Nvidia-SK Hynix pact sharpens memory race with Samsung, Micron - digitimes

NVIDIA's multi-year technology partnership with SK Hynix strengthens the South Korean chipmaker's position in the AI memory supply chain, intensifying pressure on Samsung Electronics and Micron Techno... Read original →

The Nvidia–SK Hynix alliance reflects AI’s architectural shift toward memory-compute convergence. SK’s leadership in 3D-stacked HBM4—leveraging TSV and hybrid bonding—cements its position in high-bandwidth memory, directly pressuring Samsung’s GDDR7 roadmap and forcing Micron to accelerate CoWoS-compatible capacity. Geopolitically, U.S. CHIPS Act subsidies for domestic advanced packaging are redefining supply chain eligibility: non-U.S. players risk exclusion unless integrated into Washington-endorsed ecosystems. Over the next 12–24 months, Samsung may open its FO-PLP platform to retain Nvidia, while Micron could acquire back-end packaging assets in Taiwan, China to close integration gaps. This is no longer a race of node scaling—it’s a full-stack battle spanning materials, equipment, packaging, and IP co-optimization.