|
|
楼主 |
发表于 2024-10-8 08:51:48
|
显示全部楼层
Citi on $NVDA - “ We believe AI adoption remains in 3rd/4th innings as enterprise AI demand takes off next with AI agents” Buy PT $150
We have been on the road marketing in Boston/Europe and summarize key investor topics on the stock. While we are bullish on another strong +40% Y/Y cloud data center capex growth next year, we expect the stock to likely remain range bound through CES Jan before Blackwell driven Y/Y sales and gross margin inflection in the Apr-Q. Fundamentally, we believe AI adoption remains in 3rd/4th innings as enterprise AI demand takes off next with AI agents. Maintain Buy and $150 TP. GM Trajectory: Shape of the gross margins in the near-term is a key investor topic. We model low 70’s or ~72% trough in the Jan-Q with LT gross margins to stabilize in the mid-70s% once Blackwell fully ramps. Custom ASIC vs. GPUs: At a high level, we believe the reason to choose one vs. the other has not changed in last few years. ASIC remains fixed functioned as it often runs one application. With AI applications evolving rapidly, model sizes growing, and more applications being AI enabled, GPU market will continue to grow. From a CSP perspective, running cloud AI on ASIC is not viable. From an enterprise standpoint, building applications on ASIC ties them to a specific cloud vendor which often enterprises do not want as they often take the multi-cloud approach. If an enterprise adopting a multi-cloud strategy picks a given cloud and runs its applications on an ASIC, it will have to re-write that application once they move to another cloud. With NVDA GPUs, enterprises write their applications once as it will be transferable across clouds. Moreover, NVDA’s large installed base is a strong pull for developers who aim to have the largest possible adoption of their applications. GPU Competition: While performance metrics are important, we believe the data center operators are focused on TCO and ROI that are both functions of throughput which NVDA leads. As NVDA runs various applications including AI, the data center operators rely on NVDA to have the hardware to run multiple applications rather than buying accelerators that are limited in their use cases. Vertical Integration: The slowdown of Moore’s Law forces scaling focus to systems from chips. NVDA’s products have evolved from chips, to cards to systems, and to finally racks. This approach allows the company to drive innovation as the individual pieces can be further optimized. Blackwell Sales Mix: We expect mix shift more towards GB200 format (rather than B100’s 8-GPU format) given its TCO and ROI benefits. ROI: While NVDA has emphasized ROI benefits that its products deliver for consumer internet companies in large markets such as social media, e-commerce and search, we believe investors need to be patient as Gen AI creates disruptive business models. We expect to see positive ROI data points next year led by GPU as a service providers.
由
翻译自 英语
花旗银行$NVDA - “我们认为，随着企业对人工智能的需求随着人工智能代理的出现而腾飞，人工智能的采用仍处于第三/第四阶段”买入 PT 150 美元
我们一直在波士顿/欧洲进行市场营销，并总结了投资者对该股的关键话题。虽然我们看好明年云数据中心资本支出同比再创 40% 的强劲增长，但我们预计该股可能在 1 月份 CES 期间保持区间波动，然后 Blackwell 会在 4 月份推动同比销售额和毛利率出现拐点。从根本上讲，我们认为人工智能的采用仍处于第三/第四阶段，因为企业对人工智能的需求将随着人工智能代理的出现而腾飞。维持买入和 150 美元的目标价。毛利率轨迹：短期内毛利率的形态是投资者的一个关键话题。我们预测 1 月份毛利率为 70% 出头或 ~72% 的低谷，一旦 Blackwell 完全发展，长期毛利率将稳定在 75% 左右。定制 ASIC 与 GPU：从高层次来看，我们认为选择其中一个而不是另一个的原因在过去几年中没有改变。ASIC 保持固定功能，因为它通常运行一个应用程序。随着人工智能应用的快速发展、模型规模的不断扩大以及越来越多的应用支持人工智能，GPU 市场将继续增长。从 CSP 的角度来看，在 ASIC 上运行云人工智能是不可行的。从企业的角度来看，在 ASIC 上构建应用程序会将它们绑定到特定的云供应商，而企业通常不想要这种情况，因为他们通常采用多云方法。如果采用多云策略的企业选择给定的云并在 ASIC 上运行其应用程序，则一旦迁移到另一个云，就必须重写该应用程序。借助 NVDA GPU，企业只需编写一次应用程序，即可跨云传输。此外，NVDA 庞大的安装基础对旨在最大限度地采用其应用程序的开发人员来说具有强大的吸引力。GPU 竞争：虽然性能指标很重要，但我们认为数据中心运营商专注于 TCO 和 ROI，这两者都是 NVDA 领先的吞吐量的函数。由于 NVDA 运行包括人工智能在内的各种应用程序，数据中心运营商依靠 NVDA 拥有运行多个应用程序的硬件，而不是购买使用案例有限的加速器。垂直整合：摩尔定律的放缓迫使人们将关注点从芯片转向系统。NVDA 的产品从芯片发展到卡、系统，最后发展到机架。这种方法使公司能够推动创新，因为各个部分可以得到进一步优化。Blackwell 销售组合：考虑到其 TCO 和 ROI 优势，我们预计组合将更多地转向 GB200 格式（而不是 B100 的 8-GPU 格式）。投资回报率：虽然 NVDA 强调其产品为社交媒体、电子商务和搜索等大型市场的消费者互联网公司带来的投资回报率优势，但我们认为投资者需要耐心等待，因为 Gen AI 创造了颠覆性的商业模式。我们预计明年将看到由 GPU 即服务提供商引领的正投资回报率数据点。 |
|
IP卡
狗仔卡
发表于 2024-10-7 13:09:11
提升卡
置顶卡
沉默卡
喧嚣卡
变色卡
显身卡
