对于投资者和云厂商而言，TPU最大的价值不仅仅是快，而是利润率。谷歌通过掌控TPU的全栈设计，成功绕开了"英伟达税"。与此同时，Broadcom 的毛利远低于英伟达，这让谷歌能够将算力成本压到极致。从TPU v6到最新曝光的 TPU v7 ，谷歌不仅是在造芯，更是在为即将到来的"AI推理时代"构 建一道几乎不可逾越的护城河。 英伟达盘前下跌3.5%，报道称谷歌人工智能芯片TPU获Meta洽购，价值数十亿美元。 TPU获Meta洽购，价值数十亿美元 在AI算力领域，英伟达似乎是那个不可一世的霸主。但在聚光灯之外，科技巨头谷歌正在用一种更为隐秘却极具破坏力的方式，重新定义AI芯片 的战争规则。 这张王牌，就是谷歌自研的TPU（Tensor Processing Unit）。 如果你以为这只是谷歌为了省钱搞的"备胎"，那就大错特错了。根据最新披露的深度资料，谷歌最新的TPU v7（代号Ironwood）不仅在显存容 量上追平了英伟达的B200，更在能效比上实现了对GPU的降维打击。甚至连黄仁勋本人也曾暗示，在ASIC领域，谷歌TPU是一个"特殊的存 在"。 从TPU v6 (Trillium) 到最新曝光的 TPU ...

一系列股票抛售及会计违规指控，使英伟达(NVDA.O)成为关于AI及相关股票价值的讨论焦点。针对市场质疑，英伟达近日开始反击。天风国际证券分析师 郭明錤也通过他的分析，尝试为英伟达财务报表"正名"。 英伟达发送7页备忘录回应12项质疑 上周末，英伟达投资者关系团队向华尔街分析师私下发送了一份七页备忘录，直接回应了多位持怀疑态度投资者提出的十二项质疑。 备忘录采用公司标志性的绿色字体排版，开篇即回应了迈克尔·伯里（Michael Burry）上周在社交媒体上的一则帖子，该帖批评公司基于股票的薪酬稀释及 股票回购。伯里因在2008-2009年金融危机前精准押注次贷违约而闻名，其事迹被改编成电影《大空头》（The Big Short）。 "客户根据实际寿命和使用模式，将GPU折旧周期设定为4至6年。诸如2020年发布的A100等老款GPU仍可高效运作并产生可观利润，其经济价值 远超过部分评论者所称的2至3年。" 郭明錤：英伟达的财报结果完全符合财报常识和行业现状 英伟达在备忘录中指出："自2018年以来，公司回购股份总额为910亿美元，而非伯里所称的1125亿美元。伯里先生似乎错误地将受限股票单位（RSU）的税 款 ...

今天最好的消息莫过于两国领导的通话，特朗普表示电话沟通"非常顺利"，他已同意在四月访问北 京，并邀请中方领导明年进行国事访问。 这次通话 强化了"稳定" 叙事，对目前国内出海链的企业都算是非常好的消息。 最近谷歌链越来越火： 我们之前也写过几篇关于Google的分析： 写在英伟达业绩前、谷歌十年磨一剑 从英伟达转向谷歌？液冷产业出海信息更新 巴菲特看上谷歌什么了？谷歌国内供应商梳理 说到谷歌链，那就必然绕不开TPU。 TPU 的故事要从 2013 年说起，那时，谷歌的领导层，包括 Jeff Dean 和 Jonathan Ross，发现如果每 位安卓用户每天使用语音搜索功能仅 3 分钟，全球数据中心容量就需要翻倍。这是一个警钟！于 是，谷歌团队在短短 15 个月内，从概念到部署，开发出了第一代 TPU。这款芯片专为 TensorFlow 神经网络设计，针对深度学习中的矩阵乘法进行了优化，解决了传统 CPU 和 GPU 在 AI 任务中的瓶 颈。 2015 年，TPU 开始在谷歌内部应用，比如地图、照片和翻译服务。2016 年，在 Google I/O 大会上正 式亮相。从那时起，TPU 不断迭代，如今已经发展 ...

Core Insights - Nvidia reported third-quarter revenue of $57 billion, reflecting a 62% year-on-year increase [2] - The company is heavily reliant on TSMC for its advanced chips, which poses significant geopolitical risks [5][8] Company Dependency - Nvidia's valuation reached $4.3 trillion, with its key products (H100, H200, Blackwell) dependent on TSMC's facilities in Taiwan [3] - Over 90% of the world's advanced chips are produced in Taiwan, making Nvidia's supply chain vulnerable to geopolitical tensions [4][5] Geopolitical Risks - Tensions in the Taiwan Strait have escalated, with increased military exercises and diplomatic pressure in 2025 [8] - A limited blockade by China could halt TSMC exports, disrupting over 90% of leading-edge chip production globally [8][9] Supply Chain Vulnerability - Nvidia sources 100% of its top-tier GPUs from TSMC, with no alternative sources for advanced production until at least 2027 [7] - The sophisticated packaging required for Nvidia's GPUs is also concentrated in Taiwan, further increasing dependency [7] Market Impact - A disruption lasting six months could halve Nvidia's projected revenue of $300 billion, leading to a $75 billion decrease in earnings [14] - Nvidia shares currently trade at around 43x forward earnings, which could compress significantly in the event of supply chain disruptions [11] Potential Beneficiaries - Companies like Intel and Samsung may benefit from a global re-shoring trend, as every viable fab becomes crucial [15] - ASML and Applied Materials, key suppliers in chip fabrication, will also gain regardless of location due to increased demand for fabrication tools [15]

摘要 英伟达 Q3 财报显示收入同比增长 205%，调整后毛利率达 73.6%，预 计 Q4 收入中值为 650 亿美元，同比增长 65%，调整后毛利率中枢为 75%。公司预计 2026 财年毛利率维持在 70%中段，缓解了市场对其毛 利率受存储涨价和 ASIC 影响的担忧。 英伟达正加强供应链韧性，与台积电合作在美国本土生产 Blackwell 晶 圆，并扩大与富士康、安靠等合作伙伴的美国制造能力。安靠股价盘后 上涨 10%，反映市场积极情绪。 英伟达预计 Blackwell 和 Ruby 系列 2026 年合计出货 2000 万颗，目 标收入 500 亿美元，并可能上调。新增沙特采购协议约 100 亿美元， Anthropic 新增合作需求约 300-350 亿美元，使得原定 5,000 亿美元 目标有 10%的上调空间。 英伟达 CEO 黄仁勋回应 AI 泡沫及 GPU 折旧周期问题，认为云厂商给予 GPU 卡 5-6 年折旧周期合理，且随着推理比例增加，周期可延长。客户 测试替代方案后仍选择英伟达，表明其竞争力强劲。 Q&A 英伟达最新的财报表现如何？其未来展望如何？ 英伟达最新发布的 Q3 财报显示 ...

以下文章来源于傅里叶的猫 ，作者猫叔 傅里叶的猫 . 曾任芯片EDA大厂资深工程师，聊技术、聊产业、聊投资 H200放开的消息昨天已经传的沸沸扬扬了，国内的新闻基本都是这样写的： 但这个新闻最早是出自彭博，比路透要早2个多小时。 而彭博的新闻是下面这个写的，也就是说根据彭博的这个描述，目前只是初步讨论，而且完全有可能只 是停留在讨论，永远不会放开。 这事还得回溯到前段时间中美领导层见面，川普说会谈到Blackwell，大家都以为B30A会放开。后来的 事大家也都知道了，川普说没有谈Blackwell。 但又过了两天，WSJ上的消息说是因为川普的高级顾问们都反对，所以才没有谈，我们当时在星球中就 发过这个： 两国领导开会那天上午，有朋友就发我这样的截图： 所以可能高端的Hopper要放开的事也讨论了很久了。 说话正题，这次的说法是H200要放开，先看下H200的性能： | Specification | H100 | H200 | | --- | --- | --- | | GPU Architecture | Hopper | Hopper | | GPU Memory | 80 GB HBM3 | 141 ...

Core Insights - The article presents a bearish narrative on Nvidia, suggesting it exhibits signs of a Ponzi scheme due to alarming financial metrics and operational anomalies [3][4][5][10]. Receivables Anomaly - Nvidia's accounts receivable surged by 89% to $33.4 billion, with Days Sales Outstanding (DSO) increasing from 46 to 53 days, indicating potential collection issues [3][4]. Inventory Paradox - Inventory rose by 32% to $19.8 billion within three months, contradicting claims of high demand and sold-out capacity, which historically precedes financial collapses [4][5]. Cash Flow Signal - Nvidia's free cash flow conversion rate is only 75%, with a $4.8 billion gap between reported profits and actual cash flow, significantly lower than peers like TSMC and AMD [4][5]. Circular Financing Structure - The article details a complex flow of funds among Nvidia, Microsoft, OpenAI, and others, suggesting a closed-loop system that inflates revenue figures [5][10]. "Vibe Revenue" Admission - AI executives reportedly acknowledge that current AI revenues are largely based on hype rather than actual product sales, with OpenAI's projected 2025 revenue of $3.7 billion against expenses of $9.3 billion highlighting this discrepancy [7][10]. Historical Precedent - The article draws parallels between the current AI bubble and past financial frauds, emphasizing the rapid construction and scale of the current situation [8][10]. Margin Compression Evidence - Despite Nvidia's reported gross margin of over 70%, the article claims that true margins are declining due to various factors, including increased competition and product delays [8][10]. Smart Money Exit - Notable investors, including Peter Thiel and SoftBank's Masayoshi Son, have recently sold significant Nvidia shares, indicating a lack of confidence in the stock [8][10]. Contagion Mechanism - The potential for a broader market impact is discussed, with AI startups using approximately $26.8 billion in Bitcoin as collateral, which could trigger a market downturn if Nvidia's stock falls significantly [9][10]. Regulatory Response - The article anticipates regulatory scrutiny from the SEC and Federal Reserve regarding circular financing and related transactions, potentially leading to significant penalties [9][10]. Conclusion - The author concludes that this situation represents a significant financial risk, predicting a market correction in early 2026, urging investors to divest from AI-related assets [10].

Core Insights - The article presents a bearish narrative on Nvidia, suggesting it exhibits signs of a Ponzi scheme due to alarming financial anomalies and unsustainable business practices [1][9]. Receivables Anomaly - Nvidia's accounts receivable surged by 89% to $33.4 billion, with Days Sales Outstanding (DSO) increasing from 46 to 53 days, indicating potential collection difficulties [2][3]. - The implication is that $10.4 billion may never be collected, a classic sign of financial distress [2]. Inventory Paradox - Inventory rose by 32% to $19.8 billion within three months, contradicting claims of high demand and sold-out capacity [2]. - The price of H100 rental dropped by 34% from $3.20 to $2.12 per hour, challenging the narrative of endless demand [2]. Cash Flow Signal - Nvidia's free cash flow conversion rate is only 75%, with a $4.8 billion gap between profits and actual cash flow, significantly lower than competitors like TSMC and AMD [3]. - This discrepancy suggests a façade of profitability while cash inflow remains weak [3]. Circular Financing Structure - The article details a complex flow of funds among Nvidia, Microsoft, OpenAI, and others, indicating a cycle where revenue is counted multiple times, creating an illusion of growth [3]. - CoreWeave alone owes Nvidia $5.9 billion, representing 18% of total receivables, contributing to a false sense of prosperity [3]. "Vibe Revenue" Admission - Executives from various AI companies reportedly acknowledge that current AI revenues are largely based on hype rather than actual product sales [6]. - OpenAI's projected revenue of $3.7 billion in 2025 against expenditures of $9.3 billion highlights a significant financial imbalance [6]. Historical Precedent - The article draws parallels between the current AI bubble and past financial frauds, such as the 2000 internet bubble and the Enron scandal, suggesting a similar trajectory for Nvidia [6]. Margin Compression Evidence - Despite Nvidia's reported gross margin of over 70%, the true margin is declining when accounting for one-time credits and increased competition [7]. - Future margins are expected to drop below 50%, undermining the company's valuation model [7]. Smart Money Exit - Notable investors, including Peter Thiel and SoftBank's Masayoshi Son, have recently sold significant Nvidia shares, indicating a lack of confidence in the stock [7]. - Major hedge funds have also established short positions following the earnings report [7]. Contagion Mechanism - The article warns that a decline in Nvidia's stock could trigger forced liquidations of Bitcoin collateralized loans, potentially leading to a significant drop in Bitcoin prices [7]. Regulatory Response - Anticipation of regulatory scrutiny from the SEC and Federal Reserve regarding circular financing and related transactions, which could result in mandatory restatements and penalties [8]. Conclusion - The article concludes that the current situation represents a significant financial bubble, with predictions of a market correction occurring between February and April 2026 [8].

这是一篇外网最新发布的分析文章，原文题为《 揭穿6100亿美元骗局的算法：机器智能如何揭露AI行 业的循环融资骗局 》，作者为 Shanaka Anslem Perera。 由于是AI翻译，一些表达可能有问题，但 观点论证似乎挺有说服力，建议认真看看： 正文： 1，应收账款异常 在英伟达10-Q文件第8页的标准财务报表中，隐藏着一个数字——334亿美元的应收账款。算法系统在 文件发布后的几分钟内就标记了这一数据。 2025年11月19日美国东部时间下午4点整，英伟达公司（Nvidia Corporation）发布了超出华尔街预期的 第三季度财报。其营收达到570.1亿美元，高于市场普遍预期的549亿美元；每股收益为1.30美元，也超 过了预期的1.26美元。消息公布后，该公司股价在盘后交易中飙升5%，市值增加了约1300亿美元。 然而，仅仅18个小时后，纳斯达克综合指数收盘下跌1.21%。英伟达的涨幅全部蒸发。此前一度反弹的 比特币也下跌了2.07%，跌至89,567美元。 在这18个小时内发生的事情，在金融市场中前所未有：算法交易系统比人类分析师阅读财报附注的速度 更快地发现了会计异常。 这便是那个发现过程 ...

Summary of Key Points from Conference Call Industry Overview - The conference call primarily discusses the cloud computing industry, focusing on GPU depreciation policies and their financial implications for major cloud service providers such as Amazon, Meta, Microsoft, and Google [1][2][4]. Core Insights and Arguments - **Depreciation Impact on Profit**: Amazon's depreciation adjustments in 2024 added approximately $600-700 million in quarterly profits, but a subsequent adjustment in 2025 is expected to reduce net profits by $100-300 million, highlighting the significant impact of depreciation policies on financial performance [1][2]. - **Future Profit Growth**: A projected change in accounting standards in 2025 is anticipated to result in a net profit increase of about 5.6%, with cumulative net profit additions of $300-500 billion from 2023 to 2028, totaling $1.46 trillion [1][4]. - **GPU Rental Market Dynamics**: Despite a decline in rental prices for older GPUs like the H100 and A100 (down 25% and 30% respectively since September 2024), there remains a strong market for these older models, with some being rented at 95% of their original price [1][3][11]. - **Product Iteration Speed**: NVIDIA has accelerated its product release cycle from every 2-3 years to annually, with new chips like the Blackwell offering significantly improved performance and cost efficiency, which is expected to drive faster market updates [1][7]. - **GPU Lifespan Concerns**: High utilization rates in data centers are shortening GPU lifespans to 1-3 years, raising concerns about equipment wear and tear as demand and supply dynamics shift [1][10]. Additional Important Insights - **Differing Depreciation Policies**: Major cloud providers have adjusted their GPU depreciation periods, with Amazon changing its policy from 6 years in 2024 to 4 years in 2025, while others like Microsoft and Google have extended theirs to 6 years [2]. - **Market Trends in GPU Pricing**: The rental and second-hand market for GPUs is experiencing rapid depreciation, with significant price drops observed since 2019, indicating a need for companies to adapt to changing market conditions [3][11]. - **Emerging Cloud Providers**: New cloud service providers like NeoCloud are facing more volatile pricing and higher operational costs compared to larger firms, which may impact their competitiveness and financial stability [14][15][17]. - **Financial Implications for New Providers**: Companies like Corwave adjusting their depreciation periods are expected to see substantial savings in depreciation costs, significantly affecting their operating and net profits [18]. Conclusion - The conference call highlights the critical role of GPU depreciation policies in shaping the financial landscape of the cloud computing industry, with implications for both established and emerging players. The rapid pace of technological advancement and market dynamics necessitate ongoing adjustments in strategy to maintain competitiveness and profitability [1][19].