China's CXMT Earns 16 Trillion Won in Half a Year. Samsung Should Worry.

A number that reorders the room

CXMT, China’s largest DRAM manufacturer, reported first-half revenue of 150.31 billion yuan — roughly 30.8 trillion won — up 873.64 percent from the same period last year. Net profit came in at 77.65 billion yuan, about 16 trillion won, after a 2.33 billion yuan loss in H1 2025.

To put that in perspective: CXMT’s half-year profit now exceeds the full-year net income of several major South Korean firms. Its market capitalization on the Shanghai Stock Exchange’s ChiNext board sits at nearly 4 trillion yuan, making it the most valuable company listed on mainland China’s stock exchanges as of late August 2026.

Bloomberg put it plainly: CXMT is now capturing the same boom that has lifted Samsung Electronics, SK Hynix, and Micron. But the implication of that comparison runs deeper than the headline number.

Why CXMT’s profit matters beyond China

The AI boom has driven an unprecedented squeeze on DRAM supply. Data center builders are spending heavily on high-bandwidth memory — HBM — which Samsung and SK Hynix have prioritized. That reallocation has left standard, commodity DRAM undersupplied across the board. CXMT, which doesn’t yet produce HBM at scale, is nevertheless riding the price surge on the memory chips it does make.

The consequence is asymmetric. Samsung and SK Hynix are winning on premium products but ceding volume. CXMT is winning on volume in a market where volume is suddenly scarce and expensive. Both strategies generate profit, but they generate different kinds of leverage.

For Samsung, the risk is structural. The company has been narrowing its DRAM portfolio to focus on HBM and enterprise-grade products, pulling back from the commodity segment where it once dominated. CXMT’s numbers suggest that pullback may have opened a gap the Chinese manufacturer is now filling — and doing so at a pace that outstrips Samsung’s own growth in the high-end tier.

The technology gap that isn’t closing fast enough

CXMT still trails Samsung and SK Hynix by two or three process generations, according to industry assessments. The company cannot access extreme ultraviolet lithography equipment due to U.S. export restrictions, which limits how far it can advance its most advanced nodes. That constraint is real and likely durable.

But DRAM is not only about the leading edge. Standard memory chips — the kind used in servers, consumer electronics, and automotive applications — do not require EUV. CXMT’s Hefei and Beijing fabs, producing roughly 300,000 wafers per month across three 12-inch lines, already operate at competitive process nodes for mainstream DRAM. If two new factories come online as projected, monthly capacity could double to over 600,000 wafers.

That scaling matters because the bottleneck in the current cycle isn’t advanced lithography — it’s throughput. While Samsung and SK Hynix redirect capacity toward HBM, CXMT is filling the gap in commodity DRAM with newly expanded fabs that are already profitable. The restriction on EUV equipment shields China’s competitors from the most advanced chips but does nothing to protect the rest of the market.

Samsung’s pivot — and its blind spot

Samsung’s strategy under CEO Lee Jae-yong has been clear: double down on HBM and memory products where margins are highest, and let commodity DRAM go to those willing to compete on price. The first-half results confirm the strategy is working — Samsung’s profit is strong. But the strategy also assumes that Samsung’s customers will continue sourcing HBM exclusively from Samsung and SK Hynix, and that commodity DRAM demand will remain something the Korean majors can absorb on their own terms.

CXMT’s 16 trillion won profit suggests both assumptions are fraying.

On the HBM side, Samsung and SK Hynix still hold a commanding position. NVIDIA’s GPU dominance keeps demand concentrated among the two Korean firms and Micron. That is unlikely to shift dramatically in the next 12 to 18 months. But on the commodity side, CXMT is no longer a fringe player. It is a profitable, expanding, listed company with a market cap that now ranks above many South Korean industrial giants.

For Samsung, the uncomfortable question is whether it has ceded too much of the DRAM market too quickly. The company’s own pivot was driven by legitimate margin logic. But CXMT’s rise shows that margin logic has a blind spot: it assumes your competitors won’t also be profitable in the space you leave behind.

What happens next

Reuters reported that CXMT expects the DRAM supply shortage and price increases to continue through the second half of 2026. That would extend the window of profitability for the Chinese maker just as Samsung and SK Hynix begin to feel the effects of their own capacity constraints in commodity DRAM.

The listing on the ChiNext board also changes CXMT’s incentives. It now answers to public shareholders who will expect growth, not just survival. That pressure should accelerate fab expansion and push the company further into the mainstream DRAM market that Samsung once considered its core.

For South Korean semiconductor policy, the lesson is stark. The U.S. export controls that restrict China’s access to the most advanced equipment have successfully slowed CXMT’s progression toward cutting-edge nodes. But those same controls have indirectly accelerated CXMT’s competitiveness in the standard DRAM segment — exactly the segment where Samsung has chosen to step back.

The global memory chip order is not breaking. It is being rearranged. CXMT’s 16 trillion won profit is not a fluke. It is a signal that the realignment is already underway.