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

U.S. semiconductor giant NVIDIA and South Korean conglomerate SK are set to announce a cooperation plan, as NVIDIA CEO Jensen Huang warned that the ongoing chip shortage will persist for several years... Read original →

Jensen Huang’s warning of a multi-year chip shortage reveals a structural mismatch between AI demand and manufacturing capacity. NVIDIA’s partnership with SK Group isn’t just about supply—it’s a technical stack reconfiguration: SK Hynix’s leadership in HBM3E and 3nm EUV memory directly enables NVIDIA’s next-gen Blackwell Ultra and AI supercomputers. Geopolitically, this diversifies reliance away from Taiwan, China amid U.S. CHIPS Act constraints, positioning SK as a more controllable Asian node. Compliance costs are now explicit; expect co-investment in silicon photonics and advanced packaging to sidestep export controls. TSMC and Samsung will likely accelerate AI-optimized memory integration, while Intel may push Foveros with LPDDR5X bundling. Over the next 18 months, AI chip competition will shift from transistor density to holistic efficiency across memory, interconnect, and packaging—defining who controls the heterogeneous integration ecosystem.