2025.08．25 作者 | 第一财经 周艾琳 此次，一直和美国总统特朗普唱反调的鲍威尔也终于"松口"了，演讲之初就开宗明义——要降息了。 日前，某华尔街大行的中国首席经济学家在闭门会上提及，最近全球市场似乎无惧于经济数据、政策不 确定性的影响，俨然有一个流动性驱使的"水牛"征兆。 在美国，市场对人工智能（AI）主题疯狂扎堆炒作；在欧元区，欧洲央行的持续降息创造了低利率环 境，再通胀和国防支出加码的预期导致资金涌入国防和周期股。而中国在基本面企稳和政策预期下，低 利率驱动部分居民储蓄等资金流入股市，保险资金砸下近万亿增持股市。 8月22日，美联储主席鲍威尔在杰克逊霍尔全球央行年会上表示，"基准前景和风险平衡的变化可能需要 我们调整政策立场"。高盛认为，9月会议降息25个基点（BP）是大概率事件。当前，摩根大通、摩根 士丹利、高盛等投行普遍认为，降息将给亚洲市场注入动能，即使已经历大涨。 鲍威尔口风急转 杰克逊霍尔全球央行年会是每年常规的美联储会议之外最重要的政策会议，从2010年以来，多项重要的 政策转向都是通过这个场合对外官宣的。 如2010年和2012年时任美联储主席伯南克暗示的两轮量化宽松，2014年 ...

2025.08. 25 本文字数：2790，阅读时长大约4.5分钟 作者 | 第一财经 周艾琳 日前，某华尔街大行的中国首席经济学家在闭门会上提及，最近全球市场似乎无惧于经济数据、政策 不确定性的影响，俨然有一个流动性驱使的"水牛"征兆。 在美国，市场对人工智能（AI）主题疯狂扎堆炒作；在欧元区，欧洲央行的持续降息创造了低利率 环境，再通胀和国防支出加码的预期导致资金涌入国防和周期股。而中国在基本面企稳和政策预期 下，低利率驱动部分居民储蓄等资金流入股市，保险资金砸下近万亿增持股市。 此次，一直和美国总统特朗普唱反调的鲍威尔也终于"松口"了，演讲之初就开宗明义——要降息了。 鲍威尔表示，"基准前景与风险平衡的变化可能需要我们调整政策立场。在这一年中，在经济政策发 生深刻变革的背景下，美国经济表现出了韧性。就美联储的双重使命目标而言，就业市场仍接近最大 就业水平，通胀虽然依然偏高，但已从疫情后的高点大幅回落。与此同时，风险的平衡似乎正在发生 转变。" 过去一段时间，美国关键的经济数据显示出相互矛盾的信息。最新CPI尤其是服务通胀高于预期，但 非农爆冷显示出劳动力市场正在降温甚至出现疲弱迹象，这令美联储陷入遏制潜 ...

中国股市仍存动能 日前，某华尔街大行的中国首席经济学家在闭门会上提及，最近全球市场似乎无惧于经济数据、政策不 确定性的影响，俨然有一个流动性驱使的"水牛"征兆。 在美国，市场对人工智能（AI）主题疯狂扎堆炒作；在欧元区，欧洲央行的持续降息创造了低利率环 境，再通胀和国防支出加码的预期导致资金涌入国防和周期股。而中国在基本面企稳和政策预期下，低 利率驱动部分居民储蓄等资金流入股市，保险资金砸下近万亿增持股市。 8月22日，美联储主席鲍威尔在杰克逊霍尔全球央行年会上表示，"基准前景和风险平衡的变化可能需要 我们调整政策立场"。高盛认为，9月会议降息25个基点（BP）是大概率事件。当前，摩根大通、摩根 士丹利、高盛等投行普遍认为，降息将给亚洲市场注入动能，即使已经历大涨。 鲍威尔口风急转 杰克逊霍尔全球央行年会是每年常规的美联储会议之外最重要的政策会议，从2010年以来，多项重要的 政策转向都是通过这个场合对外官宣的。 如2010年和2012年时任美联储主席伯南克暗示的两轮量化宽松，2014年时任欧洲央行行长德拉吉为量化 宽松政策做铺垫，2015年和2016年时任美联储主席耶伦提前释放加息信号，以及2020年鲍威尔 ...

Investment Rating - The report maintains a "Buy" rating for the industry, specifically recommending an "Increase" for the stock of Jinpan Technology [6]. Core Insights - The global AI market is expected to grow significantly, with a projected CAGR of 19.20% from 2025 to 2034, reaching approximately $3680.47 billion by 2034. This growth is driven by the increasing demand for AI applications and infrastructure [11][12]. - Major cloud service providers are ramping up capital expenditures, with Amazon, Google, Microsoft, and Meta all reporting substantial increases in their 2025Q2 CAPEX, indicating a strong demand for data center infrastructure [17][18]. - The introduction of SST (Solid State Transformer) technology is positioned as the next-generation solution for AIDC (AI Data Center) power supply, offering significant efficiency improvements and space savings compared to traditional systems [4][25]. Summary by Sections Global AI Resonance and CSP Capital Expenditure - The AI market is experiencing explosive growth, with the U.S. market expected to reach $1460.9 billion in 2024 and $8514.6 billion by 2034. The Asia-Pacific region is anticipated to be the fastest-growing market [11]. - CSPs are increasing their capital expenditures significantly, with Amazon's 2025Q2 CAPEX at $32.1 billion (up 28% QoQ), Google at $22.4 billion (up 30% QoQ), and Microsoft at $24.2 billion (up 13% QoQ) [17][18]. Development of 800V HVDC Architecture - NVIDIA is developing an 800V HVDC architecture to support the growing power demands of AI workloads, which are expected to exceed 1MW. This new architecture aims to improve data center efficiency by 5%, reduce copper cable usage by 45%, and lower maintenance costs by up to 70% [20][22]. - The transition from traditional 54V AC power systems to 800V HVDC is crucial for meeting the high power requirements of modern AI servers [20][21]. Advantages of SST Technology - SST technology offers a high efficiency of approximately 98%, significantly reducing energy losses compared to traditional AC systems. It also requires 50% less space and simplifies installation by integrating multiple functions into a single unit [4][28]. - The modular design of SST allows for easy maintenance and compatibility with various data center power architectures, enhancing reliability and operational efficiency [4][25]. Key Players and Recommendations - Key players in the SST technology space include Delta, Eaton, and domestic suppliers like Jinpan Technology and Xidian Electric. The report suggests monitoring these companies for potential investment opportunities [5][39][41].

Core Viewpoint - A global cloud infrastructure competition driven by AI is rapidly intensifying, with significant capital expenditure increases expected from major cloud service providers [1][2][7]. Group 1: Capital Expenditure Projections - Morgan Stanley has significantly raised its capital expenditure forecasts for the four major U.S. cloud service providers—Amazon, Google, Meta, and Microsoft—projecting a combined capital expenditure of $359 billion in 2025, a 57% year-over-year increase, and $454 billion in 2026, a 26% increase [1][2]. - The total capital expenditure for the top 11 global cloud service providers is expected to reach $445 billion in 2025, surpassing previous estimates of $400 billion [2]. Group 2: Market Dynamics - The capital expenditure as a percentage of revenue for these companies is projected to exceed 20% by 2026, marking a historical high, with 18% expected in 2025 [3]. - There is a growing demand from non-U.S. regions and Tier 2 cloud service providers, which may have even larger AI server reserves than leading players, indicating a significant market expansion [5]. Group 3: Supply Chain Improvements - Supply chain issues are easing, with improvements in the assembly yield of NVIDIA's next-generation GB200 chips, which is crucial for meeting the rising demand for AI servers [6]. - Major projects like "Stargate," a collaboration involving OpenAI, SoftBank, and Oracle, are moving from concept to execution, indicating a shift from order-based to project-based demand [6]. Group 4: Industry Outlook - Morgan Stanley maintains a positive outlook on the cloud semiconductor industry, citing strong global demand, underestimated growth areas, and improving supply chains as foundational elements for sustained industry growth in the coming years [7].

Core Viewpoint - The AI-driven global cloud infrastructure competition is accelerating, with significant capital expenditure increases expected from major cloud service providers, indicating a robust growth trajectory for the AI server market, particularly from non-U.S. regions and Tier 2 cloud providers [1][3][10]. Group 1: Capital Expenditure Projections - Morgan Stanley has significantly raised its capital expenditure forecasts for the four major U.S. cloud service providers—Amazon, Google, Meta, and Microsoft—projecting a combined capital expenditure of $359 billion in 2025, a 57% year-over-year increase, and $454 billion in 2026, a 26% increase [3][5]. - The capital expenditure for these four companies is expected to reach $100 billion in Q4 2025, reflecting a 39% year-over-year increase [5]. - Expanding the view to the top 11 global cloud service providers, total capital expenditure is projected to reach $445 billion in 2025, significantly higher than the previous estimate of $400 billion [5]. Group 2: Emerging Demand from Non-U.S. and Tier 2 Providers - The report highlights that the market may be underestimating the demand for PCIe/HGX servers from non-U.S. countries, with strong recovery in demand for B200 servers and anticipated growth for B300 servers [8]. - Tier 2 cloud service providers are catching up, with their AI server reserves potentially surpassing those of leading cloud providers, and are expected to significantly increase capital expenditure in the second half of 2026 [8]. Group 3: Supply Chain Improvements - Supply chain issues are easing, with improvements in the assembly yield of NVIDIA's next-generation GB200 chips, which is crucial for meeting the rising demand for AI servers [9]. - The GB300 sample testing is set to begin in Q3, with no significant issues reported, indicating a positive outlook for supply chain capabilities [9]. - Large-scale projects like "Stargate," involving OpenAI, SoftBank, and Oracle, are moving beyond planning stages and are engaging with Asian supply chains for server cabinet procurement, indicating a shift from "order-based" to "project-based" demand [9]. Group 4: Overall Industry Outlook - Morgan Stanley maintains a positive outlook on the cloud semiconductor industry, citing strong global demand, underappreciated growth areas, and improving supply chains as solid foundations for sustained industry growth in the coming years [10].

Core Viewpoint - The AI boom continues to reshape the global economic landscape, with Nvidia's market capitalization reaching $4 trillion, marking a historic milestone in the tech industry since the AI wave began in late 2022 with the launch of ChatGPT [1][3]. Company Performance - Nvidia's stock price rose by 2.5% to $163.9 per share at the opening on July 9, 2023, closing at $162.9, with a total market cap of $3.97 trillion, nearing the $4 trillion mark [3]. - Nvidia's revenue for Q1 of the 2026 fiscal year reached $44.06 billion, a 69% year-over-year increase, with net profit of $18.78 billion, up 26% [4]. - Nvidia's stock has seen a cumulative increase of over 1000% since the beginning of 2023, with a 20% rise since April 2025 [3][4]. Market Position - Nvidia remains the leading supplier of AI chips for AI servers, with its data center business generating $39.1 billion in revenue, accounting for 89% of total revenue, and growing 73% year-over-year [6]. - The global data center market was valued at $250 billion in the previous year, growing at an annual rate of 20% to 25%, supporting Nvidia's valuation [5]. Future Outlook - Analysts predict Nvidia's market cap could reach $6 trillion, with a target price increase from $175 to $250 per share, driven by the anticipated "golden wave" of generative AI adoption [4]. - Nvidia is entering a "decade-long AI infrastructure buildout period," with significant growth opportunities in AI and robotics, representing a multi-trillion dollar market [6]. Challenges - Nvidia faces competition from self-developed chips by major tech companies like Amazon, Microsoft, and Meta Platforms, which could reduce reliance on Nvidia's products [8]. - The company is also challenged by the rise of low-cost AI models and restrictions on sales to China, which previously accounted for 25% of its global sales [7][8]. - Concerns about quantum computing advancements pose a potential risk, with some experts suggesting breakthroughs could occur within 5 to 7 years, potentially surpassing Nvidia's current technology [8].

Group 1 - The core viewpoint of the articles highlights Nvidia's significant growth potential following the GTC Paris developer conference, with strong buy ratings from Morgan Stanley and Evercore ISI, targeting stock prices of $170 and $190 respectively [1][2] - Nvidia plans to deploy over 3000 exaflops of computing power through its Blackwell architecture, with an initial collaboration involving 18,000 GB200 chips with French AI startup Mistral [1] - In the UK market, Nvidia is collaborating with Nebius and Nscale to create a computing cluster of 14,000 GB200 chips, alongside partnerships with major telecom operators across Europe [1] Group 2 - Evercore ISI analysts emphasize Nvidia's expansion of its European ecosystem, working with five telecom operators, 19 cloud service providers, and 16 supercomputing centers to build a localized AI computing network [2] - The EU's plan to support 20 AI factories, including several super factories, is expected to generate $40 billion to $50 billion in investments per factory, creating substantial demand for American cloud service providers [2] - Both investment banks stress that Nvidia is solidifying its AI computing dominance through a dual strategy of "hardware + ecosystem," with ongoing projects and EU policy support enhancing its market position [2]

Core Viewpoint - The article emphasizes the advancements in quantum computing and AI infrastructure, highlighting NVIDIA's role in these developments and the importance of AI factories in Europe [1][3][4]. Quantum Computing Developments - NVIDIA's CEO Jensen Huang stated that quantum computing is approaching a turning point, with predictions that the number of logical quantum bits will increase tenfold every five years and a hundredfold every ten years [1]. - NVIDIA has introduced CUDA-Q, an open-source quantum development platform designed for classical quantum computing, allowing parallel computation across GPU, CPU, and QPU resources [2]. AI Factory Concept - AI factories are described as "revenue-generating machines" that produce tokens for various industries, and they are becoming part of national infrastructure [3][4]. - Huang noted that European telecom and cloud companies are collaborating with NVIDIA to build AI infrastructure, with over 20 AI factories planned, including several gigawatt-scale facilities [4]. European AI Infrastructure - NVIDIA aims to increase AI computing capacity in Europe by tenfold within two years, addressing the GPU shortage faced by researchers and startups [4]. - Huang highlighted the potential of the UK in AI development, noting the lack of local sovereign AI infrastructure and indicating NVIDIA's intention to invest in the UK [4].

Core Insights - Huang Renxun, CEO of Nvidia, emphasized the collaboration between Quantum Processing Units (QPU) and Graphics Processing Units (GPU) for next-generation computing [1][3] - Quantum computing is approaching a critical turning point, with significant advancements in logical qubits and error correction [1][3] - Nvidia's CUDA-Q platform allows for hybrid programming, enabling parallel computation across GPU, CPU, and QPU resources [3][4] Quantum Computing Developments - Huang predicts that the number of logical qubits in quantum computers will increase tenfold every five years and a hundredfold every ten years, improving error correction and performance [3] - Nvidia has developed CUDA-Q, an open-source quantum development platform designed for classical quantum computing, which can run on Nvidia's Grace Blackwell chips [3][4] AI Factory Concept - The concept of AI factories, which run AI algorithms and generate tokens, is becoming integral to national infrastructure, with Nvidia actively engaging with global leaders on this topic [4][5] - Nvidia is collaborating with various companies to build AI infrastructure in Europe, with over 20 AI factories planned, including several gigawatt-scale facilities [5] European AI Infrastructure - Nvidia aims to increase AI computing capacity in Europe by tenfold within two years, addressing GPU shortages for researchers and startups [5] - The company is working with 77 countries to establish AI technology centers, focusing on collaboration with startups and ecosystem development [5]