在人形机器人领域，马斯克给出明确的产业化时间表：特斯拉Optimus机器人已在工厂执行简单任务， 2026年底将实现复杂操作，2027年向公众销售。他认为通过"机器人生产力×数量"的指数级增长，全球 经济将迎来"前所未有的爆炸式增长"。这一判断意味着，AI与机器人的结合将不再是单一产业升级，而 是重构全球经济增长模型的关键变量。 在自动驾驶落地节奏上，马斯克给出"2026年底美国广泛普及Robotaxi"的明确目标，并计划"下个月在欧 洲申请受监督的全自动驾驶（FSD）批准"。此外，马斯克对AI智能水平的预测更具颠覆性："2026年底 最迟2027年底，AI将超越单个人类智能；2030年—2031年，AI将超越全人类集体智能"。尽管这一预测 存在争议，但已引发全球科技界对"AI奇点"的重新审视，推动产业加快技术布局。 与马斯克的技术构想相呼应，英伟达CEO黄仁勋在同期公开场合提出"AI是国家关键基础设施"的观点， 将AI竞争从"企业层面"提升至"国家战略层面"，其判断基于AI产业的三大结构性变革，且与马斯克的观 点形成跨企业协同。 黄仁勋认为，当前AI发展已形成"计算架构转型、软件范式迁移、应用形态演进"的铁 ...

周城雄(中国科学院科技战略咨询研究院研究员、数智创新与治理研究中心副主任) 近日，众多全球科技产业领袖现身达沃斯世界经济论坛，围绕AI、机器人、太空探索等领域抛出一系列前沿观点，不仅展现了全球顶尖科 技领袖对AI发展的前沿判断，更折射出当前AI产业的核心趋势与竞争格局，同时也为我国AI发展提供了重要参考。 马斯克在论坛中的观点，以"超预期时间线"和"跨领域协同"为核心，对AI发展前景带来巨大冲击。在AI算力突破方向上，马斯克提出"太 空算力中心"的创新构想，直指当前AI发展的核心瓶颈——电力供应。据其判断，AI芯片产能正呈指数级增长，但全球电力供应年增速仅 3%—4%。为解决这一矛盾，马斯克计划依托SpaceX的星舰技术，在太空部署太阳能AI数据中心。这一构想打破了"算力中心依赖地面电 力与土地资源"的传统认知，为AI算力规模化提供了全新路径。 在人形机器人领域，马斯克给出明确的产业化时间表：特斯拉Optimus机器人已在工厂执行简单任务，2026年底将实现复杂操作，2027年 向公众销售。他认为通过"机器人生产力×数量"的指数级增长，全球经济将迎来"前所未有的爆炸式增长"。这一判断意味着，AI与机器人 的结 ...

炒股就看金麒麟分析师研报，权威，专业，及时，全面，助您挖掘潜力主题机会！ 【导读】AI又来利好！ 来源：中国基金报 记者 泰勒 大家好，关注一下AI圈的利好消息，英伟达又出手了。 1月26日晚间，英伟达向CoreWeave追加投资20亿美元，以帮助加快其到2030年新增超过5吉瓦人工智能 计算能力的计划，推动CoreWeave加速建设"AI工厂"，从而在全球范围内推进AI的规模化采用。 此外，英伟达以每股87.20美元的价格购买了CoreWeave的A类普通股。作为合作的一部分，CoreWeave 将成为首批部署英伟达即将推出产品的公司之一，其中包括存储系统以及一款新的中央处理器 （CPU）。英伟达此前已是CoreWeave的投资方，并且曾同意在2032年之前从CoreWeave采购超过60亿 美元的服务。 这笔交易也是英伟达利用其雄厚资源推动更广泛AI产业发展的最新例证。作为全球市值最高的公司之 一，英伟达已承诺向AI公司投入数百亿美元，为使用其产品建设的新基础设施提供资金支持。 消息出来之后，CoreWeave盘前股价直线拉涨9%。 | 收盘价 01/23 16:00:00 (美东) | | 暨 72 2 ...

Core Algorithm - The industry has shifted towards end-to-end solutions, moving away from modular approaches, at least in public discourse [1] - The introduction of world models is prevalent, with some companies using them to generate training data, while others incorporate them into end-to-end models to enhance performance [1][8] - There is a divergence in opinions regarding the necessity of language models (VLA) in autonomous driving, with some companies arguing that language is not essential for driving tasks [1][11] Simulation and Infrastructure - The closed-loop systems have evolved from data-driven to simulation testing and training loops [2] - 3DGS is highlighted as a crucial technology for building simulation environments, as emphasized by Tesla at CVPR 2025 [5] - Infrastructure is critical, with companies like Xiaomi and Li Auto noting its benefits for development efficiency [3][14] Organizational Capability - Organizational ability is vital, as large autonomous driving teams face significant management challenges [4] - Team culture and collaboration are emphasized as essential for overcoming complex technical and management issues [5] Technical Choices Comparison - A comparison of various companies' technical choices reveals differing approaches to core technologies and the role of world models and simulation tools [9] - Companies like Li Auto advocate for a training loop that evolves from imitation to self-learning, while NVIDIA emphasizes interpretability and reasoning in AI [9] Key Non-Core Factors - R&D infrastructure and engineering efficiency are crucial for the success of autonomous driving technologies [14] - Simulation and synthetic data are becoming essential for addressing corner cases that real-world data cannot cover [14] - The scale of computing power and chip adaptation is critical, as autonomous driving is not just a software issue but also a hardware challenge [15] User Experience and Safety - User experience and safety are paramount, with companies like Xiaomi stressing the importance of balancing advanced technology with user concerns [17] - The need for a dual-stack safety mechanism is highlighted, ensuring that even aggressive end-to-end models have a fallback to traditional rule-based systems for safety [19]

敏实集团(00425)现涨超7%，截至发稿，涨6.9%，报33.76港元，成交额1.4亿港元。 消息面上，1月5日，英伟达在CES展会上发布Rubin平台，单机柜功耗将首次突破200KW，使用100%全 液冷方案。华源证券指出，Rubin液冷方案相比GB系列核心增量变化在冷板、CDU、Manifold。该行指 出，Rubin架构的Manifold在设计上有变化，Manifold管径更大，同时集成阀、传感器等，以提升稳定性 和降低漏液风险，整体集成化、智能化程度更高，价值量提升。Manifold关注敏实集团。 公开资料显示，敏实集团液冷产品包括浸没式液冷柜、CDU、液冷板、分水器。东吴证券指出，能耗 要求+机柜功率密度提升驱动服务器液冷快速发展，敏实凭借着电池盒/液冷板技术储备+全球产能布局 +头部服务器厂商长期合作关系等优势快速切入服务器液冷行业，已获得AI服务器厂商订单，2025年年 末开始交付。 ...

消息面上，1月5日，英伟达在CES展会上发布Rubin平台，单机柜功耗将首次突破200KW，使用100%全 液冷方案。华源证券指出，Rubin液冷方案相比GB系列核心增量变化在冷板、CDU、Manifold。该行指 出，Rubin架构的Manifold在设计上有变化，Manifold管径更大，同时集成阀、传感器等，以提升稳定性 和降低漏液风险，整体集成化、智能化程度更高，价值量提升。Manifold关注敏实集团。 公开资料显示，敏实集团液冷产品包括浸没式液冷柜、CDU、液冷板、分水器。东吴证券指出，能耗 要求+机柜功率密度提升驱动服务器液冷快速发展，敏实凭借着电池盒/液冷板技术储备+全球产能布局 +头部服务器厂商长期合作关系等优势快速切入服务器液冷行业，已获得AI服务器厂商订单，2025年年 末开始交付。 智通财经APP获悉，敏实集团(00425)现涨超7%，截至发稿，涨6.9%，报33.76港元，成交额1.4亿港元。 ...

2026 年，AI 数据中心发展进入了一个新的瓶颈阶段。 单纯的算力堆砌似乎已经走到了尽头，随着代理式 AI 的兴起、混合专家模型（MoE）的普及，以及百万 token 级长上下文逐渐成为常态，大型云 厂商开始意识到，竞争的关键不再是谁拥有更多 GPU，而是谁能更高效地使用，谁能把系统整体效率真正拉起来。 但传统数据中心的短板已经暴露无遗：当代理式 AI 需要处理多轮复杂对话时，它必须频繁调用不同的专家模型，同时保持对历史上下文的准确记 忆。在这个过程中，GPU 可能大部分时间处于等待状态——等待数据从内存传输过来，等待网络将其他节点的计算结果送达，等待存储系统调出历 史记忆。 正是在这样的背景下，英伟达在 CES 2026 上正式发布了 Rubin 平台和 BlueField-4。它们并非是一次硬件的常规升级，而是对传统数据中心低效 这一症结开出的新处方，从某种意义上说，英伟达"重做了一次数据中心"，构建起了真正面向 AI 原生时代的硬件底座。 我们先从一切计算工作的核心——Rubin 平台说起。 与以往围绕单一核心器件不断迭代的路径不同，Rubin 最显著的变化在于，它从一开始就放弃了"单点升级"的思路， ...

Core Viewpoint - Jefferies has raised the target price for NVIDIA from $250 to $275, maintaining a "Buy" rating, indicating strong confidence in the company's future growth potential [1] Group 1: Product Insights - Analyst Blayne Curtis expresses optimism about NVIDIA's Blackwell chip and Rubin platform, predicting these products will be significant growth drivers for the company [1] - The expectation is that NVIDIA will continue to dominate the data center acceleration solutions market [1] Group 2: Market Potential - The overall potential market (TAM) for NVIDIA's gaming business is projected to expand at a double-digit compound annual growth rate (CAGR) [1]

Investment Rating - The industry investment rating is "Positive" (maintained) [1] Core Insights - The report highlights the rapid progress and optimistic outlook for the liquid cooling industry, particularly focusing on the core increments of the Rubin liquid cooling solution, which includes microchannel technology, CDU enhancements, and manifold integration [4][5][10]. Summary by Sections 1. NVIDIA Rubin Platform Release - The Rubin platform adopts a 100% liquid cooling solution, expected to be delivered in the second half of 2026. It features a modular design with a power consumption exceeding 200KW, compared to the GB300's 80% liquid cooling [8][10]. 2. Core Change 1: Microchannel Cold Plate Material and Process Upgrades - The adoption of microchannel technology is expected to become the mainstream cooling method, significantly enhancing cooling efficiency by reducing thermal resistance and increasing heat flow density [13][18]. - Key material considerations include the use of oxygen-free copper for its superior thermal conductivity and structural strength [19]. - Connection methods must ensure quality and sealing to prevent flow channel deformation or blockage, with techniques such as laser welding and brazing being utilized [20][23]. - Flow channel processing requires high precision, with various methods like precision milling and laser processing being employed to achieve the necessary micro-level accuracy [24][29]. 3. Core Change 2: CDU Power and Control Requirements Enhancement - The CDU's overall power will increase in line with the cabinet's power, with enhanced control requirements to accommodate higher cooling demands. This is expected to elevate the CDU's overall value [27][28]. 4. Core Change 3: Manifold Integration Design - The design of the manifold has evolved to be more integrated and intelligent, featuring larger pipe diameters and the inclusion of valves and sensors to improve stability and reduce leakage risks [30][31].

Core Viewpoint - Nvidia reported strong Q3 FY2026 earnings, exceeding market expectations with revenue of $57.01 billion, a 3.48% increase over forecasts, and adjusted EPS of $1.30, surpassing analyst estimates by 3.46% [1][2] Financial Performance - Revenue grew by 26% year-over-year, primarily driven by the data center segment, which contributed $51.2 billion, reflecting a 66% increase [2] - Gross profit increased by 60% to $41.8 billion, although gross margin decreased by 1.2 percentage points to 73.4% due to a shift from selling individual chips to complete systems [2] - Operating income rose by 65% to $36 billion, with net income also increasing by 65% to $31.9 billion, translating to basic EPS of $1.31 [3] - Cash and cash equivalents grew by 40% to $60.6 billion, with total assets at $161.1 billion and total liabilities at $42.3 billion, indicating a healthy balance sheet [3] - Operating cash flow increased by 40% to $66.5 billion, while free cash flow rose by 36% to $61.7 billion, showing improved efficiency in converting sales to cash [3] Future Guidance - Management expects Q4 revenue to be around $65 billion, indicating continued strong momentum, with gross margin projected at approximately 74.8% [3] Valuation Metrics - Nvidia's current trading price is around $188 per share, with a 10% increase over the past six months and a 42% return over the past year [7] - The expected P/E ratio (GAAP) is approximately 40, which is about 26% higher than the industry average, while the expected price-to-book ratio is 29, significantly above the sector median [9] - Analysts believe that Nvidia's dominant position in the AI market justifies its premium valuation, as it reportedly holds 90% of the AI market [9][10] Growth Drivers - The reopening of the Chinese market is expected to drive growth, with over 2 million orders for H200 chips, each priced at approximately $27,000, potentially adding a full quarter's profit if successful [11] - The upcoming launch of the Rubin platform in H2 2026 is anticipated to significantly reduce the cost of running AI models, potentially leading to a substantial market expansion [12][13] - Analysts project that if Rubin captures 60% to 70% of the high-performance chip market by 2028, it could generate $150 to $200 billion in gross profit, translating to $120 to $160 billion in net profit [13] Competitive Position - Nvidia's CUDA software platform has become the industry standard, creating high switching costs for customers, which enhances its competitive moat [9][10] - Each new generation of Nvidia's chips shows exponential performance growth, reducing the attractiveness of older models and driving a cycle of upgrades [10] Conclusion - Despite geopolitical risks and concerns about an AI bubble, analysts maintain a positive outlook on Nvidia, viewing it as a leader in a potentially transformative technology for the next decade [16]