Nvidia, SK to detail cooperation plan as Huang flags prolonged chip shortage - Devdiscourse

NVIDIA and South Korea's SK Group are set to announce a significant cooperation plan, highlighting deepening collaboration between the two companies in the semiconductor industry. NVIDIA CEO Jensen Hu... Read original →

Jensen Huang’s warning on prolonged memory shortages reveals a structural mismatch between AI compute demand and DRAM/NAND capacity ramp. NVIDIA’s deepening alliance with SK Group isn’t just about securing HBM3E/HBM4—it’s co-defining next-gen AI memory architecture, forcing TSMC, Micron, and Taiwan-based suppliers to accelerate co-design. Geopolitically, SK Hynix benefits from its Korean status to partially sidestep U.S. export controls, yet any collaboration involving advanced nodes or AI chip exports risks BIS scrutiny, inflating compliance overhead. In response, Micron may double down on AMD-Microsoft partnerships, while YMTC could leverage cost advantages to infiltrate tier-2 AI server markets. Over the next 18 months, HBM supply will consolidate further, with the NVIDIA-SK axis erecting a ‘memory moat’ that marginalizes unaffiliated players from the high-end AI hardware stack.