Apple Seeks to Purchase CXMT Chips, but Production Capacity Already Booked Through 2027
Nashnova编辑部
Apple is seeking U.S. approval to buy memory chips from sanctioned Chinese maker CXMT, but with CXMT running at 95% utilization and booked through 2027, the supply Apple can actually secure is minimal — too little to shift its bargaining position against Samsung, Micron, or SK Hynix.
Why is Apple trying to buy from a sanctioned Chinese chipmaker?
Apple has asked Washington for permission to source memory chips from CXMT and YMTC, for products sold in China only — not for iPhones sold in the U.S.
This means → Apple is drawing a clear line: Chinese chips for the Chinese market, minimizing the political surface area.
The White House is reviewing the request. Sources say the approval may be announced as a "gift" after a Trump–Xi meeting scheduled for September, to signal commercial goodwill and extract concessions from Beijing.
What does Apple stand to gain?
Using local chips in China would free up capacity from Samsung, Micron, and other current suppliers, redirecting that capacity to the rest of the world.
In plain terms = Apple is not replacing its suppliers — it wants one more card at the table. Even a small CXMT allocation would give Apple extra leverage when negotiating prices with Samsung and Micron.
Whether that card is playable depends on how much capacity CXMT can actually spare.
How much can CXMT supply right now?
CXMT currently produces about 200,000 wafers per month, running at 95% utilization. Capacity is booked through 2027.
Apple sells roughly 50 million iPhones a year in China. At 12 GB of memory per device, that totals about 600 million GB of annual demand.
Based on CXMT's D1a/G4 DRAM process — its current production-grade memory technology — at 0.32 Gb per square millimeter, Apple's needs represent roughly 6.6% of CXMT's year-end capacity.
This means → even if CXMT were willing to allocate supply, Apple could secure only a very limited marginal share.
Will CXMT's expansion plans close the gap?
CXMT is scaling aggressively: monthly capacity is expected to reach 300,000 wafers by late 2026, with a planned 50,000-wafer HBM — high-bandwidth memory, used mainly in AI chips — line.
After new fabs in Shanghai and Hefei come online, total capacity could reach 600,000 wafers, potentially surpassing Micron in volume by 2030.
But Goldman Sachs estimates that even at 10 billion GB of output in 2028, CXMT would cover only half of China's 20-billion-GB total demand.
This reflects a clear timing mismatch: Apple wants to use CXMT as leverage now, but CXMT won't have meaningful spare capacity until 2028 at the earliest.
What does this mean for Apple's supply-chain leverage?
Analysts say Apple could obtain at most a very limited marginal supply from CXMT — far too little to change its bargaining dynamics with Micron, Samsung, SK Hynix, and Kioxia.
Put simply = Apple wants another supplier to push prices down, but CXMT can't even meet its existing orders — the leverage play doesn't work in the short term.
The real variable arrives after 2028: if CXMT's expansion stays on track and the political approval comes through, Apple may finally be able to bring CXMT into its supply chain in a meaningful way.
Content is for reference only, not financial advice.