Changxin Memory: With a $3 trillion market value, can it support the future of China’s chip industry?

Let’s talk about ChangXin Memory (CXMT). It listed on Monday, hit a high of 55 yuan intraday, and has a market cap of 3 trillion, becoming the A-share company with the highest market cap. Why can a newly launched company top the rankings right away? Is it real—technology and orders—or is the bubble too big? I spent some time researching.

First, ChangXin Memory (CXMT) is a storage company. It is currently the only DRAM (dynamic memory) firm in China that has achieved large-scale mass production. This technology has long been monopolized by Japanese and Korean semiconductor companies—among the “three giants” we all know: Micron, SK hynix, and Micron.

ChangXin’s technological breakthrough is of extraordinary significance for our country.

Featured products

1. DDR4: Currently the cash-flow biggest product. It’s also the most technologically mature globally. Samsung, SK hynix, and Micron have already produced it in large quantities.

2. DDR5: The key focus of future competition, and also the new standard for PCs and servers. ChangXin has already launched 16GB/24GB products with a maximum speed of 80,000 Mbps. But compared with leading-tier manufacturers, its yield is slightly weaker and power consumption is higher.

3. LPDDR5/LPDDR5X: This mainly used in phones and tablets. But in the high-end smartphone market, it’s still dominated by the first echelon.

How big is the gap between ChangXin and Samsung, SK hynix, and Micron?

1. Market share

Samsung is about 40%, SK hynix 30%, Micron 20%, and ChangXin is roughly a few percentage points (around 5%). However, its growth has been fast in the past two years. Even though the share gap is still significant, ChangXin is the only true new entrant that breaks the three-giant pattern.

2. Process node gap

ChangXin’s current storage technology uses the stacked capacitor technology from Xinyao/Chipright, and on that basis it developed the “Taihang architecture.” It doesn’t rely on US technology and is less affected by external sanctions, but the gap versus Samsung, SK hynix, and Micron is roughly 1.5 to 2 generations—translated to about a 3 to 5 year technology gap.

3. HBM gap

In this generation of AI servers, the most lacking isn’t DRAM—it’s BHM. This is ChangXin’s shortcoming right now. ChangXin is doing R&D deployment and is in a catch-up phase.

A lot of people underestimate how difficult HBM is here. SK hynix didn’t start researching HBM only after AI arrived.

It began investing in 2009, launched the first generation of HBM in 2013, and then only when AI exploded in 2023 did it truly start to reap the benefits—over more than a decade of persistence.

During the R&D period, it累计 invested several tens of billions of dollars, and continued to invest in advanced packaging, TSV, hybrid bonding, and other core technologies, before it finally built the current leading advantage.

Therefore, for ChangXin, even if R&D goes smoothly, it will likely take at least 5 years for HBM to form real competitiveness.

And whether the development speed of the AI industry will leave ChangXin such a long catch-up window is the biggest variable in the future.

Revenue situation

Full-year 2025 revenue is 61.8 billion yuan (9 billion USD), Q1 2026 revenue is 50.8 billion yuan (7.4 billion USD), and Q1 2026 profit is 33.0 billion yuan (4.8 billion USD). Currently the market cap is 3.1 trillion.

Compare this with Micron’s Q1 revenue of 13.4 billion USD, Q2 revenue 23.8 billion USD, and Q3 revenue 41.4 billion USD (just announced). Micron’s reported financials are slightly earlier, but even based on the comparison of publish timing, ChangXin’s revenue and Micron’s revenue difference is roughly 3 to 5 times.

Micron’s market cap is 1 trillion USD—converted to RMB it’s 7 trillion, which is 2 times ChangXin’s. So by this logic, ChangXin is overvalued.

Next look at SK hynix’s revenue. Q1 revenue is 52.5 trillion won (244 billion RMB), which is 5 times ChangXin Memory’s. SK hynix’s market cap is 1,282 trillion won (6 trillion RMB). So compared with SK hynix, ChangXin Memory is also seriously overvalued.

So why can ChangXin command such a big premium?

1. Domestic substitution: Memory is China’s largest category of chip imports. If ChangXin can replace 10% to 20% of that, the scale would also be huge. As you can see from the above data comparison, even though the market share gap is 5 to 8 times, the revenue gap is only 3 to 5 times.

2. Geopolitical push

Since the US restricts semiconductor equipment exports, it also pushes China’s domestic technology to progress. Yesterday, there was also news that domestic lithography machine technology is making breakthroughs. So this wave of semiconductor technology breakthroughs is also something the country needs to back up. (This reasoning is the same as the US “standing firm” with Intel.)

Summary

ChangXin Memory will most likely become the world’s fourth-largest DRAM manufacturer. If it breaks through the HBM technology blockade within 3 years, it could very likely become China’s version of SK hynix, and the outlook should be very bright.

Then, its current market cap is definitely overestimated. Using a benchmark comparison, a market cap of 1 trillion is more reasonable (compared with SK hynix + Micron).

ChangXin is worth looking at long-term. But in the short term, the valuation has already moved ahead of fundamentals. Whether it can digest the current valuation mainly depends on two factors: first, whether DDR5 market share can keep rising; second, whether HBM can achieve real breakthroughs in the next few years.