SK Hynix plans to double wafer production capacity by 2030 as chairman warns AI will keep memory tight - TweakTown

SK Hynix plans to double its wafer production capacity by 2030 in response to sustained demand from artificial intelligence (AI), with Chairman Chey Tae-won warning that memory tightness will persist ... Read original →

SK Hynix’s aggressive capacity doubling isn’t merely demand-driven—it’s a strategic lock-in on HBM’s technology moat. The Yongin and Cheongju fabs’ heavy reliance on EUV and near-3nm processes will force DRAM to evolve from commodity to AI-optimized architectures, catalyzing upgrades in TSV, CoWoS, and advanced packaging. Yet geopolitical friction is intensifying: tighter U.S. export controls on advanced memory and Korea’s domestic constraints on power/water could inflate wafer costs by over 15%. Samsung will counter with accelerated HBM4 development to claw back NVIDIA share, while Micron leverages CHIPS Act subsidies to scale Arizona output—triggering volatile pricing. A supply-demand gap in HBM may widen through 2027, but structural oversupply looms as CPO and in-memory computing erode traditional memory dependency.