Chinese DRAM maker CXMT's +719% explosion is not a rebuttal of the Big 3 bull thesis — it is a gauge of the memory shortage's depth. The bullish confirmation runs through FY27; the FY28 buildout risk belongs on a separate track.
+719% revenue is a shortage reading, not an invasion — the Tencent deal disarmed Apple's China card; a bullish confirmation bounded at FY27, with FY28 on a separate watch track
Chinese DRAM maker CXMT's +719% explosion is not a rebuttal of the Big 3 bull thesis — it is a gauge of the memory shortage's depth. The bullish confirmation runs through FY27; the FY28 buildout risk belongs on a separate track.
CXMT's Shanghai IPO (29.5B yuan raise, review cleared May 27), the June 29 Tencent 20B-yuan server DRAM supply deal, and Apple's White House lobbying over CXMT converged in a single week.
Inside FY27 — earnings leverage for all Big 3 on commodity exposure at record margins. On the FY28 track — SK Hynix most insulated via HBM mix; the heavier the commodity-mobile mix, the closer to CXMT's firing line. The biggest variable: whether the Entity List designation is published.
First, see the company clearly. is a state-backed Chinese DRAM maker founded in 2016 in Hefei, Anhui. It runs three 12-inch fabs, and its mainstay is DDR and (low-power mobile DRAM) — commodity, legacy DRAM. It effectively cannot make (high-bandwidth memory): of roughly 265 (thousand wafer starts per month) of capacity, only 5 kwspm is allocated to HBM, and about 99% of revenue comes from standard DDR and LPDDR, per SemiAnalysis.
Now the IPO filing financials. Full-year 2025: revenue of 61.8 billion yuan (+155.6%), attributable net profit of 1.87 billion yuan. Then in the first quarter of 2026 alone: revenue of 50.8 billion yuan (+719%), attributable net profit of 24.76 billion yuan. One quarter of revenue equals 80% of last year's total; one quarter of profit is 13 times last year's total. First-half guidance climbs to 110–120 billion yuan of revenue and 50–57 billion yuan of attributable net profit. As a still-private company, anything outside this filing — per-fab costs, per-product margins — remains undisclosed.
Here is the fork in the road. Read these numbers as "China's DRAM invasion" and you get it wrong. CXMT's global DRAM bit share was roughly 7.7% in 2025 (Reuters) to ~9% (SemiAnalysis model). Share barely moved from single digits — yet revenue grew sevenfold. Only one variable solves that arithmetic: price. This is a P-driven number, not a Q-driven one. Commodity DRAM contract prices jumped +90–95% in a single quarter, and commodity is exactly CXMT's home turf.
So CXMT's results are an output of memory pricing, not an input. The deeper the shortage, the faster and more violently the marginal supplier's income statement reacts. Taiwan's DRAM makers printed their best numbers at every boom peak in the 2000s — same structure. When the gauge moves, you do not fear the gauge. You read what it points to: the depth of the shortage.
CXMT's +719% is not Chinese capability. It is the depth of the shortage. A gauge is not an input.
Sign in with Google to unlock the full body of every free report instantly.
Sign in with GoogleThis site runs on ads — the tier system rewards community contributions.
This report is provided for informational purposes only and does not constitute a recommendation to buy or sell any financial instrument. Investment decisions should be made based on your own judgment and responsibility. The analysis and opinions contained herein are based on information available at the time of writing and are subject to change.