Key Takeaways:
3D NAND flash is dropping tungsten word lines for molybdenum as stacking passes 300 layers, opening a $1 billion deposition-equipment market by 2027.
3D NAND flash makers are replacing tungsten word lines with molybdenum as stacking exceeds 300 layers, a materials shift that research firm SemiAnalysis estimates will create a $1 billion deposition-equipment market by 2027.
"Molybdenum is not an optimization — it is a requirement for NAND above 300 layers," SemiAnalysis said on X on Aug. 23.
3D NAND stopped lateral scaling roughly a decade ago, relying instead on vertical stacking to lift storage density. Once layers pass about 300, the tungsten word lines connecting each memory cell hit three physical limits: excessive resistance, fluorine-based chemical processes that cause leakage, and an inability to fill ultra-deep holes. Molybdenum answers all three — lower resistance for faster read/write, no fluorine chemistry, and better filling in high-aspect-ratio structures.
The transition hands Lam Research the first equipment orders, with Tokyo Electron close behind, while Applied Materials, ASMI and China's Naura stand to benefit as DRAM and logic chips adopt the material later. SemiAnalysis projects molybdenum deposition-equipment sales growing from more than $1 billion in 2027 to about $2 billion a year by 2030.
SemiAnalysis places SK Hynix at the front of the stacking race at 321 layers, ahead of Samsung Electronics at 286, Micron Technology at 276 and Kioxia/SanDisk at 218. The leap to 3xx-4xx layers is underway, and every major manufacturer has committed to molybdenum processes.
Samsung is the pioneer, with molybdenum word-line products in mass production since 2024. Micron began large-scale production on Lam Research's ALTUS Halo equipment in 2025, becoming the second manufacturer to commercialize the process. SK Hynix plans to launch 375-layer NAND using molybdenum by end-2026, which would extend its lead in the layer-count race.
SemiAnalysis estimates molybdenum's share of total NAND capacity will climb from mid-single-digit percentages in 2025 to about 30 percent by 2027, reflecting how quickly the industry is adopting the material.
China's YMTC, through its controlling entity Changcun Holdings, is racing to close the gap. The Shanghai Stock Exchange accepted Changcun's STAR Market IPO application on Aug. 21, with a planned raise of 33 billion yuan (about $4.9 billion) for production-line upgrades and research.
Changcun swung to profit in 2024 and expanded it in 2025, posting net profit of 14.21 billion yuan (about $2.1 billion) on revenue of 63.18 billion yuan (about $9.4 billion). First-quarter net profit of 33.38 billion yuan (about $5.0 billion) already exceeded the full-year 2025 figure, with gross margin reaching 76.77 percent.
Yet YMTC has not appeared in SemiAnalysis's 300-plus-layer competitive tier, meaning it faces dual catch-up pressure in technology and capacity during the molybdenum transition window. Its 2025 revenue of 63.18 billion yuan trails Samsung Electronics' 1.62 trillion yuan, SK Hynix's 472.16 billion yuan and Micron's 299.48 billion yuan.
For equipment suppliers, the molybdenum sub-segment alone would equal the full-year revenue of a mid-sized equipment maker by 2027, and it is set to double within three years as DRAM and logic follow. Lam Research benefits first as the incumbent on Micron's ramp, while Tokyo Electron and the later adopters capture the broader transition.
For NAND makers, the manufacturers that complete the molybdenum shift first gain a performance edge in 300-plus-layer products, shaping share in the high-end storage market. Key milestones are SK Hynix's 375-layer mass production by end-2026, Micron's capacity ramp on Lam Research tools, and whether Samsung's head start converts into market share. YMTC's IPO progress and its molybdenum roadmap will also influence the global NAND competitive balance.
This article is for informational purposes only and does not constitute investment advice.
