花旗发表报告指，将英伟达的目标价从210元下调至200美元，以反映上周博通表示其XPU按年增长加速 (受新旧客户推动)后，2026年GPU销售额下降约4%；并估计上述交易将对英伟达2026年销售额造成约 120亿美元的GPU销售影响。 该行对英伟达修订后的2025、2026年预估仍高于共识预期2%、5%，因新兴云端及主权AI支出增加。重 要的是，花旗的预估未包括中国市场，若英伟达恢复对中国的GPU出货，可能带来上行空间；新目标价 200美元基于修订后的2026年每股盈利乘以一致的30倍市盈率；维持"买入"评级，看好其在AI长期增长 机会中的优势。 ...

据媒体报道，继8月底英伟达（NVDA）透露Rubin架构芯片计划明年量产后，当地时间9月8日的高盛技术会议上，英伟达高管又谈到Rubin的进展。 英伟达CFO科莱特·克雷斯（Colette Kress）表示，Rubin芯片已经在为进入市场做准备，Rubin架构将会有6个芯片，这些芯片都已经流片。 图片来源：每日经济新闻 科莱特·克雷斯表示，2026财年第二季度英伟达数据中心收入包含不同部分，其中包含了Blackwell架构的GB200、B200和Blackwell Ultra架构的GB300。第三 季度GB200和GB300还在发货。再往后看，在Rubin进入市场之前，英伟达已经看到了与之相关的几千兆瓦的需求。 Clearstead Advisors高管Jim Awad表示，投资者需要准备好面对英伟达在该领域出现更多竞争的局面，但由于AI市场本身的快速扩张，即便失去部分市场份 额，公司依然能保持稳定增长。 毛利率则有所下滑，该季度为72.4%，去年同期为75.2%。对下一季度的业绩展望，英伟达预计第三季度收入将达到540亿美元，毛利率预计在73.3%至73.5% 之间。截至收盘，英伟达涨0.77%，公司最新市 ...

2025.09.09 此前科莱特·克雷斯在8月底英伟达财报发布后的电话会议上谈到，在接下来5年内，数据中心基础设施 相关的资本开支将达到3万亿~4万亿美元。 在高盛技术会议上，她解释称，谈论接下来5年的数据中心基础设施资本支出，是为了帮助整个生态了 解市场的重要性，"我们在谈论的是未来几十年的新计算平台，我们需要从一个已经存在二三十年的标 准计算平台上开始转变。" 科莱特·克雷斯表示，可以看看四个头部云服务厂商的动作，接下来AI工厂也 将成为全球产业中很重要的一部分，主权AI需求也很重要，将这些因素放在一起，英伟达看到了3万亿 美元以上的资本开支。 科莱特·克雷斯在会议上也回答了关于2023年已部署芯片使用寿命和更换周期的问题，她表示英伟达现 在还没有看到非常大的改变，前一代的Hopper架构芯片仍运行得很好。英伟达不少客户的折旧年限为 4~6年，其中很多人将继续在数据中心保留这些芯片，因为它们仍具备高性能。 2026财年第二季度，英伟达GAAP毛利率72.4%，英伟达展望第三季度GAAP毛利率为73.3%。就如何在 产品转换的过程中提升毛利率，科莱特·克雷斯表示，现在Blackwell Ultra版本运行顺 ...

继8月底英伟达透露Rubin架构芯片计划明年量产后，当地时间9月8日的高盛技术会议上，英伟达又谈到 Rubin的进展。 英伟达CFO科莱特·克雷斯（Colette Kress）表示，Rubin芯片已经在为进入市场做准备，Rubin架构将会 有6个芯片，这些芯片都已经流片。 科莱特·克雷斯表示，2026财年第二季度英伟达数据中心收入包含不同部分，其中包含了Blackwell架构 的GB200、B200和Blackwell Ultra架构的GB300。第三季度GB200和GB300还在发货。再往后看，在 Rubin进入市场之前，英伟达已经看到了与之相关的几千兆瓦的需求。 科莱特·克雷斯说，一年前谈论的话题还是Blackwell架构和架构之间的过渡，现在Blackwell已经进入市 场，还包括Blackwell Ultra，这一年英伟达的节奏较为顺利。另一个经常讨论的话题是，在预训练、后 训练和推理阶段是否还需要大量计算？可以看到，在过去一年间计算方面仍有巨大需求，其中很重要的 一部分是由推理型模型产生的。 此前科莱特·克雷斯在8月底英伟达财报发布后的电话会议上谈到，在接下来5年内，数据中心基础设施 相关的资本开支 ...

Summary of NVIDIA Corp (NVDA.O) Conference Call Company Overview - **Company**: NVIDIA Corp - **Industry**: Graphics Processing Units (GPUs) and AI computing - **Headquarters**: Santa Clara, CA - **CEO**: Jensen Huang Key Financial Insights - **Target Price**: Revised down to $200 from $210, reflecting a ~4% decrease in 2026 GPU sales estimates due to increased competition from TPUs [1][4][19] - **Market Capitalization**: Approximately $4,058.586 billion [4] - **Expected Total Return**: 19.8% based on the new target price [4] Sales and Revenue Projections - **2026 GPU Sales Impact**: Estimated ~$12 billion reduction in GPU sales, with Meta contributing ~$2 billion and the remaining ~$10 billion from other clients [1][2] - **2026 Revenue Estimates**: Revised to $209.912 billion, maintaining a 2%/5% increase above consensus due to higher cloud and sovereign AI spending [1][10] - **Growth Rates**: Projected sales revenue growth of 60.9% in 2026, down from previous estimates [10] Market Dynamics - **GPU vs. XPU Market**: GPUs are expected to maintain over 85% market share in AI compute, but the XPU market is projected to grow 53% YoY in 2026, outpacing GPU growth of 34% YoY [2] - **Key Competitors**: Increased competition from Google, Meta, and Amazon in the XPU market [1][2] Earnings Per Share (EPS) Estimates - **2026 EPS**: Expected to be $4.58, with a growth trajectory leading to $6.68 by 2027 and $7.11 by 2028 [6][10] - **Previous EPS Estimates**: Adjustments made to align with revised revenue projections [10] Risks and Challenges - **Competition Risks**: Potential loss of market share in gaming could negatively impact stock performance [20] - **Adoption Rates**: Slower-than-expected adoption of new platforms may lead to lower sales in data centers and gaming [20] - **Market Volatility**: Fluctuations in the automotive and data center markets could add volatility to stock performance [20] Upcoming Events - **Keynote Event**: Jensen Huang's keynote at GTC on October 28 is anticipated as a significant event for the stock [3] Conclusion - **Investment Recommendation**: Maintain a Buy rating on NVIDIA due to long-term growth opportunities in AI, despite short-term challenges from increased competition and revised sales estimates [18]

8 September 2025 | 4:25PM EDT Nvidia Corp. (NVDA): Communacopia + Technology Conference 2025 — Key Takeaways | +1(212)357-2929 | | --- | | jim.schneider@gs.com Goldman Sachs & Co. LLC | | Anmol Makkar | | +1(212)357-1366 | anmol.makkar@gs.com Goldman Sachs & Co. LLC Chris Kress +1(212)902-6696 | chris.kress@gs.com Goldman Sachs & Co. LLC David Balaban +1(212)357-6369 | david.balaban@gs.com Goldman Sachs & Co. LLC Luya You +1(212)902-5297 | luya.you@gs.com Goldman Sachs & Co. LLC Presenters: Colette Kress, E ...

科莱特·克雷斯表示，2026财年第二季度英伟达数据中心收入包含不同部分，其中包含了Blackwell架构的GB200、B200和Blackwell Ultra架构的GB300。第三 季度GB200和GB300还在发货。再往后看，在Rubin进入市场之前，英伟达已经看到了与之相关的几千兆瓦的需求。 科莱特·克雷斯说，一年前谈论的话题还是Blackwell架构和架构之间的过渡，现在Blackwell已经进入市场，还包括Blackwell Ultra，这一年英伟达的节奏较为 顺利。另一个经常讨论的话题是，在预训练、后训练和推理阶段是否还需要大量计算？可以看到，在过去一年间计算方面仍有巨大需求，其中很重要的一部 分是由推理型模型产生的。 "我们谈论的是未来几十年的新计算平台，我们需要从一个已经存在二三十年的标准计算平台上开始转变。" 此前科莱特·克雷斯在8月底英伟达财报发布后的电话会议上谈到，在接下来5年内，数据中心基础设施相关的资本开支将达到3万亿~4万亿美元。 继8月底英伟达透露Rubin架构芯片计划明年量产后，当地时间9月8日的高盛技术会议上，英伟达又谈到Rubin的进展。 在高盛技术会议上，她解释称，谈论接下 ...

Question-and-Answer SessionSo maybe to start off on a few things from what we heard from you a couple of weeks ago when you reported on your latest earnings call, you stated that the data center infrastructure capital requirements could reach $3 trillion to $4 trillion by the end of the decade. I think many investors were doing a sound check at that point. Maybe give us a little bit of context for that statement and sort of impact or share some of the major building blocks to that?Colette KressExecutive VP ...

Core Viewpoint - D-Wave Quantum Inc. has seen a year-to-date increase of 60% in its stock price, but recently experienced a decline of over 10% in the past month, indicating volatility and investor caution as competition intensifies in the quantum computing sector [1][2]. Group 1: Competitive Landscape - Niche competitors like IonQ and Rigetti are actively developing their technologies, posing a significant risk to D-Wave investors [2]. - NVIDIA's investment of nearly $1 billion in Quantinuum, a joint venture with Honeywell, introduces a well-resourced competitor into the market [2][3]. - Quantinuum's recent capital raise of $600 million has elevated its valuation to $10 billion, positioning it as a formidable player in the quantum industry [4]. Group 2: Technological Approaches - D-Wave focuses on annealing technology and is increasingly exploring gate-model quantum technology, which is advantageous for specific problem types [7]. - In contrast, Quantinuum employs trapped-ion hardware, which is stable but presents scalability challenges [7]. Group 3: Financial Position and Forecast - D-Wave's stock price forecast is set at $19.27, indicating a potential upside of 25.19% based on 11 analyst ratings [8][10]. - The company has a record cash reserve exceeding $800 million, providing it with flexibility for potential acquisitions and strategic adaptations [9]. Group 4: Partnerships and Market Position - D-Wave's partnerships with organizations like Davidson Technologies and Incheon Metropolitan City highlight the perceived value of its quantum technology [4]. - The upcoming launch of Quantinuum's Helios system will allow for direct comparisons with D-Wave's Advantage2 system, impacting market dynamics [8].

Summary of NVIDIA 2025 Conference Call Company Overview - **Company**: NVIDIA (NasdaqGS: NVDA) - **Event**: Goldman Sachs BNP Paribas Pure Technology Conference - **Date**: September 08, 2025 Key Industry Insights - **Data Center Infrastructure**: Capital requirements for data center infrastructure could reach **$3 to $4 trillion** by the end of the decade, indicating a significant transformation in computing platforms driven by AI and accelerated computing [5][8][10] - **Compute Demand**: There is a sustained and increasing demand for compute power, particularly for reasoning models, which are essential for AI applications [7][18] - **Annual Product Cadence**: NVIDIA's one-year product cadence is crucial for maintaining innovation and meeting customer needs, allowing for rapid advancements in technology [6][24] Financial Performance - **Data Center Revenue Growth**: Data center revenue grew **12% sequentially** in Q2, with a projected **17% growth** for Q3, driven by strong demand across various product lines [11][12] - **Networking Growth**: Networking revenue saw significant growth, indicating a strong correlation with compute infrastructure development [31][32] - **Gross Margin Target**: NVIDIA aims to return to a gross margin of **mid-70%** by the end of the year, supported by improved operational efficiency and product mix [42][43] Product Developments - **Blackwell Architecture**: The transition to Blackwell architecture has been successful, with multiple systems like GB200 and GB300 Ultra performing well in the market [6][12][13] - **Vera Rubin Chips**: Six chips in the Vera Rubin architecture are in the maturation phase, with significant demand anticipated even before market release [37][38] - **NVLink Technology**: NVLink continues to be a competitive advantage, enabling high-performance data center solutions and integration with various architectures [27][34] Market Dynamics - **China Market**: NVIDIA has received license approvals for key customers in China, with potential revenue opportunities estimated between **$2 billion to $5 billion** pending geopolitical resolutions [15][16] - **Competition**: The competitive landscape includes the role of ASICs in training and inference markets, with NVIDIA focusing on maintaining performance leadership across various workloads [17][19] Strategic Focus - **Capital Allocation**: NVIDIA is focused on strategic investments in its ecosystem, with potential for M&A but prioritizing organic growth and stock repurchase strategies [44] - **Future Priorities**: Key priorities include advancing the next architecture, maintaining agility in operations, and ensuring readiness for future AI developments [45][46] Conclusion - **AI Evolution**: The journey towards AI integration in various industries is ongoing, with NVIDIA positioned as a key enabler of this transformation through its advanced computing solutions and architectures [48][50][51]