CXMT Plans to Build Second DRAM Wafer Fab in Beijing
Alina Collins
CXMT is in early funding talks with Beijing authorities to build a second 12-inch DRAM fab in the Yizhuang development zone; if all projects reach full capacity, monthly output would exceed 600,000 wafers — a step-change in China's domestic DRAM capability.
Why Beijing Yizhuang for a second fab?
CXMT (长鑫存储) plans to build a new 12-inch memory-chip fab in Beijing's Yizhuang Economic-Technological Development Area and has begun preliminary financing talks with Beijing municipal agencies.
The new plant would complement CXMT's existing Beijing facility and fit into a broader expansion spanning Shanghai and Hefei.
This means → CXMT is shifting from a single-city base in Hefei to a multi-city footprint, spreading capacity risk while tapping additional local-government funding.
Where will the $8.6 billion IPO money go?
CXMT just completed a record $8.6 billion IPO on Shanghai's STAR Market, becoming the flagship case of the "Hefei model" — local governments using state-backed funds to incubate strategic tech firms.
Hefei's early investment in CXMT reportedly returned roughly 5,000%, prompting other Chinese cities to replicate the semiconductor-funding playbook.
IPO proceeds are earmarked for expanding wafer capacity, investing in more advanced DRAM products, and closing the technology gap with global leaders. In plain terms = the money splits three ways: build fabs, upgrade process nodes, catch up with Samsung and SK Hynix.
Where does CXMT rank in global DRAM today?
Counterpoint Research estimates CXMT holds about 9% of global DRAM bit shipments, rising to roughly 11% by 2028.
Analysts say the company needs to reach about 15% share to sustain the investment required for next-generation memory technology.
This means → the gap from 9% to 15% is still nearly a doubling; the IPO cash is more of a ticket to compete than proof of victory.
Will the expansion crash DRAM prices?
Morningstar analyst William Kerwin argues that CXMT's share of global DRAM output remains limited and that fears of immediate oversupply are likely overstated.
Significant pricing pressure would materialize only if multiple suppliers complete large-scale expansions simultaneously over the coming years.
This reflects a critical timing question: what determines DRAM pricing is not CXMT's expansion alone, but whether its capacity ramp collides in time with those of Samsung, SK Hynix, and Micron.
Content is for reference only, not financial advice.