---
title: "CITIC Securities: Advances in AI technology drive terminal prosperity, with significant potential for AIDC demand"
description: "China Merchants Securities released a research report stating that DeepSeek has suspended API service recharges due to tight server resources. The short-term supply-demand imbalance has led to resourc"
type: "news"
locale: "en"
url: "https://longbridge.com/en/news/228153429.md"
published_at: "2025-02-13T07:55:05.000Z"
---

# CITIC Securities: Advances in AI technology drive terminal prosperity, with significant potential for AIDC demand

> China Merchants Securities released a research report stating that DeepSeek has suspended API service recharges due to tight server resources. The short-term supply-demand imbalance has led to resource constraints, but in the long term, the improvement in AI efficiency will drive the popularization of end applications. Low-cost models will help expand the domestic computing power industry chain. Overseas AI capital expenditures are strong, with companies like Meta and Amazon exceeding expectations in their 2024 financial reports. Trump plans to invest $500 billion in AI infrastructure, and France will also invest €109 billion. Risk warnings include AI terminal demand falling short of expectations

According to the Zhitong Finance APP, China Merchants Securities released a research report stating that recently, due to tight server resources, DeepSeek officially announced the suspension of API service recharge. The current tight server resources of DeepSeek are the result of a short-term supply-demand imbalance. In the long term, the improvement of AI efficiency will inevitably promote the broader application of terminals. The "efficiency improvement → cost reduction → application popularization → demand surge → resource tension" cycle, reminiscent of the Jevons Paradox, may be replayed in the AI industry. In addition, low-cost models represented by DeepSeek are expected to enhance the adaptability of the domestic computing power industry chain and promote the accelerated expansion of domestic computing power platforms.

**The failure to expand capacity in a timely manner is the main reason for the tight server resources.** This time, due to the impending end of preferential pricing (until February 8, 2025), the concentrated recharge by users further exacerbated the server load. DeepSeek has significantly reduced training and inference costs through algorithm optimization (such as mixed expert models, reinforcement learning, etc.), but the failure of hardware expansion to match the surge in user numbers has caused tight server resources, forcing it to suspend API service recharges.

**Overseas AI-related capital expenditure determination is strong.** Recently, overseas CSP companies such as Meta and Amazon have successively released their 2024 financial reports, with Q4 capital expenditure exceeding market expectations. In fiscal year 2025, the capital expenditure plans of various companies have significantly increased, with the main investment direction being AI-related infrastructure construction. In addition, Trump announced The Stargate Project, planning to invest $500 billion in AI infrastructure in the United States. Besides the U.S., France also plans to invest €109 billion in artificial intelligence projects in the coming years. Overall, there is a strong determination and clear direction from overseas governments and mainstream enterprises for investment in AI-related projects.

**Risk Warning:** AI terminal demand may fall short of expectations, corporate capital expenditure may not meet expectations, and trade barriers may intensify

### Related Stocks

- [DPSK.NA - DeepSeek](https://longbridge.com/en/quote/DPSK.NA.md)
- [AMZN.US - Amazon](https://longbridge.com/en/quote/AMZN.US.md)
- [06099.HK - CMSC](https://longbridge.com/en/quote/06099.HK.md)
- [META.US - Meta Platforms](https://longbridge.com/en/quote/META.US.md)

## Related News & Research

| Title | Description | URL |
|-------|-------------|-----|
| OpenAI 新一輪融資或突破千億美元 據報亞馬遜、軟銀、英偉達及微軟參與投資 | OpenAI 即將完成新一輪融資，預計籌集超過 1000 億美元，估值可能超過 8500 億美元。主要投資者包括亞馬遜、軟銀、英偉達和微軟。融資將分階段進行，預計在本年度內完成。亞馬遜可能投資高達 500 億美元，軟銀 300 億美元，英偉 | [Link](https://longbridge.com/en/news/276297991.md) |
| 股神巴菲特再現「神準」操作 亞馬遜急跌前大減倉 再沽蘋果美銀 買入一隻媒體股 | 股神巴菲特在卸任巴郡 CEO 後，繼續調整投資組合。根據 SEC 13F 文件，巴郡大幅減持亞馬遜 770 萬股，持倉減少 77%；同時出售約 5,080 萬股美國銀行和 1,030 萬股蘋果，連續第三季減持蘋果。巴郡首次增持傳統媒體股，購 | [Link](https://longbridge.com/en/news/276191227.md) |
| DeepSeek 灰度測試新一代模型，野村: 訓練與推理成本下降或緩解盈利壓力 | DeepSeek 正在進行新一代模型的灰度測試，預計本月中旬推出 V4 模型。該模型在上下文長度和核心能力上有顯著提升，野村證券認為 V4 將通過底層架構創新推動 AI 應用商業化，而非顛覆現有價值鏈。V4 的發布預計將顯著降低訓練與推理成 | [Link](https://longbridge.com/en/news/275723457.md) |
| 亞馬遜年度營收首超沃爾瑪，零售雙雄加碼 AI 競賽 | 亞馬遜取代沃爾瑪，成為年營收最高的公司。沃爾瑪週四公佈，其最近一個財年的年度營收為 7132 億美元，這略低於亞馬遜的 7169 億美元。兩家公司在人工智能領域採取了不同的策略：亞馬遜大幅加大了對人工智能的投資，而沃爾瑪則更依賴技術合作夥伴 | [Link](https://longbridge.com/en/news/276384959.md) |
| 機構 “最超配” 閃迪，“最低配” 英偉達 | 據摩根士丹利最新的統計：“機構對美國大型科技股的低配程度是 17 年來最大的” 相比 2025 年 Q4 的標普 500 指數權重，“$NVDA 仍然是機構低配程度最大的大型科技股，其次是蘋果、微軟、亞馬遜和博通，而存儲巨頭閃迪則是 “最超 | [Link](https://longbridge.com/en/news/276289765.md) |

---

> **Disclaimer**: This article is for reference only and does not constitute any investment advice.