On May 17, Changxin Technology updated its IPO prospectus for the Science and Technology Innovation Board and resumed its listing review.

The most eye-catching figure in the prospectus is the net profit of 33.012 billion yuan in the first quarter of 2026, which translates to approximately 400 million yuan per day.

The net profit attributable to the parent company in a single quarter is already about 13 times the net profit attributable to the parent company of RMB1.875 billion for the whole year of 2025.

The market space vacated by the three giants

DRAM is the most mainstream category of memory chips and also the area with the most dramatic price increases in this round. According to TrendForce data, the DRAM contract price in the first quarter of 2026 has been revised upward to 93% to 98% quarter-on-quarter, driven by high-bandwidth memory (HBM).

The rapid expansion of AI infrastructure has driven the demand for DRAM in a standard AI server to 8 to 10 times that of a regular server. Specifically for high-end storage such as HBM, which is designed for AI accelerator cards, the demand growth rate is even higher than that of general-purpose DRAM.

Samsung, SK Hynix, and Micron, the three giants of the global DRAM market, have shifted a large amount of advanced node capacity from general-purpose DRAM to HBM. HBM is essentially a stacked DRAM, but with a more advanced process and higher gross margin; SK Hynix's operating profit margin is as high as 72%.

The direct consequence of this capacity shift is a severe supply shortage of general-purpose DRAM such as DDR4 and DDR5.

In February of this year, the Price Monitoring Center of the National Development and Reform Commission issued a document confirming that from September 2025 to February 2026, the global memory market gap will continue to widen, and in January of this year, DRAM prices hit a record high since data became available in 2016.

The key issue is the long conversion cycle of HBM production lines. The three major players had essentially locked in their advanced node capacity for HBM by 2025, and HBM capacity was already sold out in 2026, a situation that cannot be reversed in the short term. The market expects DRAM contract prices to still rise by 58% to 60% quarter-on-quarter in the second quarter.

Changxin Technology's main products are general-purpose DRAM, and it does not have HBM business.This round of profit surge stems from the general market space that the three giants voluntarily relinquished in pursuit of HBM's high profits.

Fourth largest production capacity in the world

During this price surge, Changxin Technology achieved explosive growth thanks to the following key fundamental conditions: near-full capacity, product coverage of mainstream specifications, and customers that are mainly leading domestic cloud vendors.

Currently, the company has three 12-inch DRAM wafer fabs in Hefei and Beijing, ranking fourth globally and first in China in terms of production capacity.The prospectus shows that the capacity utilization rate increased from 87.06% in 2023 to 95.73% in 2025, which is close to full production.

In 2025, DDR series accounted for 31.87% of the main operating revenue, with DDR5 as the main growth driver, while LPDDR series accounted for 66.43%, mainly LPDDR5/5X.

The gross profit margin for the DDR series in 2025 was 41.89%, and for the LPDDR series it was 37.25%, with an overall gross profit margin of 41.02%. In contrast, the gross profit margin for the DDR series in 2023 was -108.76%, clearly demonstrating the extent of the price increase.

From 2023 to 2025, the company's operating revenue was RMB 9.087 billion, RMB 24.178 billion, and RMB 61.799 billion, respectively, with a compound annual growth rate of 160.78%. The performance guidance for the first half of 2026 is revenue of RMB 110 billion to RMB 120 billion and net profit attributable to the parent company of RMB 50 billion to RMB 57 billion.

Based on DRAM sales revenue in the fourth quarter of 2025, Changxin's global market share was 7.67%, and the three giants together accounted for more than 90% of the global market share, namely Samsung 36.6%, SK Hynix 32.9%, and Micron 22.9%.

In the first quarter of 2026, the company's net profit attributable to the parent company was RMB 24.762 billion, ranking only behind Samsung (approximately RMB 215 billion), SK Hynix (approximately RMB 183.7 billion), and Micron (approximately RMB 93.9 billion) among major global memory manufacturers.

Price-driven profit elasticity

The current surge in profits is mainly driven by industry prices, rather than by fundamental optimization of the cost structure.

DRAM is a typical capital-intensive and cyclical industry with extremely high fixed costs, the core of which is equipment depreciation. The prospectus shows that depreciation and amortization expenses will reach 24.68 billion yuan in 2025, which is 2.3 times that of 2023.

Capital expenditures for 2023, 2024 and 2025 are RMB 43.7 billion, RMB 71.2 billion and RMB 49.7 billion, respectively.By the end of 2025, the company had accumulated losses of 36.65 billion yuan, mainly due to high depreciation during the early large-scale construction period.

The inherent characteristic of asset-heavy cyclical industries is that when prices are high, after covering fixed costs, almost all incremental profits are converted into net profit, resulting in extremely high leverage; when prices decline, profits disappear at the same rate because fixed costs do not decrease with prices.

In each cycle of overcapacity in the history of the three giants, the entire industry has suffered huge losses. Changxin Technology's loss period from 2022 to 2024 was a direct manifestation of the previous downturn.

Industry insiders predict that the current boom cycle can continue until mid-2027, based on the fact that HBM's shift to general-purpose DRAM continues to exert structural pressure, and the three giants are maintaining a high degree of restraint in expanding general-purpose DRAM production.

However, if demand for AI servers slows down after 2027, or if the three giants adjust their production capacity strategies, the supply and demand balance may be broken again at any time.

Expansion of production capacity and technological positioning within the cycle window

The timing of this IPO restart is based on a clear logic: the economic cycle is still ongoing, the company has ample cash flow, and the valuation window in the capital market is open.

The company plans to raise 29.5 billion yuan, the second largest financing scale in the history of the Science and Technology Innovation Board. The funds will be invested in the technological upgrading and transformation of the mass production line of memory wafer manufacturing (7.5 billion yuan), the technological upgrading of DRAM memory (13 billion yuan), and the research and development of forward-looking technologies (9 billion yuan), with the core focus on expanding production capacity, upgrading processes, and promoting HBM research and development.

The current mass production process is at the 16nm node. In the absence of EUV lithography machines, catching up in process technology is one of the core constraints, and there is still a gap with Samsung and SK Hynix's below 12nm.

The mass production of DDR5 is a result of process technology iteration. In Q2 2026, Changxin's global market share of DDR5 had risen to 3.97%. HBM is the product line with the biggest gap compared to international giants, but it is also the category with the highest profit margin. Samsung and SK Hynix have already started mass production, and Micron is following suit.

Changxin's current mass production is concentrated on DDR4, DDR5 and LPDDR series. HBM3 samples have been delivered to domestic customers such as Huawei, but the mass production plan is for the Shanghai packaging plant to start production at the end of 2026, which is about 2 to 3 generations behind the industry frontier.

TrendForce predicts that by 2027, HBM's share of global DRAM revenue will rise from approximately 12% in 2024 to over 35%. Manufacturers unable to produce HBM will see their market share shrink in high-growth markets.

This also explains why 9 billion yuan of the funds raised are specifically earmarked for forward-looking technology research and development. Converting capital market financing into investment in technology upgrades during a boom cycle is one of the core logics behind restarting the IPO at this time.

From Cyclical Beneficiary to Global Competitor

The industry is currently still in a boom cycle, and with Changxin operating at full capacity and prices remaining high, profitability is expected to continue. However, this period, after the investment in new production capacity is completed, is also the window where the risk of a cyclical downturn is most concentrated.

Omdia predicts that the global DRAM market will reach $571 billion by 2030, with a compound annual growth rate of 30.56%. As the total market size continues to expand, competition will also intensify.

General-purpose DRAM is a mature market, while HBM and high-end storage for AI data centers are the truly high-value-added areas. Whether Changxin can achieve mass production of HBM3/3E by 2028 is a key variable determining the company's long-term competitive position.

The global DRAM market has been monopolized by three foreign companies for more than 30 years. The reason why this pattern is stable is not because of technical issues, but because DRAM has extremely high requirements for capital scale, process accumulation and supply chain depth.

Since its establishment in Hefei in 2016, Hefei State-owned Assets has indirectly held shares through entities such as Hefei Urban Construction, and has been one of the industrial capital entities that have continuously participated in the company since its inception. Today, it ranks fourth in the world and its quarterly net profit has entered the top four in the world. Such a transformation has not been easy.

Short-term high profits come from cycles, but the production capacity and technological capabilities that support cyclical profits are the result of continuous investment over the past decade.

For China's storage industry, entering the global competitive system is only the first step. Further breakthroughs in the high-end storage field will determine how far it can go in the future.

This IPO is not just a financing opportunity, but also an important milestone for Changxin Technology to move towards the next stage of competition.