当卫星在近地轨道完成数据运算并将结果实时传回地球，太空数据中心不再是科幻场景。德意志银行2026年最新研究报 告揭示，这场跨越天地的算力竞赛中，成本追平大约需要十五年以上的时间。 据德银测算，2026年太空数据中心部署成本是地面中心的6.7倍，到2032年将缩小至2.6倍，2030年后有望实现成本追 平。换句话说，大约十五年后，太空数据中心的部署成本才会体现出明显优势。 01 成本对决：四大核心维度的天地差异 (一)发射成本：太空数据中心的"头号门槛" 火箭发射是太空数据中心最核心的成本支出，也是当前最大的经济性障碍。当前可重复使用的猎鹰9号低地轨道发射成 本约为1600美元/千克，而星舰的成熟化将彻底改写这一格局：初期发射成本降至700美元/千克，最终通过完全可重复 使用和规模化运营，有望跌破70美元/千克，降幅超95%。 对比之下，地面数据中心无需承担"上天"成本，但需投入土地征用、基建审批等前期支出，仅发电系统配套建设就可能 面临供应链瓶颈，拉长投资回报周期。 (二)硬件制造成本：卫星迭代vs地面基建刚性 太空数据中心的硬件成本集中在卫星平台，星链的卫星迭代清晰展现了成本下降路径：V3卫星单位功率成本为 ...

商业航天无疑是跨年行情的强势主线，Wind数据显示，自2025年11月21日以来，截至2026年1月7日， 万得商业航天主题指数累计涨幅超60%。 对于市场中是否存在"沾边即炒"的现象，任宏道表示，通过对相关标的梳理可以看出，多数企业均在股 权或业务层面与商业航天存在实质关联，并非纯粹的概念炒作，这反映出资金对行业发展前景的理性考 量。 本轮行情的驱动逻辑呈现多维度特征，其中技术突破是核心因素。任宏道表示，自2015年SpaceX实现 火箭一子级回收以来，国内航天"国家队"与民营力量双线发力，历经十年追赶，将在2025-2026年正式 进入大运力可回收火箭集中"突破期"，这一实质性技术进展为板块行情提供了坚实支撑。政策红利则构 成重要助推力，国家航天局商业航天司的设立、专项发展基金的落地、科创板上市标准的调整与适配等 一系列政策举措，既为行业技术突破保驾护航，又引导社会资源向这一战略性新兴产业倾斜，形成产业 发展与资本市场的良性互动。 对于市场关注的板块估值与企业实际营收偏离问题，任宏道认为，这是新兴战略领域发展初期的共同特 征。从国际对比来看，SpaceX估值已达4000亿美元-1.5万亿美元，AST Sp ...

在开始阅读文章之前，辛苦大家点一下关注，方便后续讨论和分享。作者会不负大家期望，按时、按量 创作出更高质量的内容。本文作者文、编辑均为韩错。 哈喽，大家好，小韩这篇国际评论，主要分析 中美博弈下，美国针对中国所做的战略调整，以及这些调整释放了哪些重要信息。近期，美国发布了新 版国家安全战略文件，引发广泛讨论。文件中提到的不再当世界警察战略收缩等表述，让很多人误以为 中美博弈将会降温。 首先，战争消耗牌未能奏效。美国想复制科索沃、伊拉克战争的以战养霸模式，在东亚挑起可控冲突， 但遭遇了中国总体战准备——完整的产业链战争潜力和深厚的战略纵深，使美国不敢轻举妄动，最终双 方形成一种你来我拦、你走我送的默契对峙。 其次，盟友体系牌出现裂痕。美国试图拉拢盟友进行脱钩、断链，但忽略了欧洲对中国市场的高度依 赖，以及日韩与中国供应链的深度绑定，就连美国本土巨头也离不开中国市场。所谓的联盟，在实际利 益面前显得脆弱不堪。 然而，事实恰恰相反。网友戏称这份文件为逃跑宣言，实际上它反映的是美国对华战略的升级。在此前 极限施压未能达到预期效果后，美国放弃了表面上的硬碰硬，转而采取更加隐蔽、持久的竞争方式。这 才是当前中美关系的真实 ...

Group 1 - Nvidia's stock price plummeted 17% in a single day, resulting in a market value loss of nearly $600 billion, marking the largest single-day market loss in U.S. stock market history [1] - The emergence of DeepSeek, a Chinese company, has caused significant concern on Wall Street, as it achieved results comparable to OpenAI's GPT-4 with a training cost of only $5.57 million, compared to OpenAI's $100 million [2][9] - The media has referred to this event as an "Sputnik moment" in the AI field, indicating a significant shift in technological capabilities [3] Group 2 - DeepSeek's success challenges the notion that more computational power is necessary for better AI models, suggesting that smarter algorithms can achieve similar results with less resource investment [9] - The "Six Little Dragons" from Hangzhou, including DeepSeek and other tech companies, have collectively raised over 18 billion yuan, indicating a shift towards technology-driven growth in the region [12][13] - The concept of "embodied intelligence" is gaining traction, with significant investments in humanoid robots and AI that can interact with the physical world, marking a new frontier in AI development [16] Group 3 - The "Guzi economy," referring to the market for anime and gaming merchandise, has grown to a market size of 200 billion yuan, with over 667,000 related enterprises in China [19] - The rise of "digital nomads," who work remotely and travel, is reshaping workplace dynamics and challenging traditional office structures [20] - Gold prices surged by 70% in 2025, reaching over $4,500 per ounce, driven by a decline in the dollar's global reserve status and increased central bank purchases [23] Group 4 - The competition in the food delivery market intensified with JD's entry, leading to significant consumer benefits through aggressive pricing strategies [26] - Nvidia's market capitalization surpassed $5 trillion, making it the first company in history to reach this milestone, reflecting its dominance in the AI hardware market [28][29] - Warren Buffett announced his retirement as CEO of Berkshire Hathaway, raising questions about the future of value investing in the rapidly evolving tech landscape [32] Group 5 - SpaceX achieved a record 167 rocket launches in 2025, demonstrating the industrialization of space travel and significantly reducing launch costs [35] - The developments in AI, technology, and consumer behavior in 2025 indicate a transformative year, with implications for various industries and investment strategies [36]

Core Viewpoint - The article emphasizes that electricity has become a critical constraint on the development of artificial intelligence (AI), with the demand for power surging due to the exponential growth of AI models and data centers [2][14]. Group 1: AI Power Consumption - ChatGPT consumes approximately 2.9 watt-hours per response, which is nearly ten times the energy used by a traditional Google search, leading to a daily consumption of over 500,000 kilowatt-hours [3][9]. - The energy required for training AI models has significantly increased, with GPT-3 consuming 1,287 megawatt-hours for a single training session, enough to power 3,000 Tesla vehicles for 200,000 miles [9][10]. - The projected AI computing power in the U.S. is expected to require about 1,269 terawatt-hours by 2030, accounting for 22% of the total electricity consumption [10][12]. Group 2: Electricity Supply Challenges - The U.S. electricity grid is aging, with 70% of transformers exceeding their 25-year design life, leading to a low load reserve margin of only 20% [10][11]. - The average outage duration for U.S. users reached 662.6 minutes in 2024, a year-on-year increase of 80.74%, with states like Virginia and Texas experiencing even longer outages [11]. - The demand for electricity from AI is characterized by "pulse-like" spikes, which poses significant challenges to grid stability [10][11]. Group 3: Renewable Energy Solutions - Solar power combined with energy storage is viewed as the most viable solution to meet the growing electricity demands of AI, given its economic advantages and shorter construction timelines compared to other energy sources [15][17]. - The cost of solar power generation is the lowest among various energy sources, with prices ranging from $0.038 to $0.078 per kWh, making it an attractive option for data centers [17][18]. - Major tech companies like Google, Microsoft, and Amazon have set ambitious goals for achieving 100% renewable energy for their data centers by 2030, indicating a strong commitment to sustainable energy solutions [20][21].

Core Insights - The rapid development of the artificial intelligence industry has led major U.S. tech companies to explore the potential of relocating data centers to space, driven by the increasing energy consumption and associated costs of terrestrial data centers [1][2] Group 1: Industry Trends - A report from the Pew Research Center indicates that approximately 4,000 data centers are currently operational or under construction in the U.S., consuming about 4.4% of the country's electricity in 2023, projected to rise to 12% by 2028 [1] - Amazon's founder Jeff Bezos announced that space data centers powered by solar energy could emerge within the next 10 to 20 years [2] - Google's "Sun Catcher Project" aims to launch multiple satellites to create a network for collaborative computing, with plans to deploy two prototype satellites by 2027 [2] Group 2: Technological Developments - The startup Nebula has launched a satellite equipped with NVIDIA H100 chips to test the operational effectiveness of space data centers [2] - Space data centers can utilize solar energy directly and do not face cooling challenges present on Earth, making them potentially more efficient [2] Group 3: Cost Considerations - Nebula estimates that the energy consumption of space-based data centers could be only 10% or less of that of ground-based centers, despite launch costs [3] - Current launch costs for SpaceX's Falcon Heavy rocket are approximately $1,400 per kilogram, with expectations to reduce costs to below $200 per kilogram by the mid-2030s [3] - Elon Musk predicts that future SpaceX Starship launch costs could drop to between $10 and $20 per kilogram, although some analysts express skepticism about achieving these projections [3]

在SpaceX确认有意明年进行首次公开募股(IPO)据称寻求1.5万亿美元估值，以及马斯克强烈 支持在太空部署数据中心之后，连华尔街卖方机构也开始深入研究起了这一概念。 在SpaceX确认有意明年进行首次公开募股(IPO)据称寻求1.5万亿美元估值，以及马斯克强烈支持在太空 部署数据中心之后，连华尔街卖方机构也开始深入研究起了这一概念。 在过去一周，摩根士丹利和德意志银行仅相隔数日发布了两份几乎相同的研报。它们的核心结论是：虽 然实现该计划存在明显的技术挑战，但这些挑战更多似乎属于工程限制而非物理限制。此外，谷歌、 OpenAI和贝索斯的蓝色起源似乎都在探索实现途径，这一事实令分析师和科学家们备受鼓舞。 目前谷歌的"太阳捕手计划"正通过与Planet Labs合作，力争在2027年发射原型卫星。另据报道，OpenAI 的奥尔特曼曾考虑收购火箭公司Stoke Space，而谷歌前CEO埃里克·施密特则实际收购了Relativity Space ——部分动因正是他对太空数据中心的兴趣；蓝色起源团队也在该技术领域深耕了多年。 以下，不妨让我们深入剖析这项曾被视为天马行空的科幻构想将如何成为科学现实。为此我们摘录了德 ...

出品｜搜狐科技 作者｜郑松毅 太空，正成为全球 AI 基建竞争的全新赛场。 最近，在马斯克的强力"带货"下，无论是美国还是国内，"太空数据中心"都是科技圈热议的焦点话题。 他判断，太空数据中心是解决全球AI算力瓶颈的革命性方案——"5年内，太空AI成本效益将以压倒性优势彻底碾压地 面AI。" 如今，马斯克的布局已落地生根。他旗下的 SpaceX 不仅将太空算力纳入核心规划，更成功通过火箭将搭载英伟达 H100芯片的卫星送入轨道，迈出了关键实践一步。 不只是他，盯上太空基建这块"蛋糕"的，还有英伟达、亚马逊、谷歌等一众行业巨头。中国也已攥拳发力，发布了分 阶段推进的太空数据中心建设规划。 "太空算力"是什么，究竟为何如此受大佬青睐，引发新一轮行业竞争？ 为何把算力建到太空？ 不久前，马斯克旗下太空探索公司SpaceX传出推动IPO的消息，其估值有望飙升至惊人的8000亿美元。马斯克也随之 成为人类史上首位身家超过6000亿美元的富翁。 值得关注的是，在此次估值叙事中，一个超越星链的宏伟蓝图被清晰勾勒出来——太空算力。 简单来说，"太空算力"就像太空中的一张大网，上面分布着搭载着高性能计算资源（如AI芯片）的卫星 ...

Group 1 - The A-share market experienced a collective pullback on December 16, with the aerospace sector showing significant fluctuations, particularly the Aerospace ETF (159227), which narrowed its decline to 0.25% by 14:39, with a trading volume of 273 million yuan [1] - The Aerospace ETF has recently surpassed 2.1 billion yuan in scale, making it the largest of its index, with key holdings such as Aerospace Development and Aerospace Electronics reaching their daily limit [1] - The commercial aerospace sector has been gaining momentum, highlighted by the successful launch of Blue Arrow Aerospace's "Zhuque-3" rocket, marking China's first successful orbit of a reusable liquid oxygen-methane rocket [1] Group 2 - CITIC Securities reports that OpenAI is seeking to establish, invest in, or acquire a rocket company to compete in the space computing sector, indicating a growing recognition of the commercial value in space computing [2] - The commercial aerospace industry is entering a new era with the mass launch phases of China StarNet and G60 Qianfan Constellation, alongside the operationalization of Hainan's commercial launch site and rockets, leading to trends of high capacity and low cost [2] - The Aerospace ETF (159227) tracks the National Aerospace Index, with a significant 96.6% allocation to the military industry, focusing on aerospace capabilities and covering a full industry chain from fighter jets to satellites [2]

Core Insights - The future AI computing landscape is likely to be characterized by coexistence rather than a single dominant player, indicating a shift towards a more diversified market [1][7][11] Group 1: Nvidia's Market Position - Nvidia has received permission to export its H200 AI chips to China, but with a stipulation that 25% of sales revenue must be paid as tax to the U.S. government [1] - Nvidia's CEO Jensen Huang expressed concerns about the company's performance, stating that even a slight miss in quarterly results could lead to significant market repercussions [1] - Nvidia's revenue from China has plummeted by 63% year-over-year, dropping from 13% to 5% of total revenue, indicating a severe market contraction [4] Group 2: Rise of Domestic Competitors - Chinese GPU companies like Moore Threads and Muxi are rapidly advancing, with Moore Threads achieving a revenue growth rate of 208.44% from 0.46 billion yuan in 2022 to an expected 4.38 billion yuan in 2024 [3] - Muxi's IPO saw a subscription rate of nearly 3000 times, highlighting strong market interest and confidence in domestic alternatives [3] - Domestic GPU firms are making significant technological strides, with Moore Threads investing over 4.3 billion yuan in R&D and achieving competitive performance metrics against Nvidia's offerings [3] Group 3: Market Dynamics and Challenges - Nvidia's stock has seen a significant decline of approximately 16% within a month, losing over 800 billion yuan in market value, despite reporting a 62% year-over-year revenue increase in Q3 [7] - The emergence of Google's TPU as a competitive threat is notable, with the TPU's cost-effectiveness and system-level optimizations posing challenges to Nvidia's market dominance [8][9] - Major tech companies are increasingly developing their own AI chips, further fragmenting the market and reducing reliance on Nvidia [10]