New Stocks Preview | AI storage becomes a market trend, can Starry Sky integrate successfully under the watchful eyes of giants?

Due to the explosive traction of AI computing power, the demand for high-end storage chips is growing exponentially. In addition, the concentration of production capacity towards high-end products such as HBM is exacerbating structural supply shortages, leading to a global storage chip market currently experiencing a "bull market" driven by the supply-demand gap. During this industry opportunity period, many companies in the storage industry chain are seeking to expand by leveraging the capital market, including GigaDevice Semiconductor Inc. (603986.SH), which has achieved dual listing in A+H shares, Montage Technology (688008.SH), Shenzhen Longsys Electronics (301308.SZ), Biwin Storage Technology (688525.SH), as well as Lijia Storage, Xintianxia, Hongxin Yudian and other companies that have submitted applications for their first listing to the Hong Kong Stock Exchange. On January 27th, Beijing Xingchen Tianhe Technology Co., Ltd. (referred to as "Xingchen Tianhe"), which focuses on providing enterprise-level AI storage solutions, submitted an application for listing on the Main Board of the Hong Kong Stock Exchange, with Huatai International as its exclusive sponsor. According to the Zhaoshi Consulting report, based on the installed capacity in 2024, Xingchen Tianhe holds a 10.4% share in the Chinese distributed AI storage solution market, making it the second largest provider of distributed AI storage solutions in China and the largest independent provider of distributed AI storage solutions. In terms of performance, Xingchen Tianhe had revenues of 167 million yuan and 172 million yuan in 2023 and 2024 respectively, with losses of 181 million and 84.18 million yuan during the period. In the first three quarters of 2025, Xingchen Tianhe's revenue was 195 million yuan, a year-on-year increase of 65.4%, with a net profit of 8.11 million yuan, reversing a loss into a profit, compared to a loss of 65.506 million yuan in the same period of 2024. With its continued strong performance and the high demand for storage in the market, will Xingchen Tianhe's IPO in Hong Kong attract funds? Through the company's prospectus, we can find out the details. Eight rounds of fundraising have created a luxurious shareholder lineup, with unlocked selling pressure potentially becoming a future hidden concern Since its establishment in 2015, Xingchen Tianhe has developed over ten years from a storage technology provider into a leading enterprise in the domestic distributed AI storage solution market. Its development can be divided into three stages, each of which has accurately grasped technological trends and continuously built its core competitiveness. The first stage from 2015 to 2017 saw the company entering the private cloud and virtualization market early on, providing stable and reliable cloud infrastructure to customers with storage needs from Beijing Vastdata Technology, establishing a foothold in the market. The second stage from 2017 to July 2022 saw Xingchen Tianhe propose the architecture concept of "one data center, one set of storage," and developed its own data operating system (DataOS) based on this, strengthening support for multiple computing chips at the core level. This enabled the company to provide customers with a unified, intelligent data platform, completing the strategic upgrade from storage products to data platforms. The third stage from July 2022 to the present sees Xingchen Tianhe fully transitioning to AI, focusing on building end-to-end AI data infrastructure. Its independently developed XGFS (high-performance file storage) and XEOS (object storage) have been deeply integrated into AI training and inference processes. The company's AIMesh solution further breaks through data protocol barriers, achieving zero-copy data flow from data lake to training ground, significantly improving the overall efficiency of AI data processing. After ten years of continuous accumulation and strategic evolution, Xingchen Tianhe has ranked second in the Chinese distributed AI storage solution market in terms of installed capacity in 2024, firmly positioned in the industry's leading camp. This growth trajectory is not only the result of the company's keen insight into technological trends but also the continuous support of capital. According to the prospectus, as of the end of 2023, Xingchen Tianhe has completed a total of eight rounds of fundraising, with the latest being the F+ round of financing in December 2022, valuing the company at 4.58 billion yuan. After completing up to eight rounds of fundraising, Xingchen Tianhe's current shareholder structure has shown a pattern of founder dominance, professional investment institutions driving force, industry capital empowerment, and state-owned enterprise support. The founding shareholders and persons acting in concert collectively hold a 25.72% stake. Professional investment institutions include Boyu Capital, Chaojun (representing Northern Light Venture Capital), QM188 Limited (representing Qiming Venture Partners), Junlian Shengyuan (representing Junlian Capital), and China International Capital Corporation (CICC) with shareholding ratios of 9.62%, 7.18%, 4.92%, 3.85%, and 3.85%, respectively. Industry capital includes Tencent's Tencent Guangxi, Nanjing Xingcheng (representing Hundsun Technologies Inc.), and Beijing Kunlun (representing Kunlun Tech), with shareholding ratios of 5.64%, 0.77%, and 1.15% respectively. State-owned background shareholders include Shenzhen Jianwei (representing State Investment Fund) and Shanghai Guoxin, with shareholding ratios of 7.5% and 1%. It is clear that Xingchen Tianhe's shareholder structure is considered luxurious, reflecting the positive outlook of various capital entities for the company's future development. However, it is worth noting that Xingchen Tianhe's equity structure is relatively decentralized, and once the lock-up shares are released after listing, it may significantly impact the company's stock price. Three major dimensions working together to support continuous performance improvement According to the prospectus, as of now, Xingchen Tianhe mainly provides two core services to enterprises: AI storage solutions and supporting AI storage services. The solutions contribute the majority of revenue, with a share of 80% in 2024, while the storage services account for 20% of the revenue. Specifically, Xingchen Tianhe's AI storage solutions aim to build a unified data foundation covering the entire AI application lifecycle from data preparation to model training and inference. The solution consists of two key components: AI data lake storage and AI training and inference storage, with revenue shares of 40.7% and 39.3% in 2024, respectively. The AI data lake storage is responsible for integrating, cleaning, and centrally managing massive original unstructured data (such as images and videos) as the starting point of the AI data lifecycle. The AI training and inference storage target the model training and inference stages, providing high throughput, low latency data supply to GPU clusters and other high-performance computing resources, aiming to increase GPU utilization and reduce inference costs. The two components work together to ensure efficient data flow within enterprises. The AI storage services provided by Xingchen Tianhe refer to post-deployment maintenance, support, and renewal support for AI storage solutions to ensure the long-term stable operation of customer systems. In terms of performance, the significant narrowing of Xingchen Tianhe's net loss in 2024 was mainly due to three factors. Firstly, there was robust revenue growth. Benefiting from the continuous release of downstream AI transformation demand, the company's orders for AI data lake storage solutions increased from 257 to 364 in 2024, driving a 13.02% year-on-year growth in revenue for this business to 70.141 million yuan. Combined with moderate growth in AI storage services revenue, the company's total revenue for the year increased by 3.38% to 172 million yuan. Secondly, there was a significant increase in gross margin. The overall gross margin in 2024 reached 63.7%, an increase of over 8 percentage points from 55.4% in 2023. This was mainly due to the improved gross margin brought about by the optimization of product structure of AI storage solutions and the increase in gross margin of AI storage services as economies of scale manifested. Finally, there was effective cost control. Under the strategy of reducing costs and increasing efficiency, the proportion of total operating expenses to revenue decreased to 119.79% in 2024, a significant optimization from 174% in 2023. The effective control of expenses directly accelerated profit release, providing crucial support for the narrowing of losses. In the first three quarters of 2025, although Xingchen Tianhe's overall gross margin decreased by 6 percentage points to 63.7% due to the company's adoption of competitive pricing to maintain sales volume, the significant revenue growth and continuous decrease in total operating expenses resulted in the company reversing losses into profits during the reporting period. The significant revenue growth during the period was due to the accelerated AI transformation in downstream industries, driving a surge in demand for advanced storage solutions. Orders for AI data lake storage solutions increased from 255 to 371, and orders for AI training and inference storage solutions increased from 201 to 294 during the period, resulting in revenue growth of 106.8% and 54.5% respectively for the two major businesses. Coupled with a 14% increase in AI storage services revenue, these three driving forces led to a significant 65.4% increase in Xingchen Tianhe's revenue in the first three quarters of 2025. The proportion of total operating expenses to total revenue during the period also decreased from 130.99% in the first three quarters of 2024 to 65.52%, with the halving of the proportion of total operating expenses indicating a catalyst for the accelerated release of Xingchen Tianhe's profits. The continuous improvement in Xingchen Tianhe's performance is the result of the three aspects of revenue growth, gross margin improvement, and cost control working together, demonstrating a clear financial improvement trajectory and operational efficiency enhancement for Xingchen Tianhe. Chip price hikes drive the hot AI storage trend, facing tough competition from giants In fact, the significant revenue growth in the first three quarters of 2025 for Xingchen Tianhe has already shown that it is benefiting from the "bull market" in storage chips. Looking at the current storage chip price hike trend, the storage market began to bottom out and recover at the end of 2024, laying the groundwork for this round of price increases. By April 2025, the industry saw the first exploratory round of price hikes. Just five months later, the price hike wave entered a second round of large-scale increases, with prices rapidly rising again in the fourth quarter due to supply-demand tensions. For specific products, the unit price of 256GB DDR5 server memory modules for data centers has exceeded 40,000 yuan, and consumer-grade 16GB DDR5 laptop memory modules have also surged from over 380 yuan to nearly 1,400 yuan in just a few months. Research institutions widely predict that in the first quarter of 2026, contract prices for storage chips will continue to rise rapidly, with DRAM prices expected to increase by 55%-60% and NAND Flash prices potentially in the 33%-38% range. Looking at a longer-term perspective, the overall shortage of supply in the market is expected to persist at least until the end of 2026, with UBS predicting that the shortage of DRAM may continue until the first quarter of 2027. Due to the high technical barriers and slow capacity expansion, high-end HBM is expected to remain in high demand until 2028. The CEO of Micron Technology also expects the tight supply situation for HBM to continue until 2028. The bull market in storage chips is expected to benefit Xingchen Tianhe, as the company does not produce hardware chips but provides software-defined distributed AI storage solutions. Its core value lies in helping customers manage and utilize existing heterogeneous storage hardware resources more efficiently and flexibly through the intelligent data operating system (DataOS), improving storage utilization and performance. When the cost of purchasing new hardware storage (CAPEX) increases sharply due to chip price hikes, purchasing Xingchen Tianhe's solutions to optimize the efficiency of existing storage infrastructure, extend its lifecycle, and reduce total cost of ownership (TCO) becomes a highly attractive alternative or supplementary strategy. This means that between "buying more expensive hardware" and "making existing hardware work smarter," the latter demand is strongly stimulated, benefiting Xingchen Tianhe. However, the potential risks and challenges facing Xingchen Tianhe are also evident. Firstly, there is intense market competition. While Xingchen Tianhe holds a 10.4% market share in the Chinese distributed AI storage solution market, ranking second based on installed capacity in 2024, the first place is held by Huawei with a market share of 14.7%, over 4 percentage points higher than Xingchen Tianhe. The third and fourth places are held by Inspur Electronic Information Industry and New H3C respectively, with market shares of 9.3% and 9.2% respectively, which are closer to Xingchen Tianhe. Directly competing with comprehensive technology giants such as Huawei, Inspur Electronic Information Industry, and New H3C, these competitors can typically provide full-stack solutions from hardware to software, and have significant advantages in optimizing hardware coordination and brand channels. In addition, public cloud service providers like Alibaba Cloud are also important market competitors, as their cloud storage services create alternative pressure for on-premise solutions. As an independent software vendor, Xingchen Tianhe needs to continuously demonstrate the unique value of its solutions compared to the ecosystems of these giants. Secondly, amidst intense competition with giants, Xingchen Tianhe's research and development (R&D) expenses have been continually decreasing, which is not conducive to the company's long-term steady development. According to the prospectus, the company's R&D expenses were 110 million, 87.959 million, and 57.461 million yuan in 2023, 2024, and the first three quarters of 2025 respectively, accounting for 65.9%, 51%, and 29.5% of total revenue, showing a clear downward trend in both absolute value and proportion of R&D expenses. While adopting a strategy of "pursuing profitability and efficiency" during the IPO process is a common practice to optimize financial statements, it also signifies a slowdown and interruption in the pace of technology investment. This is not conducive to building technological competitiveness and if R&D spending is not increased in the future, it may gradually erode the company's product competitiveness. Additionally, Xingchen Tianhe's AI storage solutions are mainly delivered in two forms: all-in-one machines and pure software. All-in-one machines require the purchase of upstream storage hardware, and if the proportion of customers purchasing all-in-one machines increases significantly, and Xingchen Tianhe cannot effectively transfer the upstream hardware costs to downstream customers, it may exert pressure on the company's gross margin, leading to significant cost control pressure. Overall, Xingchen Tianhe's performance improvement trajectory, IPO prospects, and challenges reflect the dynamic landscape of the storage industry as it navigates through the opportunities and risks presented by the current market conditions.