Computing Power Breakthrough Sparks Application Explosion: Public Companies Enter AI Dividend Realization Period

In recent times, the densely issued 2025 annual reports and performance forecasts from listed companies show that AI’s catalytic effect on listed-company performance is turning from a dream into reality. From the compute power infrastructure supporting large-model training, to application solution strategies that empower industries across the board, and then to cost-reduction and efficiency-improvement tools that enhance internal operational efficiency, AI technology is penetrating layer by layer along the industrial chain, becoming the core engine driving revenue growth and profit improvement for enterprises.

Performance Soars Upstream in the Compute-Power Industry

The surge in AI compute power has driven a sharp spike in memory storage prices, delivering an exceptionally strong catalyst for storage-related listed companies. As Xiaomi’s president, Lu Weibing, said at Xiaomi’s March 24 earnings call, “I was already an aggressive person when it comes to anticipating memory price hikes, but judging from the actual outcome, the result turned out to be even more aggressive than my original judgment.”

Judging from listed companies’ performance, BOWEI Memory’s revenue reached 11.3B yuan in 2025, up 68.82%, with net profit of 853 million yuan, up 429.07%. The company stated that, with the rapid growth of the AI emerging edge market, products in related fields quickly scaled up, and in 2025 its revenue from AI emerging-edge storage products was about 1.75B yuan. Jiangbo Long expects that in 2025 it will achieve net profit of 1.25 billion yuan to 1.55 billion yuan, up 150.66% to 210.82%, and revenue of 22.5 billion yuan to 23.0 billion yuan, representing a substantial year-on-year increase.

The training and inference of AI large models has put forward unprecedented demand for the supply of compute-power infrastructure. This directly translates into strong procurement demand for core hardware such as chips, servers, and optical modules.

Domestic AI chip companies are accelerating their deployment. Zheng Zhun, chairman of Guoxin Technology, told reporters that the company adheres to the “RISC-V CPU + AI NPU” technology route. Its edge AI chip CCR4001S has achieved large-scale application in the intelligent commercial air-conditioning field, with shipments exceeding 100k units. “In the next 3 to 5 years, the company will use four major business units, such as its cloud AI chip business division, as engines, integrating AI and quantum security technologies to enhance competitiveness.”

The 2025 annual report disclosed by Foxconn Industrial Internet shows that AI is the core driver of performance. Benefiting from the expansion of compute-power investment by global cloud service providers, the company’s cloud computing business revenue was 100k yuan, up 88.70% year-on-year. AI server revenue increased by more than three times year-on-year. Revenue of high-speed switches above 800G surged 13 times, becoming a new growth engine. For the full year, the company’s revenue reached 602.68B yuan, up 48.22% year-on-year, and net profit increased 51.99% year-on-year.

In the core position of the compute-power industrial chain, Jinjixu Chuang’s 2025 performance also corroborates strong demand for compute power. During the reporting period, the company’s shipment volume grew rapidly. The proportion of high-speed optical modules continued to rise. The company achieved revenue of 902.89B yuan, up 60.25% year-on-year; net profit was 38.24B yuan, with a year-on-year increase of 108.81%.

In the upstream base materials segment, Shengyi Technology benefited from the growth in demand for high-performance copper-clad laminates. In 2025, both volume and pricing of copper-clad laminates increased, and product structure was optimized, resulting in an improved gross margin. In 2025, Huzheng New Materials achieved net profit of 277 million yuan, turning losses into profits. The company built three production bases in Hangzhou and Zhuhai, enabling iterative upgrades to its copper-clad laminate manufacturing capability.

Downstream Tech Companies Drive the Commercialization of AI Applications

At the levels of technological innovation and product implementation, AI’s performance-catalyzing effect is also evident. Among them, internet giants—leveraging their technological and scenario advantages—have become the main drivers of monetization.

On March 19, Alibaba Group released its 2026 fiscal third-quarter earnings report. AI and cloud computing businesses became the most eye-catching growth engine in this earnings report. The data shows that in that quarter, Alibaba Cloud Intelligence Group’s revenue was 10.8B yuan, up 36% year-on-year. On the consumer side, Alibaba’s “Qianwen App” surpassed 300 million monthly active users this February, helping broaden the adoption of AI applications. Alibaba Group CEO Wu Yongming announced that over the next five years, Alibaba Cloud and AI’s annual commercialized revenue will exceed $100 billion, with a compound annualized growth rate of about 47%.

Tencent’s 2025 revenue reached 751.77 billion yuan, up 14%. Profit attributable to equity holders was 43.28B yuan, up 16%. Among them, Tencent Cloud achieved large-scale profitability, mainly benefiting from the continuous rise in enterprise AI demand and contributions from its leading-market PaaS and SaaS products. To consolidate its advantage, Tencent continued to increase its AI investment. In 2025, total capital expenditures reached 79.2 billion yuan, and R&D expenditures were 85.75 billion yuan.

In 2025, SenseTime Group positioned AI as the core driving force for performance improvement and business breakthroughs. Revenue from its generative AI business increased 51.0% year-on-year to 224.84B yuan, accounting for as much as 72.4% of total revenue, driving the group’s total revenue up 32.9% year-on-year to 36.3B yuan. At the same time, the company’s net loss narrowed sharply by 58.6% year-on-year. In terms of commercial deployment, SenseTime’s AI applications have already penetrated both the C-end and B-end, building an AI ecosystem matrix.

Listed companies focused on large models also announced their latest performance. MiniMax founder and CEO Yan Junjie said that in February 2026, the company’s annual recurring revenue had already exceeded $150 million. Its open-platform products targeting enterprise customers and individual developers saw the number of newly registered users that month reach more than four times the figure from December 2025. Yuan Guohua, chairman of Shanghai Guotou, said that China has abundant application scenarios, a large volume of edge-side devices, and rapidly growing high-quality data, providing unique soil for AI development.

AI Helps Improve Operations and Boost Efficiency

With the innovation and experimentation of AI technologies and their practical applications, listed companies are integrating AI into their products and day-to-day operations, achieving improved operating efficiency, notable performance improvement, and driving changes in business models.

AI became the key for Huia Media to turn a loss into a profit in its 2025 fiscal year. Relying on its self-developed AI multi-agent and full-domain ad placement system, the company steadily increased its business scale, delivering revenue growth of more than 10%. At the same time, by using AI to empower material production and operational management, it drove administrative expenses down by more than 10% year-on-year, while continuously improving the quality of profitability.

Qiniu Intelligence’s annual report further reflects AI’s impact on revenue mix and R&D efficiency. In 2025, the company’s AI-related revenue was 437 million yuan, accounting for 24.7% of total revenue, becoming a core growth driver. Meanwhile, the company used AI tools to significantly improve R&D efficiency, leading to a 9.8% year-on-year reduction in R&D costs. Under the combined effects of effective cost control and revenue growth, its adjusted net loss decreased by 60.9% year-on-year.

The cybersecurity industry also integrates AI into daily operations. Based on annual report data, 360 is expected to achieve net profit attributable to shareholders of about 213 million yuan to 318 million yuan in 2025, meaning it will turn a loss into a profit. Qianxin’s performance express report shows the company expects full-year 2025 sales cash collections of about 50.15B yuan, up about 8.04% year-on-year.