Palantir三季度营收11.8亿美元，同比暴增63%，超分析师预期的10.9亿美元。全年营收指引上调至44.0亿美元，较此前 41.4-41.5亿美元大幅提升。这是今年第三次上调年度指引。盘后Palantir一度涨7%，随后快速回落。 硬·AI 作者 | 鲍亦龙 编辑 | 硬 AI Palantir营收创纪录"疯狂"增长，估值也正趋于"疯狂"。 11月3日，Palantir发布第三季度财报，公司营收和利润双双超越预期，并给出了强劲的第四季度指引。 三季度营收11.8亿美元，同比暴增63%。 根据彭博社汇编的数据显示，Palantir已连续21个季度营收超过分析师预期。 具体表现来看： 财务表现： 三季度营收11.8亿美元，同比增长63%，超分析师预期的10.9亿美元。调整后每股收益0.21美元，超预期 0.17美元。 核心业务进展： 盘后Palantir一度涨7%，随后快速回落。分析认为，尽管第三季度表现优异，但市场对于其高估值可持续 性存疑。 公司本季度净利润大增至4.756亿美元，同比增长超两倍。 全年营收指引上调至44.0亿美元，较此前41.4-41.5亿美元大幅提升。自由现金流预期提升至19-21 ...

人工智能巨头正面临一个新瓶颈：电力。微软CEO Satya Nadella透露，公司已采购的芯片因电力和数据中心不足而闲 置。Altman指出，科技公司面临两难：若现在锁定长期电力合同，未来可能因新能源技术的突破而蒙受损失；若投资不 足，又可能无法满足AI需求的爆炸式增长。 硬·AI 作者 | 龙 玥 编辑 | 硬 AI 人工智能竞赛的焦点正从算力转向电力。科技行业领袖们承认，他们正努力应对一个根本性的不确定性： 未来的人工智能究竟需要消耗多少能源。 长期以来，市场普遍认为获取先进的图形处理器（GPU）是部署人工智能服务的最大障碍。然而， Nadella的言论证实，瓶颈已经转移。当科技公司斥巨资采购的芯片无法被点亮时，算力优势便无从谈 起。 这一现象反映了习惯于快速迭代的软件和芯片公司，在面对能源和房地产等重资产、长周期行业时所遭遇 的挑战。在美国，数据中心的电力需求在过去五年中急剧上升，打破了此前长达十年的平稳态势，其增长 速度已超过了公共事业公司的发电容量规划。 这迫使数据中心开发商寻求"电网外"（behind-the-meter）供电方案，即绕过公共电网，直接从发电设施 获取电力。 02 需求迷雾：AI ...

阿里 千问 和 DeepSeek两款中国模型成为唯二盈利的模型。相比之下，四款美国顶尖模型全部亏 损。 作者 | 硬 AI 编辑 | 硬 AI 11月4日，备受关注的AI大模型实时投资比赛"Alpha Arena"落下帷幕。结果极具戏剧性：阿里 千问 Qwen 以 22.32%收益 夺得冠军 ；而 OpenAI的GPT-5和谷歌 的Gemini 2.5 Pro则遭遇惨败，分别亏损 62.66 %和 56.71 %， 大幅腰斩。 在这场为期 17天、全程无人工干预的实盘对决中，阿里 千问 和 DeepSeek两款中国模型成为唯二盈利的模型。相比之下，四款美国顶尖模型全部亏损。 这场由美国人工智能研究实验室 Nof1发起的"AI角斗场"，可能是迄今为止对AI"真实决策水平"最残酷的一次检验。 长久以来，我们习惯于用 MMLU、SWE-Bench之类的学术基准来衡量AI的"智商"。这些榜单固然重要，但它们更像是实验室里的"标准智力测验"，无法回答一个 更根本的问题：当AI离开"考场"，进入混乱、动态且充满对抗的真实世界时，它的"决策水平"究竟如何 ？ 为了回答这个问题，美国人工智能研究实验室 Nof1发起了一场名为 ...

Core Viewpoint - Tesla has shifted the training method for its humanoid robot Optimus from motion capture to pure camera data collection, utilizing video training material from employees performing daily tasks [2][3][6]. Group 1: Training Method Transition - Since June, Tesla has abandoned the previous motion capture suits and remote operation methods, opting for a camera-only data collection approach [6][8]. - Data collection workers wear helmets equipped with five cameras and carry a 30-40 pound equipment pack, repeating basic actions like wiping tables and lifting cups [6][8]. - The transition to camera data collection is expected to accelerate the scaling of data collection efforts [8]. Group 2: AI-Generated Task Instructions - Tesla has begun using AI-generated prompts to assist in training the robot, with workers receiving a series of instructions to complete actions within 3-5 seconds [11]. - Training exercises include a variety of movements, some of which may seem uncomfortable or random, but are believed to help identify areas for improvement [11]. - Data collection also occurs in the Fremont factory, where workers organize vehicle parts while wearing the camera equipment [11]. Group 3: Technical Challenges in Robot Performance - Despite showcasing capabilities in company videos, the actual performance of Optimus during training reveals significant gaps [13]. - Workers report that the robot often falls when performing tasks that require bending or tilting, and it is usually tethered to a support frame to maintain balance [15]. - Experts emphasize that demonstrations often highlight the best performances, lacking true cognitive understanding [15]. Group 4: Workforce and Data Collection - Over 100 individuals have participated in data collection, but the company has laid off several workers following a performance review in September [17]. - Workers are evaluated based on task execution, with a requirement to collect at least 4 hours of usable video material per shift [17].

硬·AI 这不仅是科技巨头间的算力军备竞赛，更对美国的电力基础设施构成了前所未有的挑战。激增的电力需求 正撞上美国现有电网的"电力墙"。电网容量不足、审批延迟和供应限制，正迫使这些科技巨头采取"自备电 力"（Bring-Your-Own-Power）策略。 据巴克莱研究，美国正在规划或建设的数据中心项目总容量超过45吉瓦，预计总投资额将超过2.5万亿美元，OpenAI、 亚马逊、Meta、微软、xAI等超大规模云厂商是主要参与者。巨额投资背后是复杂的融资结构，除了科技巨头自身的资本 支出，私募股权公司和专业基础设施基金扮演了关键角色。 01 巨头领衔： Stargate、超大规模厂商与xAI主导建设 根据巴克莱的追踪，少数几家科技巨头是这45GW项目建设狂潮的核心。 作者 | 龙 玥 编辑 | 硬 AI 一场由人工智能驱动的基建竞赛正在美国全面展开。 据追风交易台消息，巴克莱10月31日的研报显示，美国目前已规划的大型数据中心项目总容量超过45吉 瓦（GW），这股建设热潮预计将吸引超过2.5万亿美元的投资。 报告明确指出，这轮扩张的主要推手是OpenAI的Stargate项目、亚马逊、Meta、微软等超大规模 ...

Core Insights - Morgan Stanley predicts that 2026 will be a pivotal year for explosive growth in AI hardware, driven primarily by strong demand for AI server hardware [2][3] - The report highlights a significant redesign upgrade in AI server hardware, propelled by GPU and ASIC advancements, with notable upcoming releases from NVIDIA and AMD [2][4] AI Server Rack Demand Surge - The demand for AI server racks is expected to surge from approximately 28,000 units in 2025 to at least 60,000 units in 2026, representing over 100% growth [7] - The transition from the H100/H200 era to the new cycle driven by NVIDIA's GB200/300 and Vera Rubin platforms is emphasized [4] Power Consumption and Cooling Solutions - The report identifies power and cooling challenges as significant opportunities for suppliers, with power solutions expected to transition to 800V high-voltage direct current (HVDC) architecture [11][13] - By 2027, the value of power solutions for Rubin Ultra cabinets is projected to exceed ten times the current value of GB200 server cabinets [13] Liquid Cooling as a Standard - Liquid cooling has shifted from an optional solution to a necessity, with the total value of cooling components for a GB300 cabinet estimated at approximately $49,860, increasing to $55,710 for the next-generation Vera Rubin platform [15][16] PCB and Interconnect Upgrades - The upgrade of AI platforms is expected to have a profound impact on printed circuit boards (PCBs) and interconnect components, with increasing requirements for layer counts and material grades [20][21] - The transition from ultra-low-loss to extreme low-loss materials in PCB manufacturing is anticipated to create structural growth opportunities for suppliers with the necessary technical capabilities [22][23]

作者 | 硬 AI 编辑 | 硬 AI 近日，在一场罕见的同场对话中，OpenAI CEO萨姆·奥特曼(Sam Altman)与微软CEO萨提亚·纳德拉(Satya Nadella)深度讨论了AI行业的关键议题。 在这次对话中，Altman详细阐述了OpenAI重组后的愿景。关于与微软的合作协议，Altman澄清只有"无状态API"独家保留在Azure平台到2030年，其他产品包括 ChatGPT将在多个平台分发，这种安排对双方都有利。 关于OpenAI重组方案，Altman澄清只 有"无状态API"独家保留在Azure平台到2030年。面对1.4万亿算力支出质疑，Altman强调收入将远超预期，算力每增10倍收入可增近10倍。关于算力需求，Nadella认为，现在面 临最大问题不是算力过剩，而是电力问题，要不可能会有一堆芯片积压在库存里却无法使用，但算力过剩终将出现。 硬·AI （图片来源：BG2） 核心观点 Altman（OpenAI CEO）："就是因为算力不足的限制，这真的很疯狂，当我看到我们受到多大制约时... 在很多方面，过去一年我们的算力已经扩大了大约10倍。但 如果我们再有10倍的算力，我不知 ...

Core Viewpoint - The article discusses Apple's financial performance in the third quarter, highlighting significant growth in revenue and net profit, driven by strong service revenue and iPhone sales, while also addressing challenges in specific markets like Greater China [14][15]. Financial Data Summary - Revenue: In Q3, net sales reached $102.466 billion, a year-on-year increase of 7.9%, surpassing analyst expectations of $102.137 billion [3]. - EPS: The diluted earnings per share (EPS) for Q3 was $1.85, up 90.7% year-on-year, exceeding analyst expectations of $1.77 [5]. - Net Profit: Q3 net profit was $27.47 billion, reflecting an 86.4% year-on-year increase [6]. - Operating Expenses: Operating expenses in Q3 were $15.91 billion, up 11.4% year-on-year [7]. - Gross Margin: Q3 gross margin was 47.2%, compared to 46.5% in Q2, with gross profit of $48.34 billion, also up 10.2% year-on-year [7]. Segment Performance - Product Sales: Total product sales in Q3 were $73.72 billion, a 5.4% increase year-on-year [8]. - iPhone Sales: iPhone sales reached $49.025 billion, a 6.1% year-on-year increase, slightly below analyst expectations [8]. - Mac Sales: Mac sales were $8.726 billion, up 12.7% year-on-year, exceeding expectations [9]. - iPad Sales: iPad sales were $6.952 billion, showing minimal growth of 0.03% year-on-year [10]. - Wearables, Home, and Accessories: Sales in this category were $9.013 billion, down 0.3% year-on-year [10]. - Services: Service revenue was $28.75 billion, a 15.1% year-on-year increase, surpassing analyst expectations [10]. Market Performance - Americas: Q3 sales in the Americas were $44.19 billion, up 6.1% year-on-year [11]. - Europe: Sales in Europe reached $28.7 billion, a 15.2% increase year-on-year [11]. - Greater China: Sales in Greater China were $14.49 billion, down 3.6% year-on-year, contrary to analyst expectations [11]. - Japan: Sales in Japan were $6.64 billion, up 12% year-on-year [12]. - Other Asia-Pacific: Sales in this region were $8.44 billion, a 14.3% increase year-on-year [13]. Strategic Insights - Service Revenue Growth: Apple's service revenue has consistently reached new quarterly highs, with a 13.5% year-on-year increase to $109.16 billion, indicating strong performance despite hardware profit margin pressures [14]. - Future Guidance: Apple expects Q4 revenue growth of 10%-12%, with potential for significant sales driven by new product launches [15]. - iPhone Sales Trends: iPhone sales growth slowed in Q3 but is expected to rebound in Q4, with positive consumer response to the iPhone 17 series [16][17]. Cost and Investment - Tariff Costs: Apple reported an increase in tariff-related costs to $1.1 billion in Q3, up from $800 million in Q2 [19]. - AI Investment: The company is expanding investments in artificial intelligence, with progress on the new Siri expected to be released next year [20]. Analyst Ratings - Multiple analysts maintain a "Buy" rating on Apple, with target prices ranging from $279 to $320, reflecting confidence in the company's growth prospects [21][22].

亚马逊公布第三季度业绩超出市场预期，得益于AWS增速加快至20%，创三年来最大涨幅，推动股价盘后一度大涨14%。该公司正全力加码AI基础设施建设，预计2025年的资本支 出约为1250亿美元，高于分析师预测的1187.6亿美元，2026年的资本支出还会进一步增加。该公司还表示，已通过110亿美元的"雷尼尔计划"强化与Anthropic的合作，自研AI芯片 供不应求 硬·AI 净销售额： 第三季度净销售额增长13%，达到1,802亿美元，高于2024年第三季度的1,589亿美元以及分析师预期的1778.2亿美元。如果不考虑全年汇率变动带来的15 亿美元利好影响，同比增长为12%。 营业利润： 第三季度营业利润为174亿美元，与2024年第三季度的174亿美元持平。2025年第三季度的营业利润包含两项特别支出——与联邦贸易委员会（FTC）法 律和解相关的25亿美元支出，以及主要与计划裁撤岗位相关的18亿美元预计遣散费用。如果不计入这些支出，营业利润将达到217亿美元。 作者 | 赵雨荷 编辑 | 硬 AI 亚马逊周三盘后公布财报显示，该公司第三季度业绩超出分析师预期，其中云业务创下三年来最大增幅，推动该公司股价盘 ...

高盛认为，谷歌在模型、云、应用全链条卡位，拥有10亿+用户分发能力和成本优势，AI货币化前景广阔。谷歌Q3财报显示：营收首破千亿美元，搜索业务增长15%，云计算收入 增34%、利润率达23%成最大亮点。该行上调谷歌目标价至330美元并维持买入评级。 硬·AI 作者 | 硬 AI 编辑 | 硬 AI 高盛称，凭借从底层模型、云基础设施到终端应用的"全栈AI"布局，这家科技巨头正在将技术优势转化为实实在在的财务回报，维持买入评级，并将目标价上调 至330美元。 10月31日，据硬AI消息，高盛在最新研报中称，谷歌在第三季度财报中全面展示了其"全栈AI"布局的竞争优势。分析师Eric Sheridan指出， 从底层模型开发、云 计算基础设施到消费级和企业级应用，谷歌已经在AI产业链的各个环节实现卡位，这种垂直整合能力正在转化为实际的财务回报。 据 华尔街见闻此前文章 ，Alphabet第三季度综合营收首次突破1000亿美元，搜索业务付费点击和每次点击成本均同比增长7%，YouTube广告收入增长15%超预 期，而谷歌云34%的同比增长率和23%的运营利润率成为本次财报的最大亮点。高盛称， 管理层在财报会议上对AI驱 ...