chips, with SK Hynix leading the development. The company plans to introduce 16-layer HBM3E and 12-layer HBM4 products in the first half of 2025. The entire production for 2025 is already sold out. SK also emphasizes energy-efficient solutions for AI data centers, which Chey views as a growth priority.
The collaboration between SK Hynix and Nvidia aims to advance AI technology and create promising hardware solutions that may drive innovative transformations across various industries.
