<p><strong>英伟达</strong>(NVDA -3.01%) 在过去两年半的人工智能（AI）热潮中成为了明星。它的 GPU 是每个大型数据中心项目的基石，这些项目来自于投资数百亿美元于 AI 的超大规模云客户。这导致了英伟达在过去两年中收入和盈利的巨大增长。</p>
<p>投资者得到了应有的回报。英伟达的股价在 2023 年上涨了 239%，在 2024 年又上涨了 171%。如此强劲的价格上涨使得该公司短暂地登上了最有价值公司的榜单。但现在，在最近的股市抛售之后，英伟达的股价目前距离 1 月份创下的历史高点下跌了约 25%。</p>
<p>虽然许多投资者可能在关注这家 AI 领军企业并考虑在较低价格时购买其股票，但另外两家人工智能公司似乎有望在长期内超越这只半导体股票，得益于它们的竞争优势，使其在 AI 故事如何发展时都能保持强大的竞争力。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTQ4OTkvZ2V0dHlpbWFnZXMtYWktYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtY2lyY3VpdC1jaGlwLmpwZyZ3PTcwMA?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-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTQ4OTkvZ2V0dHlpbWFnZXMtYWktYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtY2lyY3VpdC1jaGlwLmpwZyZ3PTcwMA"/></p>
<p>图片来源：Getty Images。</p>
<h2>1. Meta Platforms</h2>
<p><strong>Meta Platforms</strong>(META -0.22%) 可能是全球在生成性 AI 进步中受益最大的公司之一，并且它正在大量投资以实现这一目标。</p>
<p>管理层表示，今年其资本支出将高达 650 亿美元，因为它希望通过新的数据中心建设来利用生成性 AI 的潜力。虽然这个预算不如三大云计算平台那么大，但 Meta 并不将其服务器租给任何人——这完全是为了自身业务使用。</p>
<p>如果 Meta 没有看到巨大的商业机会，它就不会花这么多钱。它已经从 AI 的进展中看到了积极的回报。Meta 将其在开发大型语言模型中的经验应用于其推荐应用程序，将其扩展到跨内容格式的更一般推荐。结果是用户参与度提高，用户在其应用程序上花费的时间增加。</p>
<p>与此同时，它的早期广告创意 AI Advantage+ Creative 也得到了强劲的采用，目前有 400 万广告客户在使用它。</p>
<p>扩展生成性 AI 可能为更多小企业打开在 Meta 应用程序上投放广告的大门，即使它们没有太多经验或开销。Meta 最终可能会提供一个 AI 代理，消除开发广告活动的所有繁琐工作。Meta 首席执行官马克·扎克伯格在去年第二季度财报电话会议上表示：“从长远来看，广告客户基本上只需告诉我们一个商业目标和预算，我们就会为他们完成剩下的工作。” 这为 Meta 解锁了一大批新的广告客户，此外，广告效果的持续改善也意味着更高的平均广告价格。</p>
<p>Meta 还在为 WhatsApp 和 Messenger 开发 AI 聊天机器人，这些机器人可以为企业处理客户服务和高接触销售。根据至少一位分析师的说法，每个企业拥有一支 AI 代理的潜力可能价值 1000 亿美元。它为企业解锁的价值巨大，可能推动 Meta 的收入再上一个台阶。</p>
<p>作为一家软件公司，该公司受益于非常高的运营杠杆。AI 的创新可以为数十亿人解锁价值，这是其他公司无法比拟的规模。因此，随着其 AI 服务获得 traction 并产生更多收入，Meta 应该会看到强劲的利润增长。</p>
<p>Meta 股票的潜在价值在其股价中并未完全反映。截至目前，股票的交易价格约为未来收益的 21.2 倍。但随着用户参与度的提高和市场领先的广告产品组合的扩展，Meta 在构建其 AI 能力时应该会看到可观的盈利增长。</p>
<h2>2. 台积电</h2>
<p>如果没有 <strong>台积电</strong>(TSM 0.13%) 的制造能力，英伟达今天不会有这样的成就。这家芯片制造商主导着行业，占据了超过三分之二的半导体制造支出。而且这一份额还在不断增长。</p>
<p>公司在 TSMC 的支出增加是有充分理由的，而其他晶圆厂则在原地踏步：TSMC 是唯一能够大规模制造世界上最先进芯片的公司，而目前对先进芯片的需求非常高。</p>
<p>更重要的是，其技术领先和规模使其能够以低于竞争对手的成本制造每个芯片。这创造了一个良性循环，TSMC 赢得了更多来自领先芯片设计公司的大合同，这使其能够重新投资于技术进步和扩大制造能力。因此，TSMC 与竞争对手之间的差距正在不断扩大。</p>
<p>这意味着 TSMC 具有卓越的持续竞争力。虽然关税对半导体需求构成短期威胁，因为最终用户的价格将不得不上涨，但从长远来看，需求将继续攀升。TSMC 有望赢得这一需求的超大份额，尤其是随着云客户需要最先进的芯片，以高效利用数据中心的空间和能源。</p>
<p>没有保证英伟达会继续保持目前的地位：向超大规模客户提供绝大多数 AI 加速器。事实上，英伟达所有最大的客户都在开发自己的定制硅解决方案，同时寻找替代英伟达高价芯片的方案。但无论谁设计用于 AI 数据中心的芯片，TSMC 将是制造和封装它们的公司。几乎没有其他选择。</p>
<p>这意味着 TSMC 不应该随着时间的推移而看到任何利润侵蚀，管理层应该能够轻松维持其 53% 或更高的目标毛利率。同时，股票的交易价格仅为未来收益的 17.4 倍，这接近自 AI 热潮开始以来的最低估值。</p>

台积电

<p>Nvidia 在人工智能热潮中一直处于领先地位，近年来股价大幅上涨。然而，有预测认为，Meta Platforms 和台积电将在长期内超越 Nvidia。Meta 正在大力投资生成式人工智能，提升其广告能力和用户参与度。作为主导芯片制造商的台积电，凭借其技术领先和成本优势，确保了在半导体行业的持续增长和需求。这两家公司都处于有利位置，可以利用不断发展的人工智能市场</p>

<p><strong>Nvidia</strong>(NVDA -3.01%) has been the star of the artificial intelligence (AI) boom over the last two-and-a-half years. Its GPUs are the cornerstone of every big data center project from the hyperscale cloud customers investing tens of billions of dollars in AI. That's led to massive revenue growth and even stronger earnings growth over the last two years for Nvidia.</p>
<div class="lb-trans"><p><strong>英伟达</strong>(NVDA -3.01%) 在过去两年半的人工智能（AI）热潮中成为了明星。它的 GPU 是每个大型数据中心项目的基石，这些项目来自于投资数百亿美元于 AI 的超大规模云客户。这导致了英伟达在过去两年中收入和盈利的巨大增长。</p>
</div><p>Investors have been duly rewarded. Nvidia's stock price climbed 239% in 2023 and another 171% in 2024. That strong price appreciation led the company to briefly top the list of most valuable companies. But now, after the recent stock market sell-off, Nvidia shares currently sit around 25% off their all-time high reached in January.</p>
<div class="lb-trans"><p>投资者得到了应有的回报。英伟达的股价在 2023 年上涨了 239%，在 2024 年又上涨了 171%。如此强劲的价格上涨使得该公司短暂地登上了最有价值公司的榜单。但现在，在最近的股市抛售之后，英伟达的股价目前距离 1 月份创下的历史高点下跌了约 25%。</p>
</div><p>While many investors may be looking at the AI leader and considering buying shares at the lower price, two other artificial intelligence companies look poised to outperform the semiconductor stock over the long run, helped by competitive advantages that give them great staying power, no matter how the AI story plays out.</p>
<div class="lb-trans"><p>虽然许多投资者可能在关注这家 AI 领军企业并考虑在较低价格时购买其股票，但另外两家人工智能公司似乎有望在长期内超越这只半导体股票，得益于它们的竞争优势，使其在 AI 故事如何发展时都能保持强大的竞争力。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTQ4OTkvZ2V0dHlpbWFnZXMtYWktYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtY2lyY3VpdC1jaGlwLmpwZyZ3PTcwMA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F814899%2Fgettyimages-ai-artificial-intelligence-circuit-chip.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>1. Meta Platforms</h2>
<div class="lb-trans"><h2>1. Meta Platforms</h2>
</div><p><strong>Meta Platforms</strong>(META -0.22%) may be the biggest long-term beneficiary of advances in generative AI of any company in the world, and it's spending heavily to make that a reality.</p>
<div class="lb-trans"><p><strong>Meta Platforms</strong>(META -0.22%) 可能是全球在生成性 AI 进步中受益最大的公司之一，并且它正在大量投资以实现这一目标。</p>
</div><p>Management said its capital expenditures will climb as high as $65 billion this year as it looks to capitalize on the potential of generative AI with a new data center build-out. While that budget isn't as big as the three big cloud computing platforms, Meta doesn't rent its servers to anyone -- it's all for its own business use.</p>
<div class="lb-trans"><p>管理层表示，今年其资本支出将高达 650 亿美元，因为它希望通过新的数据中心建设来利用生成性 AI 的潜力。虽然这个预算不如三大云计算平台那么大，但 Meta 并不将其服务器租给任何人——这完全是为了自身业务使用。</p>
</div><p>The company wouldn't be spending so much if it didn't see a massive opportunity for its business. It's already seeing positive returns from its advances in AI so far. Meta applied its learnings from developing large language models to its recommendation app, expanding it to more general recommendations across content formats. The result was higher engagement and more time spent on its apps.</p>
<div class="lb-trans"><p>如果 Meta 没有看到巨大的商业机会，它就不会花这么多钱。它已经从 AI 的进展中看到了积极的回报。Meta 将其在开发大型语言模型中的经验应用于其推荐应用程序，将其扩展到跨内容格式的更一般推荐。结果是用户参与度提高，用户在其应用程序上花费的时间增加。</p>
</div><p>Meanwhile, it's seeing strong adoption of its earliest ad creative AI, Advantage+ Creative, with 4 million advertisers using it.</p>
<div class="lb-trans"><p>与此同时，它的早期广告创意 AI Advantage+ Creative 也得到了强劲的采用，目前有 400 万广告客户在使用它。</p>
</div><p>Expanding generative AI could open the door for more small businesses to buy ads on Meta's apps with little experience or overhead. Meta could eventually offer an AI agent that takes all the hard work out of developing an ad campaign. &#34;Over the long term, advertisers will basically just be able to tell us a business objective and a budget, and we're going to go do the rest for them,&#34; CEO Mark Zuckerberg said during Meta's second-quarter earnings call last year. That unlocks a slew of new advertiser for Meta, on top of continuously improving ad efficacy, which translates into higher average ad prices.</p>
<div class="lb-trans"><p>扩展生成性 AI 可能为更多小企业打开在 Meta 应用程序上投放广告的大门，即使它们没有太多经验或开销。Meta 最终可能会提供一个 AI 代理，消除开发广告活动的所有繁琐工作。Meta 首席执行官马克·扎克伯格在去年第二季度财报电话会议上表示：“从长远来看，广告客户基本上只需告诉我们一个商业目标和预算，我们就会为他们完成剩下的工作。” 这为 Meta 解锁了一大批新的广告客户，此外，广告效果的持续改善也意味着更高的平均广告价格。</p>
</div><p>Meta is also working on AI chatbots for WhatsApp and Messenger, which could handle customer service and higher-touch sales at scale for businesses. The potential of each business having a fleet of AI agents on WhatsApp could be worth $100 billion, according to at least one analyst. The value it unlocks for businesses is massive and could drive another leg up in revenue for Meta.</p>
<div class="lb-trans"><p>Meta 还在为 WhatsApp 和 Messenger 开发 AI 聊天机器人，这些机器人可以为企业处理客户服务和高接触销售。根据至少一位分析师的说法，每个企业拥有一支 AI 代理的潜力可能价值 1000 亿美元。它为企业解锁的价值巨大，可能推动 Meta 的收入再上一个台阶。</p>
</div><p>The company benefits from very high operating leverage as a software company. Innovations in AI can unlock value for literally billions of people, a scale no other company is working with. So as its AI services gain traction and generate more revenue, it should see strong profit growth.</p>
<div class="lb-trans"><p>作为一家软件公司，该公司受益于非常高的运营杠杆。AI 的创新可以为数十亿人解锁价值，这是其他公司无法比拟的规模。因此，随着其 AI 服务获得 traction 并产生更多收入，Meta 应该会看到强劲的利润增长。</p>
</div><p>A lot of the potential value of AI to Meta stock isn't fully reflected in its stock price. Shares trade for about 21.2 times forward earnings, as of this writing. But with growing engagement and an expanding suite of market-leading ad products, Meta should see substantial earnings growth as it builds out its AI capabilities.</p>
<div class="lb-trans"><p>Meta 股票的潜在价值在其股价中并未完全反映。截至目前，股票的交易价格约为未来收益的 21.2 倍。但随着用户参与度的提高和市场领先的广告产品组合的扩展，Meta 在构建其 AI 能力时应该会看到可观的盈利增长。</p>
</div><h2>2. Taiwan Semiconductor Manufacturing</h2>
<div class="lb-trans"><h2>2. 台积电</h2>
</div><p>Nvidia wouldn't be where it is today without the manufacturing capabilities of <strong>Taiwan Semiconductor Manufacturing</strong>(TSM 0.13%), commonly known as TSMC. The chip manufacturer dominates the industry, accounting for over two-thirds of all spending on semiconductor fabrication. And that share is growing.</p>
<div class="lb-trans"><p>如果没有 <strong>台积电</strong>(TSM 0.13%) 的制造能力，英伟达今天不会有这样的成就。这家芯片制造商主导着行业，占据了超过三分之二的半导体制造支出。而且这一份额还在不断增长。</p>
</div><p>There's good reason companies are spending more with TSMC while other fabs tread water: TSMC is the only company capable of manufacturing the world's most advanced chips at scale, and there's very high demand for advanced chips right now.</p>
<div class="lb-trans"><p>公司在 TSMC 的支出增加是有充分理由的，而其他晶圆厂则在原地踏步：TSMC 是唯一能够大规模制造世界上最先进芯片的公司，而目前对先进芯片的需求非常高。</p>
</div><p>What's more, its technology lead and scale allow it to manufacture chips at lower costs per chip compared to competition. That creates a virtuous cycle, whereby TSMC wins more big contracts from leading chip designers, which allows it to reinvest in advancing its technology and scaling its manufacturing capabilities. As a result, the gap between TSMC and the competition is growing wider and wider.</p>
<div class="lb-trans"><p>更重要的是，其技术领先和规模使其能够以低于竞争对手的成本制造每个芯片。这创造了一个良性循环，TSMC 赢得了更多来自领先芯片设计公司的大合同，这使其能够重新投资于技术进步和扩大制造能力。因此，TSMC 与竞争对手之间的差距正在不断扩大。</p>
</div><p>That means TSMC has exceptional staying power. While tariffs pose a short-term threat to demand for semiconductors as prices will have to increase for end users, in the long run, demand will continue to climb. TSMC is positioned to win an outsized portion of that demand, especially as cloud customers require the most advanced chips to make efficient use of space and energy in data centers.</p>
<div class="lb-trans"><p>这意味着 TSMC 具有卓越的持续竞争力。虽然关税对半导体需求构成短期威胁，因为最终用户的价格将不得不上涨，但从长远来看，需求将继续攀升。TSMC 有望赢得这一需求的超大份额，尤其是随着云客户需要最先进的芯片，以高效利用数据中心的空间和能源。</p>
</div><p>There's no guarantee Nvidia will remain in the position it currently finds itself: supplying the vast majority of AI accelerators to hyperscale customers. In fact, all Nvidia's biggest customers are developing their own custom silicon solutions while looking for alternative's to Nvidia's high-priced chips. But no matter who's designing the chips used in AI data centers, TSMC will be the one printing and packaging them. There's practically no other option.</p>
<div class="lb-trans"><p>没有保证英伟达会继续保持目前的地位：向超大规模客户提供绝大多数 AI 加速器。事实上，英伟达所有最大的客户都在开发自己的定制硅解决方案，同时寻找替代英伟达高价芯片的方案。但无论谁设计用于 AI 数据中心的芯片，TSMC 将是制造和封装它们的公司。几乎没有其他选择。</p>
</div><p>That means TSMC shouldn't see any margin erosion over time, and management should easily be able to maintain its target gross margin target of 53% or higher. Meanwhile, the stock trades for just 17.4 times forward earnings, which is near its lowest valuation since the start of the AI boom.</p>
<div class="lb-trans"><p>这意味着 TSMC 不应该随着时间的推移而看到任何利润侵蚀，管理层应该能够轻松维持其 53% 或更高的目标毛利率。同时，股票的交易价格仅为未来收益的 17.4 倍，这接近自 AI 热潮开始以来的最低估值。</p>
</div>

Prediction: These 2 Artificial Intelligence (AI) Giants Will Outperform Nvidia Over the Next 5 Years

- Nvidia 在人工智能领域的成功推动了其收入和股价的快速增长。  
- Meta 和台积电被认为在长期内有潜力超越 Nvidia，分别通过 AI 广告和芯片制造优势。  
- Meta 的 AI 投资和台积电的技术领先地位将推动各自的业务增长。  

预测：这两家人工智能（AI）巨头将在未来五年内表现优于英伟达