<p>如果回顾过去几年，几乎没有哪只股票像 <strong>Nvidia</strong>(NVDA 0.58%) 那样受到如此多的关注。当前的人工智能（AI）热潮使 Nvidia 成为焦点，并成为全球最有价值的上市公司。截至 8 月 8 日市场收盘，Nvidia 的市值为 4.55 万亿美元，比第二名 <strong>Microsoft</strong> 大出 5700 亿美元。</p>
<p>关于 Nvidia 的业务，尤其是与 AI 相关的方面，有很多值得喜爱的地方，但如果你在寻找一只可能在 2030 年前超越它的 AI 相关股票，我的预测是 <strong>台积电</strong>(TSM 0.77%)。</p>
<p>我并不是说到 2030 年，TSMC 的市值会超过 Nvidia——目前 TSMC 的市值约为 Nvidia 的 3.2 万亿美元_少于_。但如果今天是起点，TSMC 的业务使其股票在未来五年内有可能超越 Nvidia。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MjgzNDYvZ2V0dHlpbWFnZXMtMTgwOTY2MDE0OC5qcGcmdz03MDA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" width="700" height="393" original-src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MjgzNDYvZ2V0dHlpbWFnZXMtMTgwOTY2MDE0OC5qcGcmdz03MDA"/></p>
<p>图片来源：Getty Images。</p>
<h2>为什么 Nvidia 的股票如此受欢迎？</h2>
<p>在过去三年中，Nvidia 的股票上涨了 920%。当你关注真正重要的事情时，Nvidia 的业务归结为图形处理单元（GPUs）。GPUs 是数据中心的关键部分，支持和促进 AI 的训练、部署和扩展。</p>
<p>Nvidia 无疑是 AI 专用 GPUs 的主要供应商，市场份额估计超过 90%。这种虚拟垄断为 Nvidia 带来了丰厚的回报（字面意义上），使其收入从 2023 财年第一季度的 83 亿美元增长到 2026 财年第一季度的 440 亿美元（截至 4 月 27 日）。</p>
<p><img src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/0435c78262f021a3b9f48074db77eb71.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" width="720" height="441" original-src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/0435c78262f021a3b9f48074db77eb71.png"/></p>
<p>NVDA 收入（季度）数据来自 YCharts</p>
<h2>一切始于 TSMC</h2>
<p>尽管 Nvidia 及其 GPUs 对 AI 世界极为重要，但没有 TSMC，Nvidia 无法生产这些产品，TSMC 是 Nvidia 的 GPUs 和先进芯片的独家制造商。</p>
<p>如果说 Nvidia 是 AI 生态系统中的一辆炫酷跑车，那么 TSMC 就是使其性能得以实现的引擎，这就是为什么我愿意称这家半导体制造商为 AI 股票，尽管它并没有以这种方式呈现自己，也没有面向消费者的 AI 产品。</p>
<p>通过逆向思维来看 TSMC 的影响是有帮助的。面向消费者的产品如生成性 AI（ChatGPT、Gemini 等）必须使用大量数据进行训练。这种训练没有数据中心是无法实现的。这些数据中心大多数运行在 Nvidia 的芯片上，而 Nvidia 的芯片是由 TSMC 制造的。</p>
<h2>TSMC 的业务面临的长期风险较少</h2>
<p>尽管对 TSMC 制造的 AI 相关芯片的需求推动了其近期收入的增长，但它并不需要仅依赖这些芯片来实现长期成功。其高性能计算部门，包括其 AI 相关芯片，占其第二季度 300 亿美元收入的 60%。这确实是相当可观的，但 TSMC 的业务更多依赖于广泛的科技领域，而不仅仅是 AI 的发展。</p>
<p>许多公司依赖 TSMC 来制造他们的芯片。从智能手机到笔记本电脑，再到汽车、游戏机和数据中心服务器，所有这些都包含由 TSMC 制造的半导体。其客户名单包括 <strong>Apple</strong>、<strong>Broadcom</strong>、<strong>AMD</strong> 和 <strong>Qualcomm</strong> 等。</p>
<p>TSMC 多样化的收入来源减少了一些潜在的波动。如果 AI 支出和需求放缓，TSMC 还有其他部门可以弥补损失。另一方面，Nvidia 的业务将受到重大打击。上个季度，数据中心占其收入的 88% 以上。</p>
<h2>估值很重要</h2>
<p>在撰写本文时，Nvidia 的市盈率为 42.4 倍，而 TSMC 的市盈率为 24.7 倍。尽管 TSMC 在许多标准下并不被认为是 “便宜”，但它远没有 Nvidia 那么昂贵，估值也更加合理。</p>
<p><img src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/a101aeb3c6f574e15f6ef802256416c0.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" width="720" height="456" original-src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/a101aeb3c6f574e15f6ef802256416c0.png"/></p>
<p>NVDA 市盈率（前瞻性）数据来自 YCharts</p>
<p>TSMC 较低的估值在未来五年内留出了更多的增长空间。在 Nvidia 的估值下，似乎其未来的增长已经在股票中被充分定价，这使其在未能达到高期望时容易出现剧烈回调。</p>
<p>这并不是说 TSMC 的股票不能经历回调，但其估值给了它更大的误差范围和更多的上行空间。当我们在五年后回顾时，我一点也不会惊讶看到 TSMC 的股票在这一点上超越了 Nvidia。</p>

台积电

半导体 3 倍做多 - Direxion

<p>文章预测，到 2030 年，台积电将在人工智能领域超越英伟达。尽管英伟达在 GPU 市场占据了 90% 的份额，但台积电作为这些芯片的独家制造商至关重要。台积电的收入来源多样化，且其估值（市盈率为 24.7 倍）低于英伟达（市盈率为 42.4 倍），这使其在增长前景和抵御市场波动方面具有更好的优势。作者认为，尽管英伟达目前在市场上处于领先地位，台积电的股票在未来五年将表现得更为强劲</p>

<p>If you look back at the past couple of years, there's arguably no stock that has received as much attention as <strong>Nvidia</strong>(NVDA 0.58%). The current artificial intelligence (AI) boom has propelled Nvidia into the spotlight and to being the world's most valuable public company. As of market close on Aug. 8, Nvidia's market cap is $4.55 trillion, more than $570 billion larger than second-place <strong>Microsoft</strong>.</p>
<div class="lb-trans"><p>如果回顾过去几年，几乎没有哪只股票像 <strong>Nvidia</strong>(NVDA 0.58%) 那样受到如此多的关注。当前的人工智能（AI）热潮使 Nvidia 成为焦点，并成为全球最有价值的上市公司。截至 8 月 8 日市场收盘，Nvidia 的市值为 4.55 万亿美元，比第二名 <strong>Microsoft</strong> 大出 5700 亿美元。</p>
</div><p>There are plenty of things to love about Nvidia's business, especially regarding AI, but if you're looking for an AI-related stock that could outperform it between now and 2030, my prediction would be <strong>Taiwan Semiconductor Manufacturing Company</strong>(TSM 0.77%) (TSMC).</p>
<div class="lb-trans"><p>关于 Nvidia 的业务，尤其是与 AI 相关的方面，有很多值得喜爱的地方，但如果你在寻找一只可能在 2030 年前超越它的 AI 相关股票，我的预测是 <strong>台积电</strong>(TSM 0.77%)。</p>
</div><p>I'm not saying that in 2030, TSMC will be worth more than Nvidia -- TSMC's market cap is currently around $3.2 trillion <em>less</em> than Nvidia's -- but if today is the starting point, TSMC has a business that puts its stock in a good position to outperform Nvidia's in the next five years.</p>
<div class="lb-trans"><p>我并不是说到 2030 年，TSMC 的市值会超过 Nvidia——目前 TSMC 的市值约为 Nvidia 的 3.2 万亿美元_少于_。但如果今天是起点，TSMC 的业务使其股票在未来五年内有可能超越 Nvidia。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MjgzNDYvZ2V0dHlpbWFnZXMtMTgwOTY2MDE0OC5qcGcmdz03MDA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F828346%2Fgettyimages-1809660148.jpg&amp;op=resize&amp;w=700"/></p>
<p>Image source: Getty Images.</p>
<div class="lb-trans"><p>图片来源：Getty Images。</p>
</div><h2>Why has Nvidia's stock been such a hit?</h2>
<div class="lb-trans"><h2>为什么 Nvidia 的股票如此受欢迎？</h2>
</div><p>Over the past three years, Nvidia's stock is up 920%. When you look at what really matters, Nvidia's business comes down to graphic processing units (GPUs). GPUs are a key part of data centers, which power and enable AI training, deployment, and scaling.</p>
<div class="lb-trans"><p>在过去三年中，Nvidia 的股票上涨了 920%。当你关注真正重要的事情时，Nvidia 的业务归结为图形处理单元（GPUs）。GPUs 是数据中心的关键部分，支持和促进 AI 的训练、部署和扩展。</p>
</div><p>Nvidia is far and away the main supplier of AI-focused GPUs, with an estimated market share of over 90%. This virtual monopoly has paid off for Nvidia (quite literally), increasing its revenue from $8.3 billion in its 2023 fiscal first quarter to $44 billion in its 2026 fiscal first quarter (ended April 27).</p>
<div class="lb-trans"><p>Nvidia 无疑是 AI 专用 GPUs 的主要供应商，市场份额估计超过 90%。这种虚拟垄断为 Nvidia 带来了丰厚的回报（字面意义上），使其收入从 2023 财年第一季度的 83 亿美元增长到 2026 财年第一季度的 440 亿美元（截至 4 月 27 日）。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/0435c78262f021a3b9f48074db77eb71.png" alt="" original-src="https://media.ycharts.com/charts/0435c78262f021a3b9f48074db77eb71.png"/></p>
<p>NVDA Revenue (Quarterly) data by YCharts</p>
<div class="lb-trans"><p>NVDA 收入（季度）数据来自 YCharts</p>
</div><h2>It all starts with TSMC</h2>
<div class="lb-trans"><h2>一切始于 TSMC</h2>
</div><p>Despite how valuable Nvidia and its GPUs are to the AI world, Nvidia wouldn't be able to produce them without TSMC, the exclusive manufacturer of Nvidia's GPUs and advanced chips.</p>
<div class="lb-trans"><p>尽管 Nvidia 及其 GPUs 对 AI 世界极为重要，但没有 TSMC，Nvidia 无法生产这些产品，TSMC 是 Nvidia 的 GPUs 和先进芯片的独家制造商。</p>
</div><p>If Nvidia is a flashy sports car in the AI ecosystem, TSMC is the engine that makes its performance possible, which is why I'm comfortable calling the semiconductor manufacturer an AI stock despite not presenting itself as such or having a consumer-facing AI product.</p>
<div class="lb-trans"><p>如果说 Nvidia 是 AI 生态系统中的一辆炫酷跑车，那么 TSMC 就是使其性能得以实现的引擎，这就是为什么我愿意称这家半导体制造商为 AI 股票，尽管它并没有以这种方式呈现自己，也没有面向消费者的 AI 产品。</p>
</div><p>It helps to see TSMC's impact by working backward. Consumer-facing products like generative AI (ChatGPT, Gemini, etc.) must be trained using tons of data. This training isn't possible without data centers. Most of these data centers run on Nvidia chips. Nvidia chips are built by TSMC.</p>
<div class="lb-trans"><p>通过逆向思维来看 TSMC 的影响是有帮助的。面向消费者的产品如生成性 AI（ChatGPT、Gemini 等）必须使用大量数据进行训练。这种训练没有数据中心是无法实现的。这些数据中心大多数运行在 Nvidia 的芯片上，而 Nvidia 的芯片是由 TSMC 制造的。</p>
</div><h2>TSMC's business has fewer long-term risks</h2>
<div class="lb-trans"><h2>TSMC 的业务面临的长期风险较少</h2>
</div><p>Although demand for TSMC-manufactured AI-related chips has driven much of its recent revenue growth, it doesn't have to rely solely on them for long-term success. Its high-power computing segment, which includes its AI-related chips, was 60% of its $30 billion in revenue in the second quarter. That's admittedly a decent bit, but TSMC's business relies more on the broad tech world, and not just AI developments.</p>
<div class="lb-trans"><p>尽管对 TSMC 制造的 AI 相关芯片的需求推动了其近期收入的增长，但它并不需要仅依赖这些芯片来实现长期成功。其高性能计算部门，包括其 AI 相关芯片，占其第二季度 300 亿美元收入的 60%。这确实是相当可观的，但 TSMC 的业务更多依赖于广泛的科技领域，而不仅仅是 AI 的发展。</p>
</div><p>Lots of companies rely on TSMC to manufacture their chips. Everything from smartphones to laptops to cars to gaming consoles to data center servers all contain semiconductors manufactured by TSMC. Its client list includes <strong>Apple</strong>, <strong>Broadcom</strong>, <strong>AMD</strong>, and <strong>Qualcomm</strong>, just to name a few.</p>
<div class="lb-trans"><p>许多公司依赖 TSMC 来制造他们的芯片。从智能手机到笔记本电脑，再到汽车、游戏机和数据中心服务器，所有这些都包含由 TSMC 制造的半导体。其客户名单包括 <strong>Apple</strong>、<strong>Broadcom</strong>、<strong>AMD</strong> 和 <strong>Qualcomm</strong> 等。</p>
</div><p>TSMC's diversified revenue streams reduce some potential hiccups. If AI spending and demand slows, TSMC has other segments that can pick up the slack. Nvidia's business, on the other hand, will take a major hit. Data centers accounted for over 88% of its revenue last quarter.</p>
<div class="lb-trans"><p>TSMC 多样化的收入来源减少了一些潜在的波动。如果 AI 支出和需求放缓，TSMC 还有其他部门可以弥补损失。另一方面，Nvidia 的业务将受到重大打击。上个季度，数据中心占其收入的 88% 以上。</p>
</div><h2>Valuation matters</h2>
<div class="lb-trans"><h2>估值很重要</h2>
</div><p>At the time of this writing, Nvidia is trading at 42.4 times earnings estimates, and TSMC is trading at 24.7 times earning estimates. Although TSMC isn't considered &#34;cheap&#34; by many standards, it's nowhere near as expensive as Nvidia and is much more fairly valued.</p>
<div class="lb-trans"><p>在撰写本文时，Nvidia 的市盈率为 42.4 倍，而 TSMC 的市盈率为 24.7 倍。尽管 TSMC 在许多标准下并不被认为是 “便宜”，但它远没有 Nvidia 那么昂贵，估值也更加合理。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://media.ycharts.com/charts/a101aeb3c6f574e15f6ef802256416c0.png" alt="" original-src="https://media.ycharts.com/charts/a101aeb3c6f574e15f6ef802256416c0.png"/></p>
<p>NVDA PE Ratio (Forward) data by YCharts</p>
<div class="lb-trans"><p>NVDA 市盈率（前瞻性）数据来自 YCharts</p>
</div><p>TSMC's lower valuation leaves much more room for growth over the next five years. At Nvidia's valuation, it seems to have a lot of future growth already priced into the stock, which leaves it susceptible to a sharp pullback if it falls short of its lofty expectations.</p>
<div class="lb-trans"><p>TSMC 较低的估值在未来五年内留出了更多的增长空间。在 Nvidia 的估值下，似乎其未来的增长已经在股票中被充分定价，这使其在未能达到高期望时容易出现剧烈回调。</p>
</div><p>That's not to say that TSMC's stock can't experience a pullback, but its valuation gives it a larger margin of error and more upside. When we look back five years from now, I wouldn't be the least bit surprised to see TSMC's stock has outperformed Nvidia's from this point.</p>
<div class="lb-trans"><p>这并不是说 TSMC 的股票不能经历回调，但其估值给了它更大的误差范围和更多的上行空间。当我们在五年后回顾时，我一点也不会惊讶看到 TSMC 的股票在这一点上超越了 Nvidia。</p>
</div>

Prediction: This Artificial Intelligence (AI) Stock Will Outperform Nvidia Through 2030

- Nvidia 因 AI 热潮成为全球最有价值公司，市值达 4.55 万亿美元。  
- TSMC 作为 Nvidia 的独家制造商，未来五年有望超越 Nvidia 表现。  
- TSMC 的多元化收入降低风险，估值更具增长空间。  

预测：这只人工智能（AI）股票将在 2030 年之前的表现超过 Nvidia