In-depth Analysis of AI Investment Strategy of RichInfo Co., Ltd. (300634.SZ)

Based on the collected data and market information, I provide you with an in-depth analysis report on the RMB 1.46 billion AI investment of RichInfo Co., Ltd.

---

RichInfo Co., Ltd. currently demonstrates strong financial conditions and market performance [0]. As of January 14, 2026, the company’s market capitalization reached USD 16.12 billion, with the current stock price at RMB 35.95, representing a 74.09% increase from a year ago and a 39.18% growth in the past three months. From the perspective of financial indicators, the company’s current ratio is as high as 5.68, indicating extremely ample working capital and short-term solvency; the net profit margin remains at a healthy level of 13.00%, with ROE at 8.17%, and the overall financial risk is assessed as “low risk” category [0].

From the perspective of business growth trajectory, RichInfo’s total operating revenue grew from RMB 887 million in 2021 to RMB 1.652 billion in 2024, while net profit attributable to parent company excluding non-recurring gains and losses increased from RMB 130 million to RMB 218 million, basically achieving double growth. In the first half of 2025, the company’s AI and computing power business revenue reached approximately RMB 177 million, accounting for nearly 20% of total revenue, indicating that AI-related products and services have formed substantial revenue contributions [1]. This data is of great significance—it proves that the company is not merely engaging in concept speculation, but has truly achieved commercialization of its AI business.

According to the company’s announcement, the raised funds will be allocated to three core areas [1][2]:

Project Name	Investment Focus	Strategic Significance
Intelligent Computing Center Construction Project	Add 12,000P computing power services and build a 10,000-card cluster AI intelligent computing center	Consolidate the computing power base to meet the needs of large model training and inference
R&D and Upgrade of Rich AIBox Platform	Expand the carrying capacity of AI middle-office agents and strengthen AI ecosystem construction	Enhance platform capabilities and build a technological moat
Development of Enterprise-level AI Agent Applications	Develop enterprise knowledge bases, intelligent data query, intelligent customer service, smart marketing, AI voice agents, etc.	Deepen industry applications and realize the upgrade from “tool assistance” to “independent decision-making”

This investment structure reflects RichInfo’s complete AI ecosystem layout of “computing power base + development platform + industry applications” [3]. It is worth noting that the company also has a clear layout in the field of embodied intelligence. It recently plans to subscribe for fund shares to invest in robot companies, while focusing on the layout of core technologies such as voiceprint replication and hyper-realistic speech synthesis in the direction of AI voice agents.

According to IDC data, the market size of China’s accelerated server market reached USD 3.1 billion in the first half of 2023, a year-on-year increase of 54%, and is expected to reach USD 16.4 billion in 2027 [4]. Global AI server shipments continue to grow at a high speed, with an expected 1.2 million units in 2023, a year-on-year increase of 38.4%, and a compound annual growth rate of 22% from 2022 to 2026 [4].

The construction of intelligent computing centers is entering an acceleration period. According to incomplete statistics, there have been over 100 intelligent computing center projects in 2024, showing three significant characteristics [5]:

• Scale Upward Transition
  : The overall computing power planning of a single government intelligent computing center is close to the 1,000P level, while the scale of intelligent computing centers operated by carriers/internet companies basically reaches 10,000 cards and above
• Increased Capital Investment
  : For example, China Mobile’s RMB 19.1 billion AI server tender in 2024
• Rising Demand for Professional Operation Capabilities
  : System architecture deployment optimization, computing stability guarantee, and efficient operation and maintenance of large-scale clusters have become key factors

The current intelligent computing center market presents a “multi-level, differentiated” competitive landscape:

┌─────────────────────────────────────────────────────────────┐
│                    Competitive Landscape of Intelligent Computing Centers                          │
├─────────────────────────────────────────────────────────────┤
│  First Tier: Internet Giants (ByteDance/Alibaba/Tencent/Baidu)                  │
│           Computing Power Scale: 100,000-card level                                 │
│           Advantages: Capital strength, technological accumulation, application scenarios                 │
├─────────────────────────────────────────────────────────────┤
│  Second Tier: Carrier Intelligent Computing Centers (China Mobile/China Telecom/China Unicom)               │
│           Computing Power Scale: 10,000-card level                                   │
│           Advantages: Network resources, government-enterprise relationships, customer channels                 │
├─────────────────────────────────────────────────────────────┤
│  Third Tier: Professional AI Service Providers (RichInfo Co., Ltd., etc.)                     │
│           Computing Power Scale: 1,000-card to 10,000-card level                             │
│           Advantages: Industry know-how, customized services, flexible response           │
├─────────────────────────────────────────────────────────────┤
│  Fourth Tier: Local Government Intelligent Computing Centers                                  │
│           Computing Power Scale: 1,000P level                                    │
│           Advantages: Policy support, localized services                         │
└─────────────────────────────────────────────────────────────┘

1. Differentiated Positioning
   : Different from the “large and comprehensive” approach of internet giants and carriers, RichInfo focuses on enterprise-level AI services, and is dedicated to embedding AI into enterprise business processes [3]. The company clearly states that “large model manufacturers are good at algorithms, cloud manufacturers are good at computing power, while RichInfo understands enterprise business logic best.”

2. Proven Delivery Capability
   : The company has built intelligent computing centers with a scale of over 10,000 cards for carriers, financial institutions, intelligent driving enterprises, etc., with mature large-scale delivery and optimization capabilities [3]. This “proven” capability is difficult for new entrants to replicate quickly.

3. Platform Advantages
   : The Rich AIBox platform supports multilingual, multi-scenario, multiple agent construction paradigms, and multi-agent collaboration frameworks, providing enterprises with a modular, rapid-deployment AI development environment [3].

4. Cost Efficiency
   : The company emphasizes heterogeneous computing power scheduling and low-cost optimization capabilities, which can reduce the overall AI investment cost of enterprises while ensuring high performance, which is attractive to current budget-sensitive enterprise customers.

1. Scale Gap
   : 12,000P (approximately 1200 PFlops) still has an order of magnitude gap compared to internet giants (10,000-card clusters), with limited competitiveness in scenarios such as large model training.

2. Capital Pressure
   : The RMB 1.46 billion investment is moderately sized for intelligent computing center construction, but continuous investment is still required for subsequent operation and iterative upgrading.

3. Chip Supply
   : Against the backdrop of localization replacement, improving the localization rate of computing power is a key variable, and the company needs to ensure the stability and performance of chip supply [4].

RichInfo’s competitive advantage moat can be evaluated from the following dimensions:

Moat Type	Current Assessment	Construction Path
Technological Barriers	⭐⭐⭐⭐	Multi-agent scheduling architecture, heterogeneous computing power scheduling capabilities, integrated with mainstream large models such as DeepSeek
Industry Know-how	⭐⭐⭐⭐⭐	Deeply engaged in vertical fields such as communications/transportation/energy/finance/government affairs, accumulated a wealth of scenario experience
Customer Relationships	⭐⭐⭐⭐⭐	Core partner of carriers, awarded A-level information disclosure evaluation by the Shenzhen Stock Exchange for 5 consecutive years
Platform Ecosystem	⭐⭐⭐⭐	Continuous iteration of the Rich AIBox platform, joined China Mobile’s “Juxing Plan”
Economies of Scale	⭐⭐⭐	The 12,000P scale is moderate, and further expansion is needed to form scale effects

1. Project Implementation Progress
   : The construction progress, actual utilization rate, and customer acquisition of the 12,000P intelligent computing center will directly affect investment returns.

2. AI Application Penetration
   : The company’s AI applications have been implemented in industries such as telecommunications, energy, and transportation, but there is still room for improvement in the depth and breadth of penetration.

3. Technological Iteration Capability
   : AI technology is evolving rapidly, and the company needs to continuously follow up on emerging large model technologies such as DeepSeek to maintain platform advancement.

4. Talent Reserve
   : The operation of large-scale intelligent computing centers requires professional AI engineering teams, and talent competition is a key challenge.

Risk Type	Risk Description	Risk Level
Technological Risk	AI technology evolves rapidly, which may lead to the obsolescence of the invested technical route	Medium
Market Competition	Ascend-based integrated machine manufacturers (such as Huakun Zhenyu, Kunlun Technology, etc.) are rising rapidly [4]	Medium
Policy Risk	US chip export control policies may be further tightened, affecting chip supply	Medium-High
Project Risk	The construction progress and actual utilization rate of the intelligent computing center may fall short of expectations	Medium
Valuation Risk	The current P/E ratio of 69.66 times is at a relatively high level, and the stock price faces high volatility risk	Medium-High

RichInfo’s RMB 1.46 billion AI investment has the potential to be converted into sustainable competitive advantages, but the following key points need to be noted:

1. Correct Strategic Direction
   : The investment direction aligns with the national policy guidelines of “AI+” and “moderately advanced” construction of computing power infrastructure, as well as the industry trend of AI shifting from “model competition” to “implementation competition” [1].

2. Clear Differentiated Positioning
   : The company clearly focuses on enterprise-level AI agent solutions, leveraging its core advantage of “understanding enterprise business logic” to avoid direct competition with internet giants.

3. Solid Capability Foundation
   : It has achieved commercialization of AI business (accounting for 20% of revenue), has experience in delivering 10,000-card-level intelligent computing centers, and its platform has integrated mainstream large models such as DeepSeek.

4. Challenges Remain
   : The 12,000P computing power scale still lags behind leading manufacturers, and continuous investment is required to maintain competitiveness; the current valuation level is relatively high.

: RichInfo has in the , with its core advantages lying in deep industry understanding and platform-based delivery capabilities, rather than computing power scale. Investors are advised to focus on key indicators such as project implementation progress, sustainability of AI business revenue growth, and platform user activity.

---

Chart Description: The chart above shows RichInfo’s stock price trend, investment amount allocation, comparison of intelligent computing center computing power scale, and competitive advantage analysis. Data shows that the company’s stock price has increased by 39.18% in the past three months, with investment focusing on intelligent computing center construction (accounting for approximately 70%), and it has relative advantages in industry applications and platform capabilities.

---

[0] Jinling AI - Company Profile and Financial Data of RichInfo Co., Ltd. (300634.SZ)

[1] 21st Century Business Herald - “Shanghai Weixing New Materials, Yobond Ceiling, Guangyun Technology Prompt Risks; CITIC Securities’ Net Profit Exceeded RMB 30 Billion Last Year | Announcement Highlights” (https://www.21jingji.com/article/20260114/herald/e3e6804686334d4ac3da3a144475b534.html)

[2] Cailianshe - “Breaking! 14-consecutive-listed Packaging and Printing Leading Stock Releases Performance Forecast | Post-market Announcement Highlights” (https://www.cls.cn/detail/2257950)

[3] Official Website of RichInfo Co., Ltd. - “Bai Lin, CEO of RichInfo: From ‘Local Deep Cultivation’ to ‘AI+Going Global’ — Building a Global Ecosystem with Platform Capabilities” (https://www.richinfo.cn/page/newsdetail.html?date=2025-10-12)

[4] Shanxi Securities - “Special Report on Computer AI Computing Power” (https://pdf.dfcfw.com/pdf/H301_AP202505271680068238_1.pdf)

[5] IDC - “Computing Power First, the Intelligent Computing Service Market Continues to Grow Rapidly” (https://my.idc.com/getdoc.jsp?containerId=prCHC52894624)

[6] Sina Finance - “RichInfo Plans to Issue Convertible Bonds: Capital Tool Selection Aligns with the Strategic Investment Logic of ‘AI Application + Computing Power’” (https://finance.sina.com.cn/roll/2026-01-14/doc-inhhhtem8208697.shtml)