<p><strong>英伟达</strong>(NVDA 0.54%) 是全球领先的数据中心图形处理单元 (GPU) 供应商，这些 GPU 被用于人工智能 (AI) 的开发。在过去两年中，GPU 销售帮助英伟达增加了 3.2 万亿美元的估值。</p>
<p>该公司刚刚在 11 月 20 日市场收盘后公布了 2025 财年第三季度（截至 10 月 27 日）的业绩，结果远超华尔街的预期。它刚刚开始发货基于其强大 Blackwell 架构的新一代 GPU，需求远远超过供应。</p>
<p>尽管如此，第三季度报告发布后，股价在盘后交易中下跌了 2.5%。我预测未来 12 个月内股价将飙升，因此任何疲软可能都是一个买入机会。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy83OTg1OTIvbnZpZGlhcy1oZWFkcXVhcnRlcnMtd2l0aC1hLWJsYWNrLW52aWRpYS1zaWduLWluLXRoZS1mb3JlZ3JvdW5kLmpwZyZ3PTcwMA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" original-src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy83OTg1OTIvbnZpZGlhcy1oZWFkcXVhcnRlcnMtd2l0aC1hLWJsYWNrLW52aWRpYS1zaWduLWluLXRoZS1mb3JlZ3JvdW5kLmpwZyZ3PTcwMA"/></p>
<p>图片来源：英伟达。</p>
<h2>Blackwell 的需求是 &#34;惊人的&#34;</h2>
<p>过去，数据中心是使用中央处理单元 (CPU) 建造的，这些 CPU 非常适合高效处理少量特定任务。然而，GPU 是为并行处理而设计的，这意味着它们可以同时处理大量任务，具有非常高的吞吐量。</p>
<p>这在训练 AI 模型和执行 AI 推理时至关重要，因为这些工作负载需要能够快速吸收和处理数万亿数据点的芯片。</p>
<p>基于英伟达 Hopper 架构的 GPU——如 H100 和 H200——迄今为止一直是 AI 开发的首选。像 <strong>微软</strong> 和 <strong>亚马逊</strong> 这样的数据中心运营商购买数以万计的这些 GPU，并将其计算能力租给无法负担自己基础设施的企业和 AI 开发者（单个 H100 的售价可高达 40,000 美元）。</p>
<p>现在，随着英伟达的 Blackwell GPU 架构的到来，AI 计算进入了一个新时代。基于 Blackwell 的 GB200 NVL72 系统的 AI 推理速度比相应的 H100 系统快 30 倍。</p>
<p>最近的估计表明，NVL72 系统中的单个 GB200 GPU 的成本约为 83,333 美元，因此开发者以仅是 H100 两倍的价格获得了 30 倍的 AI 推理性能提升。</p>
<p>换句话说，Blackwell GPU 应该会带来令人难以置信的成本效率提升，因此更多企业和开发者能够负担得起部署最先进的 AI 大型语言模型 (LLM)。</p>
<p>在第三季度，英伟达向客户发货了 13,000 个 Blackwell GPU 样品。<strong>微软</strong>、<strong>戴尔</strong> 和 CoreWeave 已经开始建设基于 Blackwell 的数据中心，而 <strong>甲骨文</strong> 的客户将很快能够访问拥有惊人 131,000 个 Blackwell GPU 的计算集群。</p>
<p>英伟达首席执行官黄仁勋表示，Blackwell 的需求是 &#34;惊人的&#34;。GPU 发货量在接下来的几个月可能会激增超过 20 倍，但我稍后会进一步讨论这一点。</p>
<h2>英伟达刚刚超越了华尔街的预期</h2>
<p>在第三季度报告发布之前，华尔街分析师的共识预期是英伟达将实现 332 亿美元的总收入。该公司以 351 亿美元的销售额大幅超出预期，同比增长 94%。</p>
<p>仅数据中心部门就占据了 308 亿美元的总收入，增长率为 112%。大部分收入归功于 GPU 销售。</p>
<p>该公司在对当前 2025 财年第四季度（将于 1 月底结束）的指引中也超出了预期。它告诉投资者，计划实现 375 亿美元的总收入，而华尔街的预期为 371 亿美元。</p>
<p>黄仁勋之前表示，他预计 Blackwell GPU 在第四季度将贡献 &#34;数十亿美元&#34; 的收入，但他现在表示公司有望超出该预期，尽管他没有提供具体的美元数字。</p>
<h2>我为什么认为英伟达股票在未来 12 个月内会飙升</h2>
<p>正如我之前提到的，股票在第三季度报告发布后在盘后交易中小幅下跌。这令人惊讶，因为该公司在各个层面上都超出了预期。但今年股价上涨了 202%，因此一些投资者可能只是决定获利了结。</p>
<p>让我们谈谈估值。根据英伟达过去 12 个月的每股收益为 2.62 美元，其股票的市盈率 (P/E) 为 54.2。这实际上低于其过去 10 年的平均市盈率 58.6。</p>
<p>重要的数字是，华尔街认为英伟达在即将到来的 2026 财年将产生每股收益 4.21 美元，这将在几个月后开始。这使得其前瞻市盈率为 33.8，这意味着其股票在明年必须上涨 73%，才能使其市盈率与过去 10 年的平均水平 58.6 持平。</p>
<h2>激增的 Blackwell 发货量可能支持我的预测</h2>
<p>在我看来，华尔街对 2026 财年收益的预期可能实际上是 <em>过于</em> <em>保守</em>！我之前提到英伟达在第三季度向客户发货了 13,000 个 Blackwell GPU。<strong>摩根士丹利</strong> 表示，该公司有望在 2024 年最后三个月内发货多达 300,000 个单位，随后在 2025 年的前三个月内发货多达 800,000 个单位。</p>
<p>换句话说，Blackwell 的发货量在英伟达当前的 2025 财年第四季度可能会比第三季度激增至少 20 倍。</p>
<p>此外，摩根士丹利认为，<strong>微软</strong>、<strong>亚马逊</strong>、<strong>谷歌母公司</strong> 和 <strong>Meta Platforms</strong> 明年将在 AI 数据中心基础设施上花费总计 3000 亿美元，其中相当一部分将用于 GPU。这仅仅是英伟达的四个主要客户；OpenAI、甲骨文，甚至 <strong>特斯拉</strong> 也是大客户。</p>
<p>因此，我认为股票明年有很大机会实现 73% 或更高的涨幅。</p>

英伟达

<p>英伟达 (NVDA) 预计在未来 12 个月内将实现显著的股票增长，这得益于其新款 Blackwell GPU 架构的强劲需求，该架构的性能优于之前的型号。公司在 2025 年第三季度报告了 94% 的收入增长，超出了华尔街的预期。尽管在报告发布后股价略有下跌，但分析师预测英伟达的收益将继续上升，前瞻市盈率暗示股票价值有潜在 73% 的增长，以与历史平均水平相符。英伟达的 GPU 对人工智能的发展至关重要，预计出货量将显著增加</p>

<p><strong>Nvidia</strong>(NVDA 0.54%) is the world's leading supplier of graphics processing units (GPUs) for data centers, which are used in the development of artificial intelligence (AI). Over the last two years alone, GPU sales have helped Nvidia add $3.2 trillion to its valuation.</p>
<div class="lb-trans"><p><strong>英伟达</strong>(NVDA 0.54%) 是全球领先的数据中心图形处理单元 (GPU) 供应商，这些 GPU 被用于人工智能 (AI) 的开发。在过去两年中，GPU 销售帮助英伟达增加了 3.2 万亿美元的估值。</p>
</div><p>The company just reported results for the fiscal 2025 third quarter (ended Oct. 27) after the market closed on Nov. 20, and they obliterated Wall Street's expectations. It just started shipping a new generation of GPUs based on its powerful Blackwell architecture, and demand is heavily outstripping supply.</p>
<div class="lb-trans"><p>该公司刚刚在 11 月 20 日市场收盘后公布了 2025 财年第三季度（截至 10 月 27 日）的业绩，结果远超华尔街的预期。它刚刚开始发货基于其强大 Blackwell 架构的新一代 GPU，需求远远超过供应。</p>
</div><p>Nevertheless, the stock sank 2.5% in after-hours trading following the third-quarter report. I predict shares are going to soar over the next 12 months, so here's why any weakness might be a buying opportunity.</p>
<div class="lb-trans"><p>尽管如此，第三季度报告发布后，股价在盘后交易中下跌了 2.5%。我预测未来 12 个月内股价将飙升，因此任何疲软可能都是一个买入机会。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy83OTg1OTIvbnZpZGlhcy1oZWFkcXVhcnRlcnMtd2l0aC1hLWJsYWNrLW52aWRpYS1zaWduLWluLXRoZS1mb3JlZ3JvdW5kLmpwZyZ3PTcwMA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F798592%2Fnvidias-headquarters-with-a-black-nvidia-sign-in-the-foreground.jpg&amp;op=resize&amp;w=700"/></p>
<p>Image source: Nvidia.</p>
<div class="lb-trans"><p>图片来源：英伟达。</p>
</div><h2>Blackwell demand is &#34;staggering&#34;</h2>
<div class="lb-trans"><h2>Blackwell 的需求是 &#34;惊人的&#34;</h2>
</div><p>In the past, data centers were built with central processing units (CPUs), which were great for handling a small number of specific tasks with high efficiency. However, GPUs are designed for parallel processing, meaning they can handle numerous tasks at the same time with a very high throughput.</p>
<div class="lb-trans"><p>过去，数据中心是使用中央处理单元 (CPU) 建造的，这些 CPU 非常适合高效处理少量特定任务。然而，GPU 是为并行处理而设计的，这意味着它们可以同时处理大量任务，具有非常高的吞吐量。</p>
</div><p>That's crucial when it comes to training AI models and performing AI inference, because those workloads require chips that can rapidly absorb and process trillions of data points.</p>
<div class="lb-trans"><p>这在训练 AI 模型和执行 AI 推理时至关重要，因为这些工作负载需要能够快速吸收和处理数万亿数据点的芯片。</p>
</div><p>GPUs built on Nvidia's Hopper architecture -- like the H100 and H200 -- have been the go-to choice for AI development so far. Data center operators like <strong>Microsoft</strong> and <strong>Amazon</strong> buy tens of thousands of those GPUs and rent their computing power to businesses and AI developers, which can't afford to build their own infrastructure (a single H100 can sell for up to $40,000).</p>
<div class="lb-trans"><p>基于英伟达 Hopper 架构的 GPU——如 H100 和 H200——迄今为止一直是 AI 开发的首选。像 <strong>微软</strong> 和 <strong>亚马逊</strong> 这样的数据中心运营商购买数以万计的这些 GPU，并将其计算能力租给无法负担自己基础设施的企业和 AI 开发者（单个 H100 的售价可高达 40,000 美元）。</p>
</div><p>Now, a new age of AI computing has arrived with Nvidia's Blackwell GPU architecture. The Blackwell-based GB200 NVL72 system can perform AI inference 30 times faster than the equivalent H100 system.</p>
<div class="lb-trans"><p>现在，随着英伟达的 Blackwell GPU 架构的到来，AI 计算进入了一个新时代。基于 Blackwell 的 GB200 NVL72 系统的 AI 推理速度比相应的 H100 系统快 30 倍。</p>
</div><p>A recent estimate suggests an individual GB200 GPU within an NVL72 system costs around $83,333, so developers are getting that 30-fold increase in AI inference performance for a mere twofold increase in price compared to the H100.</p>
<div class="lb-trans"><p>最近的估计表明，NVL72 系统中的单个 GB200 GPU 的成本约为 83,333 美元，因此开发者以仅是 H100 两倍的价格获得了 30 倍的 AI 推理性能提升。</p>
</div><p>In other words, the Blackwell GPUs should drive an incredible increase in cost efficiency, so more businesses and developers can afford to deploy the most advanced AI large language models (LLMs).</p>
<div class="lb-trans"><p>换句话说，Blackwell GPU 应该会带来令人难以置信的成本效率提升，因此更多企业和开发者能够负担得起部署最先进的 AI 大型语言模型 (LLM)。</p>
</div><p>Nvidia shipped 13,000 Blackwell GPU samples to customers during the third quarter. Microsoft, <strong>Dell</strong>, and CoreWeave have already started building Blackwell-based data centers, and <strong>Oracle</strong> customers will soon be able to access computing clusters with a staggering 131,000 Blackwell GPUs.</p>
<div class="lb-trans"><p>在第三季度，英伟达向客户发货了 13,000 个 Blackwell GPU 样品。<strong>微软</strong>、<strong>戴尔</strong> 和 CoreWeave 已经开始建设基于 Blackwell 的数据中心，而 <strong>甲骨文</strong> 的客户将很快能够访问拥有惊人 131,000 个 Blackwell GPU 的计算集群。</p>
</div><p>Nvidia CEO Jensen Huang says Blackwell demand is &#34;staggering.&#34; GPU shipments could soar more than 20-fold in the next few months alone, but I'll discuss that further in a moment.</p>
<div class="lb-trans"><p>英伟达首席执行官黄仁勋表示，Blackwell 的需求是 &#34;惊人的&#34;。GPU 发货量在接下来的几个月可能会激增超过 20 倍，但我稍后会进一步讨论这一点。</p>
</div><h2>Nvidia just obliterated Wall Street's expectations</h2>
<div class="lb-trans"><h2>英伟达刚刚超越了华尔街的预期</h2>
</div><p>Coming into the third-quarter report, the consensus estimate among Wall Street analysts suggested Nvidia would deliver $33.2 billion in total revenue. The company blew that out of the water with $35.1 billion in sales, which was a 94% increase from the year-ago period.</p>
<div class="lb-trans"><p>在第三季度报告发布之前，华尔街分析师的共识预期是英伟达将实现 332 亿美元的总收入。该公司以 351 亿美元的销售额大幅超出预期，同比增长 94%。</p>
</div><p>The data center segment alone accounted for $30.8 billion of that total, which represented 112% growth. The majority of that money was attributable to GPU sales.</p>
<div class="lb-trans"><p>仅数据中心部门就占据了 308 亿美元的总收入，增长率为 112%。大部分收入归功于 GPU 销售。</p>
</div><p>The company also exceeded expectations with its guidance for the current fiscal 2025 fourth quarter (which will finish at the end of January). It told investors it plans to deliver $37.5 billion in total revenue, compared to Wall Street's estimate of $37.1 billion.</p>
<div class="lb-trans"><p>该公司在对当前 2025 财年第四季度（将于 1 月底结束）的指引中也超出了预期。它告诉投资者，计划实现 375 亿美元的总收入，而华尔街的预期为 371 亿美元。</p>
</div><p>Huang previously said he expected Blackwell GPUs to contribute &#34;several billion dollars&#34; in revenue during the fourth quarter, but he now says the company is on track to exceed that estimate, although he didn't offer a specific dollar figure.</p>
<div class="lb-trans"><p>黄仁勋之前表示，他预计 Blackwell GPU 在第四季度将贡献 &#34;数十亿美元&#34; 的收入，但他现在表示公司有望超出该预期，尽管他没有提供具体的美元数字。</p>
</div><h2>Why I think Nvidia stock will soar over the next 12 months</h2>
<div class="lb-trans"><h2>我为什么认为英伟达股票在未来 12 个月内会飙升</h2>
</div><p>As I mentioned earlier, the stock suffered a modest dip in after-hours trading following the release of its third-quarter report. That's surprising because the company beat expectations on every level. But the stock is up 202% this year, so it's possible some investors simply decided to take profits.</p>
<div class="lb-trans"><p>正如我之前提到的，股票在第三季度报告发布后在盘后交易中小幅下跌。这令人惊讶，因为该公司在各个层面上都超出了预期。但今年股价上涨了 202%，因此一些投资者可能只是决定获利了结。</p>
</div><p>Let's talk about valuation. Based on Nvidia's trailing-12-month earnings per share of $2.62, its stock trades at a price-to-earnings ratio (P/E) of 54.2. That's actually a discount to its average P/E of 58.6 over the last 10 years.</p>
<div class="lb-trans"><p>让我们谈谈估值。根据英伟达过去 12 个月的每股收益为 2.62 美元，其股票的市盈率 (P/E) 为 54.2。这实际上低于其过去 10 年的平均市盈率 58.6。</p>
</div><p>The important number is that Wall Street thinks Nvidia will generate $4.21 in earnings per share in fiscal 2026, which starts in a few months. That places it at a forward P/E of 33.8, which means its stock will have to soar 73% throughout next year just for its P/E to trade in line with its 10-year average of 58.6.</p>
<div class="lb-trans"><p>重要的数字是，华尔街认为英伟达在即将到来的 2026 财年将产生每股收益 4.21 美元，这将在几个月后开始。这使得其前瞻市盈率为 33.8，这意味着其股票在明年必须上涨 73%，才能使其市盈率与过去 10 年的平均水平 58.6 持平。</p>
</div><h2>Surging Blackwell shipments could support my prediction</h2>
<div class="lb-trans"><h2>激增的 Blackwell 发货量可能支持我的预测</h2>
</div><p>In my opinion, Wall Street's fiscal 2026 earnings estimate might actually be <em>too</em> <em>conservative</em>! I mentioned earlier that Nvidia shipped 13,000 Blackwell GPUs to customers during the third quarter. <strong>Morgan Stanley</strong> says the company is on track to ship up to 300,000 units in the final three months of calendar year 2024, followed by up to 800,000 units in the first three months of 2025.</p>
<div class="lb-trans"><p>在我看来，华尔街对 2026 财年收益的预期可能实际上是 <em>过于</em> <em>保守</em>！我之前提到英伟达在第三季度向客户发货了 13,000 个 Blackwell GPU。<strong>摩根士丹利</strong> 表示，该公司有望在 2024 年最后三个月内发货多达 300,000 个单位，随后在 2025 年的前三个月内发货多达 800,000 个单位。</p>
</div><p>In other words, Blackwell shipments could surge by at least 20-fold in Nvidia's current fiscal 2025 fourth quarter compared to the third.</p>
<div class="lb-trans"><p>换句话说，Blackwell 的发货量在英伟达当前的 2025 财年第四季度可能会比第三季度激增至少 20 倍。</p>
</div><p>Plus, Morgan Stanley thinks Microsoft, Amazon, <strong>Alphabet</strong>, and <strong>Meta Platforms</strong> will spend a combined $300 billion on AI data center infrastructure next year, with a significant portion going toward GPUs. Those are only four of Nvidia's top customers; OpenAI, Oracle, or even <strong>Tesla</strong> are also big spenders.</p>
<div class="lb-trans"><p>此外，摩根士丹利认为，<strong>微软</strong>、<strong>亚马逊</strong>、<strong>谷歌母公司</strong> 和 <strong>Meta Platforms</strong> 明年将在 AI 数据中心基础设施上花费总计 3000 亿美元，其中相当一部分将用于 GPU。这仅仅是英伟达的四个主要客户；OpenAI、甲骨文，甚至 <strong>特斯拉</strong> 也是大客户。</p>
</div><p>Therefore, I think there is a very good chance that the stock delivers a gain of 73% or more next year.</p>
<div class="lb-trans"><p>因此，我认为股票明年有很大机会实现 73% 或更高的涨幅。</p>
</div>

Prediction: Nvidia Stock Is Going to Soar Over the Next 12 Months

- Nvidia 报告第三季度业绩，销售额达 5.1 亿美元，超出预期 94%。
- Blackwell GPU 需求强劲，预计未来几个月出货量将激增 20 倍。
- 分析师预测，Nvidia 股票在未来 12 个月内可能上涨 73%。

预测：英伟达的股票将在未来 12 个月内大幅上涨