Summary of Key Points
Changxin Technology, the only domestic company capable of producing DRAM (essential memory for mobile phones, servers, etc.), is about to list on the STAR Market. Over the past three years, it has turned from a huge loss of 19.2 billion yuan in 2023 to a profit of 7.1 billion yuan in 2025, mainly due to the surge in demand driven by AI and a reversal in the industry cycle. However, the strong cyclical nature of the industry, the gap with overseas giants, and its insufficient presence in high-end AI products remain significant challenges. The market is concerned about whether its profitability can be sustained and whether the funds raised from the 29.5 billion yuan IPO will help it close the gap with companies like Samsung and Micron.
I. From a Loss of 19.2 Billion Yuan to a Profit of 7.1 Billion Yuan: Two Key Reasons for the Profit Turnaround
Changxin's profit turnaround was not accidental; it was the result of both external factors and internal capabilities:
- External Factors: Surge in AI Demand + Industry Price Increases
Since the second half of 2025, there has been a sharp increase in demand for DRAM in AI servers, leading to tight supply and demand and rising prices. The average selling price of Changxin's main DRAM products increased by 33.69%, with DDR products seeing a price rise of 61%, which directly boosted revenue.
- Internal Factors: Cost Reductions + Product Portfolio Optimization
In addition to price increases, Changxin also reduced its costs—DDR product unit costs fell by 26%, and LPDDR costs by 22%. Moreover, it sold more high-margin DDR5 products, replacing older DDR4 products. As a result, the comprehensive gross margin jumped from -1.93% in 2023 (losing 19 cents for every yuan sold) to 40.99% in 2025, approaching the levels of Samsung and Micron. This indicates that its scale effects (lowering costs with increased production volume) and improved yield rates (more qualified chips) have been effective.
II. The Industry Cycle: A Potential Threat to Profitability
The DRAM industry is highly volatile, with prices fluctuating significantly, and it is dominated by three oligarchs: Samsung, SK Hynix, and Micron. Changes in supply and demand can greatly affect profits. Changxin's main risks include:
- Downward Cycle Risks: If AI demand does not meet expectations, or if new production capacity is released (e.g., by the giants), prices may fall again. Changxin has high annual depreciation expenses (24.6 billion yuan in 2025), and a downward industry trend could lead to even larger losses.
- How to Survive the Cycle? Experts suggest that the key is not to earn huge profits every year, but to minimize losses during downturns and continue research and development to recover more quickly. Although Changxin has demonstrated profitability, it still needs to reduce costs and increase the proportion of high-end products to withstand cyclical fluctuations.
III. The 29.5 Billion Yuan IPO: Where Will the Funds Be Used? What Is Still Lacking?
The funds raised from the IPO will be used for technology upgrades and capacity expansion:
- 7.5 billion yuan for upgrading wafer production lines,
- 130 billion yuan for DRAM technology improvements,
- 90 billion yuan for research and development in advanced products (such as HBM).
However, it still lags behind the overseas giants:
- Market Share: In the fourth quarter of 2025, Changxin's global market share was only 7.67%, compared to 91.85% for Samsung, SK Hynix, and Micron combined.
- Gaps to Fill: It needs to improve its ability to stably produce advanced technologies (e.g., higher yield rates and lower cost per bit), gain access to high-end customer bases (e.g., more AI server manufacturers), and develop advanced packaging technologies (such as for HBM). Experts predict that Changxin's market share could reach 17% by 2028, but it will need to continue investing in expansion and research.
IV. Challenges with the AI Wave: Can It Keep Up with High-End Products?
AI servers require more advanced storage solutions (e.g., HBM, which offer higher bandwidth and faster speeds). Currently, Changxin's revenue from AI-related products is low, and it has not disclosed any significant HBM sales. In contrast, overseas giants are already generating substantial profits from HBM.
- Short Term: Focus on DDR5: As DDR5 replaces DDR4, Changxin can maintain profitability by selling more of this product.
- Long Term: Emphasize HBM Development: If it cannot keep up with the demand for high-end storage, it may fall behind. Experts point out that AI will set the industry's limits, and Changxin must upgrade its traditional DRAM products while quickly developing HBM to close the gap with international competitors.
V. Conclusion: What Lies Ahead for Changxin?
Changxin has made a significant breakthrough in domestic DRAM production, but to become a global leader, it must overcome three challenges:
1. Coping with Cyclical Downturns: Minimize losses during industry downturns and maintain R&D investment.
2. Cost Reduction and Efficiency Improvement: Improve yield rates and scale to reduce costs per bit to competitive levels.
3. High-End Breakthroughs: Quickly mass-produce high-end storage solutions like HBM and secure more orders from premium customers.
Listing is just the beginning; whether Changxin can become a "Chinese Samsung" depends on its ability to address these challenges successfully.