过去一年，尽管亚马逊云科技（AWS）增长重新加速，且公司利润率最高的商业化业务持续扩 张，亚马逊（AMZN）的股价表现仍落后于整体科技市场。 公司当前估值较历史水平存在折 价，市场担忧的核心在于其 2025 年资本支出预计将接近 1250 亿美元，且 2026 年还可能 进一步增加。市场关注短期现金流的逻辑虽可以理解，但投资者忽略了其商业模式中蕴含的 潜 在期权价值 —— 这一价值将推动 2026 年成为公司基础设施投资转化为利润率扩张的关键之 年。 通过发力定制化芯片（Trainium3）并向服务主导型营收结构转型（目前服务收入占总销售额 的 60%），亚马逊正将自身定位为 低成本算力提供商 ，同时在电商基础上搭建更高价值的商 业化平台。对公司而言，关键在于当推理任务成为主导型工作负载时，现有算力规模所能支撑 的利润率水平与定价能力。 【如需和我们交流可扫码添加进社群】 亚马逊的投资逻辑并非赢得整个人工智能军备竞赛，而是通过 单位经济效益规模化 ，同时维 持多元化的商业化渠道，以对冲业务执行风险。 尽管自由现金流有所压缩，但多元化的现金来 源与持续改善的资本回报率，仍使亚马逊具备长期投资吸引力。 资 本 配 ...

从财报数据来看，亚马逊（Amazon, AMZN）的表现可圈可点。人工智能热潮持续发酵，上季度云计 算业务增速加快，同比增幅达到 20%；营收与利润双双创下历史新高。 目前，公司市值达 2.4 万亿美元，市盈率略高于 32 倍。整体而言，财报数据表现优异，展现出一家健 康成长的企业面貌。但问题在于，所有财务数据与报告反映的都是过去的业绩。而投资股票，看的不是 过去的辉煌，而是未来的预期 —— 且这些预期往往与历史数据毫无关联。 人工智能算力需求爆发的始末 首先，让我们把时间拉回 2023 年，也就是人工智能热潮的起点。彼时，人工智能作为云计算领域的全 新应用场景，催生了该行业的爆发式增长，而当时科技巨头们的算力储备，根本不足以满足激增的市场 需求。 分析师将这一轮需求拆解为两大组成部分： 第一，随着人工智能技术的问世，数千家企业纷纷开始开发面向智能设备的新应用，同时对现有应用进 行人工智能化改造。想必你也注意到，过去几年里，应用商店（App Store 和谷歌应用商店）中涌现出 了大量新应用。而这些应用的运行，正是云计算需求的核心来源。这些企业基本不会自建数据中心，而 是选择在云计算市场租赁算力。 更需注意的 ...

2025 年，亚马逊（Amazon, AMZN）的表现相对疲软，年内涨幅仅为 4%；反观其大型科技 股同行，谷歌（GOOGL）年内涨幅高达 69%，英伟达（NVDA）也录得 34% 的涨幅。在大 型科技股阵营中，每年往往会涌现出一两只领跑个股 —— 比如 2020 年的特斯拉 【如需和我们交流可扫码添加进社群】 数 据 中 心 帝 国 ： 掌 控 A I 实 体 经 济 命 脉 过去两年，市场的目光始终聚焦于英伟达的芯片、微软与谷歌的软件，却严重低估了亚马逊的 潜力 —— 在分析师看来，亚马逊正蓄势主导人工智能实体经济领域。 归根结底，亚马逊的投资价值远不止于电商、数字广告和亚马逊云科技（AWS）这三大业务。 但分析师认为，投资者更应聚焦于亚马逊隐藏的王牌 —— 其庞大的数据中心帝国。 亚马逊云科技常被视为高增长的云计算业务，但许多人忽略了其背后庞大的实体运营规模。目 前，亚马逊的数据中心业务已在全球 50 个国家布局超 900 座设施。更关键的是，其中大量 设施为 主机托管（colocation）数据中心 —— 即亚马逊租用的第三方场地，用以拓展自身业 务覆盖范围。 这一布局带来两大核心优势： 1.构筑深厚 ...

Market Overview - On November 25, the A-share market showed a broad-based rally, with all three major indices rising and trading volume increasing, indicating heightened market enthusiasm [1] - The Shanghai Composite Index rose by 0.87%, the Shenzhen Component Index increased by 1.53%, and the ChiNext Index gained 1.77% [1] - Total trading volume in the Shanghai and Shenzhen markets reached 1.8262 trillion yuan, an increase of 85.8 billion yuan from the previous day [1] - Over 4,300 stocks in the market rose, with 95 stocks hitting the daily limit [1] AI Hardware Sector - On November 25, AI computing hardware experienced a strong rebound, with CPO, PCB, and optical chip concepts leading the market [2] - The CPO concept saw the highest gains, with stocks like Dekeli and Changguang Huaxin hitting the daily limit with a 20% increase [2] - Amazon announced plans to invest up to $50 billion to expand AI and supercomputing capabilities for its AWS government clients, with a project set to break ground in 2026 [2] - Meta Platforms is considering a multi-billion dollar acquisition of Google's TPU for its data center development [2] - Google aims to double its AI computing power every six months to meet rising demand for AI services [2] Flu Prevention Stocks - Flu prevention stocks surged as respiratory infectious diseases enter a high incidence season, with stocks like Jindike hitting the daily limit with a 20% increase [4] - The latest report from the Chinese CDC indicates that flu activity is on the rise and will continue for some time [4] - Investment opportunities in flu vaccines, virus testing, cold medicines, and specific drugs are highlighted due to increased public and market attention [4] Market Outlook - Dongguan Securities anticipates a structural recovery in the A-share market, suggesting that the current adjustment phase may end, paving the way for early positioning for the spring market [5] - The report notes that the end of the year often sees a loosening of liquidity, which could support the market [5] - Guosen Securities indicates that while the market may experience short-term fluctuations, the foundation for a gradual upward trend remains intact due to ongoing global tech investment enthusiasm and supportive policies [5][6]

Core Insights - Amazon's data center operations are significantly larger than previously understood, with over 900 facilities across more than 50 countries, providing stable and efficient cloud computing services globally [1] Group 1: AWS Infrastructure - Amazon Web Services (AWS) has established large-scale data center hubs in Virginia and Oregon, becoming influential power bases in the industry [1] - AWS operates giant campuses either owned or through long-term leases, which represent only a portion of its global footprint [1] - AWS has deployed server racks in hundreds of colocation facilities, with these leased spaces accounting for 20% of Amazon's total available computing resources as of last year [1] Group 2: Deployment Flexibility - The deployment in colocation facilities varies from lightweight setups with a few server racks to large-scale deployments occupying entire buildings in major cities like Frankfurt and Tokyo [1] - These colocation facilities are divided into multiple zones to meet diverse customer needs, allowing AWS to offer flexible service allocation globally [1] - AWS is one of the largest tenants of colocation data center space, leveraging this advantage for global service flexibility [1] Group 3: Security Considerations - AWS does not disclose the specific locations of individual data centers due to security and operational confidentiality concerns [2]

Group 1 - The U.S. government has launched a new national initiative called the "Genesis Project" aimed at transforming scientific research through artificial intelligence [2] - The Federal Reserve is expected to lower interest rates by 25 basis points in December, with an 82.9% probability of this occurring [2] - Apple is undergoing a restructuring of its sales department, resulting in potential layoffs for dozens of employees, particularly in sales roles targeting enterprise, education, and government clients [2] Group 2 - Major U.S. stock indices saw collective gains, with the Dow Jones up 0.44%, S&P 500 up 1.55%, and Nasdaq up 2.69%, with significant contributions from large tech stocks [3][5] - Notable increases in chip stocks were observed, with Broadcom rising 11%, Micron up nearly 8%, and AMD increasing over 5% [3][5] - Chinese concept stocks also experienced gains, with the Nasdaq Golden Dragon China Index rising 2.82% [3] Group 3 - Amazon announced a significant investment of up to $50 billion to expand its cloud computing capabilities for U.S. government clients, focusing on AI and supercomputing [12] - Industrial Fulian clarified that it has not lowered its profit target for Q4 and that production and shipment are proceeding as planned [7] - Lei Jun, founder of Xiaomi, invested over HKD 100 million to acquire 2.6 million shares, increasing his stake to 23.26% [8]

Core Points - Amazon announced an investment of up to $50 billion to expand artificial intelligence (AI) and supercomputing capabilities for its U.S. government clients through Amazon Web Services (AWS) [1] - The investment plan is set to break ground in 2026, adding nearly 1.3 gigawatts of supercomputing capacity in AWS Top Secret, AWS Secret, and AWS GovCloud regions [1] - Additionally, the company will invest $15 billion to build a new data center in northern Louisiana, designed to have a capacity of 2.4 gigawatts and expected to create 1,100 jobs [1]

Group 1 - The company announced an investment of up to $50 billion to expand artificial intelligence (AI) and supercomputing capabilities for its U.S. government clients through Amazon Web Services (AWS) [1] - The investment plan is set to break ground in 2026, aiming to add nearly 1.3 gigawatts of supercomputing capacity in AWS Top Secret, AWS Secret, and AWS GovCloud regions [1] - Additionally, the company will invest $15 billion to build a new data center in northern Louisiana, designed to have a capacity of 2.4 gigawatts and expected to create 1,100 jobs [1]

Core Insights - Amazon announced an investment of up to $50 billion to expand artificial intelligence (AI) and supercomputing capabilities for its U.S. government clients through Amazon Web Services (AWS) [2] - The investment plan is set to break ground in 2026, aiming to add nearly 1.3 gigawatts of supercomputing capacity across AWS Top Secret, AWS Secret, and AWS GovCloud [2] - Additionally, the company will invest $15 billion to build a new data center in northern Louisiana, designed to have a capacity of 2.4 gigawatts, which is expected to create 1,100 jobs [2]

Core Viewpoint - Amazon's recent layoffs of 14,000 employees are aimed at increasing organizational flexibility, despite the company's strong performance, raising questions about the true motivations behind these cuts [2][4][12]. Layoff Details - Amazon has conducted multiple rounds of layoffs in recent years, including 18,000 in January 2023, 9,000 in March 2023, and additional cuts in various departments [2]. - The most affected group in the latest layoffs appears to be software engineers, with 25% of the 2,300 layoffs in Washington state being from this category [2]. Organizational Restructuring - The company claims that a streamlined structure with fewer layers and clearer responsibilities is necessary to respond more quickly to customer needs and drive business growth [4][8]. - However, the rationale for layoffs targeting lower-level employees rather than management raises doubts about the effectiveness of this explanation [4]. Investment in AI and Data Centers - Following the layoffs, Amazon announced the launch of its largest AI computing platform, Project Rainer, which requires significant capital investment [5][7]. - The company has invested $11 billion in this project, which includes 500,000 self-developed Trainium2 chips, indicating a shift in resource allocation towards AI and data center development [7]. Economic Concerns - The layoffs may be a preemptive measure in response to potential declines in consumer spending, as indicated by trends in other sectors, such as the restaurant industry [12][13]. - Companies like Chipotle have reported decreased consumer spending, particularly among lower-income groups, which could impact Amazon's retail business [12]. - UPS has also announced significant layoffs, suggesting a broader trend of reduced package delivery volumes, which may reflect declining consumer activity [13]. Market Sensitivity - Amazon is particularly sensitive to changes in consumer spending patterns, more so than other tech giants that rely heavily on advertising or enterprise spending [14]. - The company's decision to lay off employees may be a strategic response to anticipated reductions in consumer expenditure across its platform [14].