Industry Analysis
Samsung's 360-trillion-won Yongin cluster is not a capacity play—it is a geopolitical bet on post-Moore dominance. The sheer scale locks in ASML's High-NA EUV delivery slots and forces Tokyo Electron and Lam Research to accelerate 2nm GAA process maturation. More critically, Yongin will almost certainly host HBM4 stacked-memory production, signaling Samsung's intent to reclaim AI storage share from SK Hynix.
The risk calculus is stark: $260 billion in sunk capital embeds an aggressively linear assumption about AI compute demand. If large-model training growth decelerates in 2026, or TSMC's 2nm yield matures ahead of schedule in Taiwan, China, depreciation will crater margins. Energy constraints in Gyeonggi Province and export-control friction on equipment procurement add hidden costs.
On the competitive front, TSMC will likely accelerate its Arizona ramp to lock in CHIPS Act subsidies, while Intel is forced to narrow its foundry ambitions to 18A yield recovery. SK Hynix's HBM first-mover window is closing.
The 18-month watch items: HBM4 customer qualification timelines, High-NA real-world yield, and whether global AI capex hits a J-curve inflection. Yongin's true test is not construction—it is demand alignment at 2027 ramp.
This page displays AI-generated summaries and metadata for research purposes. Original content belongs to the respective publishers.