On July 16, 2026, CXMT (ChangXin Memory Technologies) opened its subscription on the STAR Market. The offering price is RMB 8.66 per share, corresponding to a market capitalization of approximately RMB 579.2 billion, with a maximum fundraising target of RMB 66.6 billion. This marks the second-largest IPO in the history of the STAR Market, second only to SMIC (Semiconductor Manufacturing International Corporation). Founded in 2016, the company was still operating at a loss last year, but its net profit for the first half of this year is expected to exceed RMB 50 billion.
What Is CXMT
CXMT focuses on DRAM, or Dynamic Random-Access Memory, and is the only domestic manufacturer capable of mass-producing DRAM. For the first half of 2026, its revenue is projected to reach RMB 110 billion to RMB 120 billion, a year-on-year growth of 612% to 677%; net profit attributable to shareholders is expected to be RMB 50 billion to RMB 57 billion, a year-on-year growth of 2244% to 2544%. In contrast, the net profit attributable to shareholders in 2023 was a loss of RMB 16.34 billion. The core reason for this performance turnaround is straightforward: AI has surged DRAM demand, while supply has failed to keep pace, leading to a significant price increase.
In terms of its technology roadmap, CXMT did not climb step by step from low-end process nodes; instead, it leaped generations in R&D. In 2019, it produced its first 8Gb DDR4 chip, achieving a zero-to-one breakthrough in domestic DRAM. The pace then accelerated: DDR5 entered mass production in 2024, and LPDDR5X was introduced to the market in 2025, with a maximum speed reaching 10667Mbps, benchmarking against international mainstream levels.
In terms of capacity, there are three 12-inch wafer fabs in total across Hefei and Beijing. The monthly capacity will reach 300,000 wafers by the end of 2025, with a target of 400,000 wafers in 2026. According to Omdia data, in the fourth quarter of 2025, CXMT's DRAM sales accounted for 7.67% of the global market, ranking fourth. The top three are Samsung, SK Hynix, and Micron, which together account for over 90%. The 7.67% share still represents an order-of-magnitude gap compared to the leading players. Moreover, this share was captured during a period of tight industry supply and demand, and whether it can be sustained requires continuous tracking.
Opportunities for Domestic Semiconductor Equipment Manufacturers
Capital-level binding has already been established. The strategic placement list for this IPO includes Piotech, AMEC (Advanced Micro-Fabrication Equipment), Anji Technology, TFME (Tongfu Microelectronics), and Simgui AMEC was allocated 18.24 million shares, corresponding to an amount of approximately RMB 158 million. Equipment manufacturers serve as both suppliers and shareholders. Under this structure, the communication costs for equipment verification and introduction will be reduced, and the advancement speed will theoretically be faster. Capacity expansion brings clear incremental orders; the monthly capacity increases from 300,000 to 400,000 wafers, representing a 33% growth. DRAM manufacturing has a higher demand intensity for etching, thin-film deposition, and cleaning equipment compared to logic chips. Brokerage research reports also mention that the high-aspect-ratio etching and advanced thin-film deposition for memory chips result in an exponential increase in process complexity. Piotech's PECVD, AMEC's etchers, and ACM Research's cleaning equipment will all directly benefit.
Furthermore, the verification window for domestic equipment has opened. Previously, wafer fabs were relatively cautious about domestic core equipment, with long verification cycles and high thresholds. During the capacity expansion phase, CXMT pays more attention to supply chain security and cost control, and is willing to give domestic equipment more opportunities to collect data and iterate on mass production lines. This is helpful for equipment manufacturers to narrow the gap with foreign giants.
Impact on Global DRAM Capacity and Prices
CXMT's global market share is less than 8%, which is insufficient to shake the pricing power of Samsung, SK Hynix, and Micron. However, a marginal impact exists—the new supply of 400,000 wafers per month can alleviate some of the tension in the current tight-supply market, but it is not enough to change the overall supply and demand landscape.
DRAM is a highly cyclical industry. The year-on-year fluctuation in product unit prices was 55% in 2024 and 34% in 2025. CXMT itself also explicitly warned in its prospectus: if downstream demand changes or the supply-demand relationship reverses, the high growth in the first half of this year may not be sustainable. This statement is worth careful consideration.
Conclusion
In global competition, CXMT still plays the role of a challenger, but it has already become a player that cannot be ignored. For the domestic equipment supply chain, it is a tangible source of orders and a provider of verification scenarios. For the global market, it is an evolving variable—it cannot change the current landscape at present, but it is worth watching in the long term.