Industry Analysis
The insatiable demand from AI data centers has transformed memory chips from cyclical commodities into structurally scarce assets. Technologically, HBM3E/HBM4 adoption is diverting TSMC’s CoWoS capacity toward GPU-integrated memory, squeezing standard DRAM wafer starts; meanwhile, QLC NAND’s role in AI training caches is forcing SanDisk to accelerate its BiCS7-to-BiCS8 transition. On the compliance front, prolonged U.S. export controls on semiconductor equipment are delaying ramp-ups at new fabs in Taiwan, China and Korea, inflating industry-wide capex. Faced with Micron and SanDisk trading below market P/E despite soaring fundamentals, Samsung may abandon price wars in favor of AI-optimized memory customization, while SK Hynix could deepen its NAND partnership with Intel at the Dalian fab. Even as new capacity comes online over the next 18 months, explosive growth in AI inference will sustain bit-demand growth above 20%, permanently elevating memory makers’ pricing power and gross margin floors.
This page displays AI-generated summaries and metadata for research purposes. Original content belongs to the respective publishers.