摩根士丹利科技对话：乔·摩尔谈 NVDA 季度后表 现,AVGO vs NVDA vs AMD 偏好鉴于 GOOGL 热 情,MU SNDK 最新亚洲调查+EPS 实力 20251125 摘要 英伟达库存增加主要用于扩充机架，客户库存并不充足，DSO 处于历史 平均水平，无需过度担忧。公司正进行战略性投资，以利用资产负债表 发展业务和巩固合作关系，长期来看将受益于人工智能项目，如 OpenAI 的 1,000 兆瓦项目。 尽管新产品周期面临挑战，但对 2026 年 Blackwell 到 Ruben 的产品 交接持乐观态度。亚洲市场信息显示产品上市情况良好，需求旺盛，预 计 Vera Rubin 需求量巨大，市场信心有望恢复。 谷歌 Gemini 3 的良好表现可能提振 TPU 业务，预计 2025 年谷歌与 TPU 相关业务收入将保持中等水平并有所增长。博通处理器业务实现三 位数增长，英伟达、AMD 等公司预计实现同比翻番或接近这个水平。 内存领域需求强劲，来自云计算、AI 及传统服务器 DDR5 内存的需求均 显著增长。供给严重不足导致价格大幅上涨，部分产品价格涨幅达 40%-70%，甚至临时采购价格较最 ...

Core Insights - Nvidia has demonstrated significant sales growth, achieving $57 billion in sales, a 62% increase year-over-year, and a GAAP net income of $31.9 billion, up 21% sequentially and 65% year-over-year [1][3] - The company is recognized as a leading beneficiary of the AI boom, growing from a $360 billion company at the start of 2023 to the largest publicly traded company, briefly reaching a market cap of $5 trillion [3] - Nvidia's CUDA software platform is crucial for maximizing the performance of its GPUs, anchoring customers to its ecosystem [7][10] Sales and Financial Performance - Nvidia's fiscal third-quarter results highlighted its first-mover advantage in the AI market, with strong sales and profit figures [2] - The company has achieved a GAAP gross margin exceeding 70%, driven by high demand for its GPUs and a scarcity of AI-GPUs [6] Market Position and Future Prospects - The AI market is projected to be a $15.7 trillion opportunity by 2030, with Nvidia positioned as a leader in this space [4][5] - Despite current success, there are concerns about the sustainability of Nvidia's growth, particularly regarding the upgrade cycles of its hardware [8][11] Risks and Challenges - Nvidia's CFO indicated that older GPUs, such as the A100, remain relevant due to software improvements, which may reduce the incentive for clients to upgrade their hardware [9][11] - The price of previous-generation GPUs is declining, which could further delay upgrade cycles and impact Nvidia's future growth potential [12][13]

Core Viewpoint - Industrial Fulian's stock price has experienced significant volatility, with a recent drop of 7.80% to 55.94 yuan per share, marking a two-month low and a cumulative decline of over 14% in the last three trading days. The company has denied rumors of a downward adjustment in its fourth-quarter performance targets, asserting that operations are proceeding as planned and customer demand remains strong [3][4]. Financial Performance - For the first three quarters of 2025, Industrial Fulian reported revenue of 603.93 billion yuan, a year-on-year increase of 38.4%, and a net profit attributable to shareholders of 22.49 billion yuan, up 48.52%. In the third quarter alone, revenue reached 243.17 billion yuan, reflecting a 42.81% increase, while net profit surged by 62.04% to 10.37 billion yuan [5]. Market Dynamics - The company's stock price has seen a significant increase of over 170% year-to-date, but recent profit-taking has led to a decline of more than 30% since October 30. The performance of Industrial Fulian's stock may also be influenced by fluctuations in Nvidia's stock price amid discussions surrounding the "AI bubble" [4][5]. Product Development - Industrial Fulian is making progress with its new AI superchip platform, Vera Rubin, with multiple new products being developed in collaboration with partners. These products are expected to enhance computing power density, energy efficiency, and system reliability, becoming a core growth driver for the company's high-end AI server deliveries [4]. Shareholder Actions - In response to inquiries about stock buyback plans, the company indicated that the maximum buyback price was set at 20 yuan per share, which will be adjusted to 19.36 yuan after the 2024 annual equity distribution. The company plans to expedite the buyback process following a board meeting to adjust the buyback price limit [4].

关于近期这些对工业富联堪称"破财"的小作文，我们也注意到，早在11月13日的业绩说明会上，四季度 及2026年订单、业绩问题就已成为焦点。有分析师直接问出"近期有称公司下调了机柜出货与业绩预期 的传闻，请问当前公司应该如何理解"。工业富联方面的回应是针对相关传闻，公司"未发布任何相关口 径，现有客户项目进度与交付节奏正常"。并提示"无须过度解读市场传闻"，称公司将"继续拓展与客户 的多元合作，实现稳健交付，提升利润质量，做到一季比一季好"。 同时，工业富联方面还表示，目前新品推进顺利，在算力密度、电源能效及系统可靠性方面全面升级， 特别在供电架构、散热效率与集成强度方面带来显著优化。后续，相关产品也会成为公司高端AI服务 器交付的核心增长引擎之一。 同时，一些AI社群中也流传有"辟谣贴"，称英伟达为了标准化设计和加速生产切入L10或是行业趋势， 但目前肯定无法像有些传闻一样切入L11（机架级集成）和L12（多机架集群集成）。也有机构报告 称，产业链对英伟达切入L10可能会有较多反对意见，这意味着目前来看，这远非事实。 此外，有业内人士指出，目前工业富联业绩中主要的交付产品是GB200和GB300，可能要到明年 ...

Core Insights - NVIDIA and TSMC are pivotal players in the semiconductor industry, with NVIDIA focusing on AI-driven GPUs and TSMC specializing in advanced chip manufacturing [1][2] NVIDIA Performance - NVIDIA is experiencing significant growth, with Q3 fiscal 2026 revenues increasing by 62% year-over-year to $57 billion and EPS rising by 60% to $1.30 [3][5] - The Data Center segment is the primary growth driver, generating $51.22 billion in revenues, accounting for 89.8% of total sales, reflecting a 66% year-over-year increase [5] - NVIDIA's partnership with OpenAI is expected to enhance long-term demand for its GPUs, solidifying its position as a leading AI chip supplier [6] TSMC Performance - TSMC is a leader in semiconductor foundry services, with revenues in Q3 2025 increasing by 41% year-over-year to $33.1 billion and EPS rising by 39% to $2.92 [8] - The company is investing heavily, with capital expenditures projected between $40 billion and $42 billion in 2025, focusing on advanced manufacturing processes [9] Growth Outlook Comparison - NVIDIA's growth outlook is stronger in the near term, with projected revenue and EPS growth of 57.9% and 50.8% for fiscal 2026, respectively [13] - TSMC's growth projections for 2025 indicate a more modest revenue growth of 33.7% and EPS increase of 44% [14] Stock Performance and Valuation - Both NVIDIA and TSMC have seen strong stock performance in 2025, with NVIDIA shares up 34.5% and TSMC shares up 40.5% [15] - TSMC has a forward P/E ratio of 23.47, while NVIDIA's is higher at 30.38, reflecting NVIDIA's stronger growth profile [16] Investment Recommendation - NVIDIA is positioned as a better investment choice due to its leadership in AI hardware, robust product lineup, and superior growth profile [20]

Jensen, it's been an astonishingly busy day for you in Washington, D. C. , so I'm grateful for your time.You're sold out of Blackwell. That's what you said. But also that $500 billion forecast, which is Blackwell Rubin has room to grow.How do those fit together. I said sales are off the charts for Blackwell. And and video.GPUs in the cloud are sold out. We got plenty of black walls to sell you. We have lots of black walls coming. We're making a lot of black wells. And we have a bunch of Vera Rubin is coming ...

Core Insights - The company is experiencing unprecedented demand for its Blackwell GPUs, with sales described as "off the charts" and a forecast of $500 billion for growth potential [1][2][3] - The upcoming Vera Rubin architecture is on track for delivery in Q3 of next year, with significant engineering efforts underway to ensure a smooth rollout [5][6][8] Supply Chain and Production - The company boasts the largest supply chain globally, with strong partnerships with TSMC and major memory providers like SK Hynix, Micron, and Samsung [3] - The transition from Blackwell to Vera Rubin is expected to be seamless due to well-planned supply chain logistics [7][8] Market Dynamics - The company has zero sales forecast for the Chinese market currently, emphasizing the importance of re-engaging with China for future growth [10][13] - The Chinese market is estimated to be worth around $50 billion, highlighting its significance for both the U.S. and Chinese economies [12] Regulatory Environment - The U.S. Commerce Department has allowed the export of up to 35,000 Blackwell chips to specific Middle Eastern countries, with strict controls to prevent technology transfer to China [14][15] Energy and Infrastructure - The company is growing at a rate of 60% annually, which presents challenges in energy sourcing and infrastructure development [16][17] - A large network of partnerships with cloud service providers enables the company to find diverse energy solutions globally [18] AI Demand and Partnerships - There is exponential growth in compute demand for AI applications, with significant partnerships with companies like Anthropic and OpenAI [25] - The company is focused on disciplined investment and build-out strategies to meet the increasing demand for AI computing power [22][24]

Core Insights - Nvidia reported strong Q3 FY26 earnings, with revenue of $57.006 billion, a 62% year-over-year increase, and net income of $31.910 billion, up 65% year-over-year, surpassing market expectations [1][2][3] Financial Performance - Revenue: $57,006 million, up 22% quarter-over-quarter and 62% year-over-year [2] - Gross Margin: 73.4%, a slight increase of 1.0 percentage points from Q2 FY26, but down 1.2 percentage points from Q3 FY25 [2] - Operating Income: $36,010 million, reflecting a 27% increase year-over-year [2] - Diluted Earnings Per Share: $1.30, a 20% increase year-over-year [2] Business Segments - Data Center Revenue: $51.2 billion, up 66% year-over-year and 25% quarter-over-quarter, accounting for nearly 90% of total revenue [4] - GPU Products: Contributed $43 billion in revenue, with Blackwell products being a key driver of growth [4][5] - Other Segments: Gaming and AI PC business generated $4.3 billion, up 30% year-over-year; Professional Visualization revenue was $760 million, up 56% year-over-year; Automotive and Robotics revenue reached $592 million, up 32% year-over-year [5] Market Position and Outlook - Nvidia's CEO, Jensen Huang, emphasized the ongoing demand for AI and GPU resources, countering claims of an AI bubble [3][7] - The company provided a strong Q4 FY26 revenue guidance of $65 billion, exceeding market expectations of $62 billion [3] - Nvidia's strategic partnerships, including a $100 billion investment in Anthropic, indicate a robust demand for AI infrastructure [6] Investor Sentiment - Following the earnings report, Nvidia's stock rose by 2.85% to $186.52, with a total market capitalization of $4.53 trillion [4] - Despite some hedge funds reducing their positions in Nvidia, the overall market sentiment remains positive due to strong demand and growth prospects [6][7] - Analysts maintain a "buy" rating on Nvidia, with Citigroup raising the target price from $210 to $220, citing supply constraints for AI chips [8]

Product & Supply Chain - Blackwell sales are off the charts, and GPUs in the cloud are sold out, but the company has plenty of Blackwells to sell with a well-planned supply chain [1][2][3] - Vera Rubin silicon chips are back in the labs, with bring-up on track for Q3 of next year, continuing the company's annual cycle [5][6] - The rack scale architecture, including Envy Link 72, is revolutionary and will transition smoothly from Grace Blackwell to Vera Rubin, leveraging existing supply chains [6][7][8] China Market - The company's forecast for the China market is zero, despite it being a potentially $50 billion market [10][12] - The company aims to re-engage with the Chinese market with competitive products, pending government approvals [11][13] Compliance & Diversion - The company is permitted to export up to 35,000 Blackwell chips each to Saudi Arabia and the UAE, with measures in place to prevent technology diversion to China [14] - The company has been rigorous in preventing diversion, testing data centers and implementing technological and process controls [15][16] Energy & Infrastructure - The company's growth rate makes everything a challenge, requiring world-class supply chain management and collaboration with energy providers [16][17] - The company's large network of partners, including cloud service providers and OEMs, helps in finding power solutions globally [18] AI & Market Demand - The company's architecture runs virtually every model, including premier frontier models like OpenAI, Anthropic, and Gemini [19] - The company is disciplined in aligning with partners like OpenAI and Anthropic, considering demand visibility and financing capabilities before building out capacity [21][22] - There is exponential growth in compute demand for AI, with companies struggling to keep up with demand, highlighting the need for increased capacity and optimization [23][24][25]

Core Insights - The company is experiencing unprecedented demand for its Blackwell GPUs, with sales described as "off the charts" and a forecast of $500 billion for growth potential [1][2][3] - The supply chain is robust, with strong partnerships with major manufacturers like TSMC and memory partners such as SK Hynix, Micron, and Samsung, ensuring a well-planned supply chain for upcoming products [3][4][5] Product Development - The Vera Rubin chip development is progressing well, with around 20,000 engineers working on bringing it from silicon to systems, indicating a strong commitment to innovation and product readiness [5][6] - The company is on track to deliver Vera Rubin by Q3 of next year, maintaining an annual product release cycle, which is expected to be a significant success [6][8] Market Dynamics - The Chinese market is currently forecasted to contribute zero revenue due to regulatory challenges, but the company remains committed to re-engaging with this important market in the future [10][12][13] - The U.S. Commerce Department has allowed the export of Blackwell chips to specific markets, indicating a cautious but strategic approach to international sales [14][15] Energy and Infrastructure - The company is growing at a rate of 60% annually, which presents challenges in energy sourcing and infrastructure development, necessitating collaboration with energy providers to ensure readiness for market deployment [16][17] - The extensive network of partnerships across various cloud service providers positions the company advantageously to meet diverse energy needs globally [18] AI and Market Demand - There is exponential growth in compute demand driven by AI applications, with significant partnerships with companies like Anthropic, indicating a strong market presence and future growth potential [25][22][23] - The company is focused on disciplined investment and build-out strategies to support the rapid scaling of AI technologies, ensuring alignment with market demand and financing capabilities [21][22]