In-Depth Analysis of Xiechuang Data's (300857.SZ) RMB 21.2 Billion AI Computing Power Gamble
Unlock More Features
Login to access AI-powered analysis, deep research reports and more advanced features

About us: Ginlix AI is the AI Investment Copilot powered by real data, bridging advanced AI with professional financial databases to provide verifiable, truth-based answers. Please use the chat box below to ask any financial question.
Related Stocks
Based on the in-depth research above, I now provide you with a comprehensive analysis report on Xiechuang Data’s RMB 21.2 billion server procurement plan:
Xiechuang Data disclosed server procurement plans in five batches in 2025, with a cumulative maximum amount of
| Announcement Date | Procurement Amount | Main Purpose |
|---|---|---|
| Early March 2025 | ≤ RMB 3 Billion | High-end GPU computing power server cluster |
| May 28, 2025 | ≤ RMB 4 Billion | Computing power infrastructure construction |
| August 16, 2025 | ≤ RMB 1.2 Billion | Intelligent computing power center construction |
| October 22, 2025 | ≤ RMB 4 Billion | Domestic and international computing power network layout |
| December 17, 2025 | ≤ RMB 9 Billion | Improvement of global computing power network |
As of September 30, 2025, the company’s total assets were
Financial data shows that the company’s asset-liability ratio skyrocketed from
In the first half of 2025, intelligent computing power products and services achieved revenue of
- The FCloud Intelligent Agent Training and Inference Platform has been launched and is in operation
- The FCloud Omnibot Embodied Intelligent Robot Development Platform was released
- Obtained NVIDIA NCP (NVIDIA Cloud Partner) qualification
- Established distributed computing power networks in Shanghai, Ningbo, Chengdu, Ulanqab, as well as in Southeast Asia and the United States
From the perspective of DuPont Analysis, Xiechuang Data’s ROE increased from
| Indicator | 2022 | 2024 | Change |
|---|---|---|---|
| Net Profit Margin | 4.16% | 9.27% | +5.11pct |
| Equity Multiplier | 1.96 | 2.26 | +0.30 |
| Total Asset Turnover | Remained basically flat | Remained basically flat | — |
The gross profit margin increased from
Current Valuation Indicators:
- P/E: 57.08x(TTM)
- P/B: 12.10x
- P/S: 6.48x
Compared with the industry average:
- Industry average P/E is about 35x, Xiechuang Data has a 63% premium
- Industry average P/B is about 3.5x, Xiechuang Data has a 246% premium
From a technical perspective, the KDJ indicator has a K value of 80.3 and D value of 79.6, RSI is in the overbought zone, and the stock price fluctuates in the range of
| Risk Indicator | Value | Risk Assessment |
|---|---|---|
| Asset-Liability Ratio | 78.33% | High Leverage |
| Operating Cash Flow | -RMB 1.374 Billion | Net Outflow |
| Monetary Funds | RMB 1.455 Billion | Relatively Limited |
| Interest-Bearing Liabilities | RMB 10.9 Billion | Debt Repayment Pressure |
The company is alleviating capital pressure through the following methods:
- Bank Credit: Applied for a credit line ofRMB 20.5 Billionin 2025
- Hong Kong Stock IPO: Plans to raise funds for AI computing power infrastructure construction
- Shareholder Loan: Upper limit ofRMB 2 Billionfor borrowing from controlling shareholder Xiechuang Wisdom
- Upstream: Changes in NVIDIA GPU supply pattern (recent concerns about market share)
- Midstream: Inspur Information, NingChang, and H3C occupy over 50% of the domestic AI server market
- Downstream: Large cloud service providers show a clear trend of building their own computing power centers
- Ren Zhengfei: Computing power surplus is an inevitable trend
- Joseph Tsai: Need to be vigilant against the risk of data center construction bubbles
- Geng Kangming: The next three years are a good time to enter the computing power sector [2]
- NVIDIA NCP Qualification: Subsidiary Aojia Software obtained it in October 2024, ensuring computing power supply
- “Storage-Computing-Optical” Collaborative Strategy: Self-developed storage chips + layout of optical chips/optical modules
- Customer Resources: Bound with key customers such as Lenovo, the top customer accounts for39.44%of sales
- First-Mover Advantage: One of the listed companies in China that early laid out computing power leasing
| Scenario | Key Assumptions | Impact on Valuation |
|---|---|---|
Optimistic |
Sustained strong demand for computing power, full orders, capacity utilization >80% | Current valuation is reasonable or undervalued |
Neutral |
Steady growth in computing power demand, capacity utilization 60-70% | Valuation is overstated, risk of correction |
Pessimistic |
Computing power surplus leads to price wars, capacity utilization <50% | Valuation is significantly overstated |
- Revenue Growth Verification: In the first three quarters of 2025, revenue reachedRMB 8.331 billion, with a year-on-year increase of 54.43%
- Order Certainty: The company stated that it has sufficient orders on hand, mainly long-term fixed-price contracts
- Business Transformation Effect: The gross profit margin of the intelligent computing power business is 20.6%, higher than that of the traditional storage business
- Policy Support: AI computing power is a national strategic emerging industry
⚠️
⚠️
⚠️
⚠️
- Supporting Factors:
- The gross profit margin of the computing power business is 20.6%, higher than that of traditional businesses
- 100% revenue growth rate, in a rapid growth stage
- Bound with NVIDIA resources, with first-mover advantage
- Risk Factors:
- The asset-liability ratio of 78.33% is at an extremely high level
- The RMB 21.2 billion procurement amount far exceeds the company’s total assets
- P/E ratio of 57x with a significant premium, requiring strong growth support
- Risks of intensified industry competition and computing power surplus
- Short-term: Technical indicators show overbought status, it is recommended to wait and see or reduce positions moderately
- Mid-term: Need to pay attention to order delivery and capacity utilization of the computing power business
- Long-term: If it can successfully transform into a “storage + computing power” dual-driven model, the valuation will have support; but need to be vigilant against capital chain risks
[1] NetEase Finance - “Xiechuang Data Bets RMB 21 Billion on AI!” (2026-01-08)
[2] Sina Finance - “Xiechuang Data’s RMB 21.2 Billion Gamble on Computing Power: Can Geng Kangming Catch the Wind Again?” (2025-12-24)
[3] 21st Century Business Herald - “Xiechuang Invests Another RMB 4 Billion in Servers; Computing Power Leasing May Enter a High Prosperity Period” (2025-10-22)
[4] Sohu Finance - “RMB 1.455 Billion in Monetary Funds Leverages RMB 21.2 Billion Procurement Order” (2025-12-17)
[0] Jinling AI Financial Database (Real-time Market and Technical Analysis Data)

Insights are generated using AI models and historical data for informational purposes only. They do not constitute investment advice or recommendations. Past performance is not indicative of future results.
About us: Ginlix AI is the AI Investment Copilot powered by real data, bridging advanced AI with professional financial databases to provide verifiable, truth-based answers. Please use the chat box below to ask any financial question.
