<p>2024 年最热门的两只人工智能（AI）股票无疑是 <strong>Palantir</strong>（PLTR -11.49%）和 <strong>Nvidia</strong>（NVDA -7.03%）。然而，在去年取得巨大收益后，这两只股票在最近的市场回调中都远离了其高点。</p>
<p>让我们看看目前哪只股票是投资者更好的投资选择。</p>
<h2>两个 AI 赢家</h2>
<p>Palantir 和 Nvidia 有着截然不同的业务，但两者都是 AI 领域的赢家。</p>
<p>Nvidia 是一家半导体公司，生产图形处理单元（GPU）。由于其快速的处理速度，这些芯片已成为 AI 基础设施的支柱，证明非常适合训练 AI 模型和进行推理。该公司通过其 CUDA 软件平台建立了广泛的护城河，该平台早在 2006 年就开发出来，以便让其芯片能够被编程用于各种用途。如今，它在此基础上构建了领先的库和服务，专为 AI 设计，使其芯片备受追捧。</p>
<p>另一方面，Palantir 是一家软件分析公司。它最初在政府领域崭露头角，其数据收集和分析能力被用于诸如反恐等关键任务。然而，随着其 AI 平台的出现，它已转型为一家 AI 操作系统公司，帮助客户设计和部署各种用例的 AI 解决方案。</p>
<p>这两家公司都在经历强劲的增长。由于大型科技公司和 AI 初创企业竞相构建 AI 基础设施，Nvidia 的收入在过去两年中翻了一番。AI 基础设施的支出持续上升，主要由三大云计算公司主导，这三家公司计划在今年在与构建 AI 基础设施相关的资本支出（capex）上花费高达 2500 亿美元。就 Nvidia 而言，该公司预测到 2028 年，整体数据中心相关的资本支出将超过 1 万亿美元。</p>
<p>与此同时，Palantir 的增长加速，商业客户纷纷涌向其 AI 平台，联邦政府也开始接受 AI。上个季度，整体收入增长了 36%，而美国商业收入飙升了 64%，美国政府收入增长了 45%。随着公司通过其 AI 训练营吸引新商业客户，其客户数量增长了 43%。</p>
<p>到目前为止，许多商业客户仍处于概念验证阶段。因此，Palantir 在将这些客户的解决方案投入生产以解决现实问题时，面临着巨大的机会。</p>
<h2>风险</h2>
<p>谈到风险，Nvidia 面临的最大风险是 AI 基础设施支出的放缓。该公司的 CUDA 软件平台是免费的，因此没有大规模的经常性收入流。相反，它必须不断销售更多的芯片以继续增长。</p>
<p>尽管 AI 基础设施支出仍在上升，但有一些担忧认为这种支出的速度最终会放缓。据报道，Nvidia 的最大客户 <strong>Microsoft</strong> 已在一些数据中心项目上有所收缩，因为它认为供需之间可能存在过剩。然而，TD Cowen 的分析师指出，云计算竞争对手 <strong>Alphabet</strong> 和 <strong>Amazon</strong> 已介入以填补这一空缺。</p>
<p>只要公司继续竞相构建更好的 AI 模型，它们就需要更多的计算能力，而这通常由 GPU 提供。事实上，随着模型的进步，它们往往需要指数级增加的 AI 芯片进行训练。例如，<strong>Meta Platforms</strong> 和 xAI 的最新模型所需的 GPU 数量是其之前版本的约 10 倍。</p>
<p>与此同时，Palantir 面临着与美国政府当前预算削减相关的潜在风险，该政府是其最大客户，去年占其收入的 40% 以上。该公司与国防部（DOD）及军事相关支出特别紧密相关。作为政府效率部（DOGE）的一部分，特朗普政府已要求国防部在未来五年内每年削减 8% 的预算。</p>
<p>这对国防部的预算是一个巨大的削减，将影响许多项目。它对 Palantir 或其增长机会的影响程度仍然未知。Palantir 首席执行官 Alex Karp 公开表示支持 DOGE，并暗示公司可能会受益。然而，他和其他公司内部人士也在抛售 Palantir 的股票。不过，如果 Palantir 的 AI 平台能够证明其提高了效率并帮助降低成本，确实有可能成为 DOGE 的赢家。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTI4NzAvZ2V0dHlpbWFnZXMtMTU4MjY1MzQ4OC5qcGcmdz03MDA?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-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTI4NzAvZ2V0dHlpbWFnZXMtMTU4MjY1MzQ4OC5qcGcmdz03MDA"/></p>
<p>图片来源：Getty Images。</p>
<h2>估值</h2>
<p>Nvidia 和 Palantir 股票之间的一个重大区别是它们的估值。在撰写本文时，Nvidia 的股票相对便宜，基于今年的分析师预估，其股票的前瞻市盈率（P/E）约为 24 倍，市盈率与增长比（PEG）比率略高于 0.4。PEG 比率低于 1 的股票通常被认为被低估，因此根据这一指标，Nvidia 是一笔不错的交易。</p>
<p>而 Palantir 的股票则相对昂贵。该股票的前瞻市销率（P/S）为 53，是 2021 年类似增长率的服务型软件（SaaS）股票的峰值市销率的两倍多。请注意，对于 Nvidia 的股票，我们关注的是基于收益的估值，而对于 Palantir 的股票，我们关注的是基于收入的估值，因此两者之间的差异相当巨大。</p>
<p>鉴于它们的经常性收入模型，软件公司应该比半导体公司有更高的估值。然而，这两家公司之间的估值差异仍然相当明显。两家公司都有潜在的增长动力和潜在的风险。因此，我目前更倾向于英伟达，如果人工智能基础设施支出继续增长，它将是一个更大的投资机会。</p>

英伟达财报，一站看懂！

英伟达

PLTR

<p>在 2024 年，Palantir 和 Nvidia 是领先的人工智能股票，但两者在去年收益后都出现了下滑。Nvidia 是一家半导体公司，在其 GPU 的支持下，在人工智能基础设施方面表现出色，而 Palantir 是一家软件分析公司，专注于人工智能解决方案。Nvidia 的收入翻了一番，得益于人工智能基础设施支出的增加，而 Palantir 的收入增长了 36%，在商业和政府部门取得了显著的增长。风险包括 Nvidia 可能面临的人工智能支出放缓，以及美国政府预算削减对 Palantir 的影响。估值方面，Nvidia 的股票目前被视为便宜</p>

<p>Two of the hottest artificial intelligence (AI) stocks in 2024 were, no doubt, <strong>Palantir</strong>(PLTR -11.49%) and <strong>Nvidia</strong>(NVDA -7.03%). However, after posting huge gains last year, both stocks now find themselves well off their highs following the recent market pullback.</p>
<div class="lb-trans"><p>2024 年最热门的两只人工智能（AI）股票无疑是 <strong>Palantir</strong>（PLTR -11.49%）和 <strong>Nvidia</strong>（NVDA -7.03%）。然而，在去年取得巨大收益后，这两只股票在最近的市场回调中都远离了其高点。</p>
</div><p>Let's look at which stock is the better investment option for investors right now.</p>
<div class="lb-trans"><p>让我们看看目前哪只股票是投资者更好的投资选择。</p>
</div><h2>Two AI winners</h2>
<div class="lb-trans"><h2>两个 AI 赢家</h2>
</div><p>Palantir and Nvidia have two very different businesses, but both have been big AI winners.</p>
<div class="lb-trans"><p>Palantir 和 Nvidia 有着截然不同的业务，但两者都是 AI 领域的赢家。</p>
</div><p>Nvidia is a semiconductor company that makes graphic processing units (GPUs). Its chips have become the backbone of AI infrastructure due to their fast processing speeds, which have proven ideal for training AI models and running inference. The company has created a wide moat through its CUDA software platform, which it developed way back in 2006 to allow its chips to be programmed for various purposes. Today, it has built on top of this program to have leading libraries and services designed for AI, which makes its chips so desirable.</p>
<div class="lb-trans"><p>Nvidia 是一家半导体公司，生产图形处理单元（GPU）。由于其快速的处理速度，这些芯片已成为 AI 基础设施的支柱，证明非常适合训练 AI 模型和进行推理。该公司通过其 CUDA 软件平台建立了广泛的护城河，该平台早在 2006 年就开发出来，以便让其芯片能够被编程用于各种用途。如今，它在此基础上构建了领先的库和服务，专为 AI 设计，使其芯片备受追捧。</p>
</div><p>Palantir, on the other hand, is a software analytics company. It initially made a name for itself in the government space, where its data gathering and analytics capabilities were used for mission-critical tasks such as fighting terrorism. However, with the advent of its AI platform, it has transformed into an AI operating system company that helps customers design and deploy AI solutions for various use cases.</p>
<div class="lb-trans"><p>另一方面，Palantir 是一家软件分析公司。它最初在政府领域崭露头角，其数据收集和分析能力被用于诸如反恐等关键任务。然而，随着其 AI 平台的出现，它已转型为一家 AI 操作系统公司，帮助客户设计和部署各种用例的 AI 解决方案。</p>
</div><p>Both companies have been seeing strong growth. Nvidia's revenue has more than doubled each of the past two years as large tech companies and AI start-ups race to build out AI infrastructure. Spending for AI infrastructure continues to rise, led by the big three cloud computing companies, which combined plan to spend a whopping $250 billion on capital expenditures (capex) this year related to building out their AI infrastructure. For its part, Nvidia has predicted that overall data center-related capex will rise to over $1 trillion by 2028.</p>
<div class="lb-trans"><p>这两家公司都在经历强劲的增长。由于大型科技公司和 AI 初创企业竞相构建 AI 基础设施，Nvidia 的收入在过去两年中翻了一番。AI 基础设施的支出持续上升，主要由三大云计算公司主导，这三家公司计划在今年在与构建 AI 基础设施相关的资本支出（capex）上花费高达 2500 亿美元。就 Nvidia 而言，该公司预测到 2028 年，整体数据中心相关的资本支出将超过 1 万亿美元。</p>
</div><p>Palantir, meanwhile, has seen accelerating growth as commercial customers flock to its AI platform and the federal government starts to embrace AI. Overall revenue growth climbed 36% last quarter, while U.S. commercial revenue soared 64%, and U.S. government revenue jumped 45%. Its customer count has grown 43% as the company attracts new commercial customers through its AI bootcamps.</p>
<div class="lb-trans"><p>与此同时，Palantir 的增长加速，商业客户纷纷涌向其 AI 平台，联邦政府也开始接受 AI。上个季度，整体收入增长了 36%，而美国商业收入飙升了 64%，美国政府收入增长了 45%。随着公司通过其 AI 训练营吸引新商业客户，其客户数量增长了 43%。</p>
</div><p>Thus far, many of its commercial customers are still in the proof-of-concept phase. So, Palantir has a big opportunity as it moves these customers' solutions into production to tackle real-world problems.</p>
<div class="lb-trans"><p>到目前为止，许多商业客户仍处于概念验证阶段。因此，Palantir 在将这些客户的解决方案投入生产以解决现实问题时，面临着巨大的机会。</p>
</div><h2>Risks</h2>
<div class="lb-trans"><h2>风险</h2>
</div><p>When it comes to risks, the biggest for Nvidia is a slowdown in AI infrastructure spending. The company's CUDA software platform is free, so it doesn't have a big recurring revenue stream. Instead, it must sell more and more chips to continue to grow.</p>
<div class="lb-trans"><p>谈到风险，Nvidia 面临的最大风险是 AI 基础设施支出的放缓。该公司的 CUDA 软件平台是免费的，因此没有大规模的经常性收入流。相反，它必须不断销售更多的芯片以继续增长。</p>
</div><p>While AI infrastructure spending is still on the rise, there are some concerns that the pace of this spending will eventually slow. It has been reported that <strong>Microsoft</strong>, which is Nvidia's largest customer, has pulled back on some data center projects, as it thinks there could be an overcapacity of supply versus demand. However, analysts at TD Cowen have noted that cloud computing rivals <strong>Alphabet</strong> and <strong>Amazon</strong> have stepped in to backfill this capacity.</p>
<div class="lb-trans"><p>尽管 AI 基础设施支出仍在上升，但有一些担忧认为这种支出的速度最终会放缓。据报道，Nvidia 的最大客户 <strong>Microsoft</strong> 已在一些数据中心项目上有所收缩，因为它认为供需之间可能存在过剩。然而，TD Cowen 的分析师指出，云计算竞争对手 <strong>Alphabet</strong> 和 <strong>Amazon</strong> 已介入以填补这一空缺。</p>
</div><p>As long as companies continue to race to build out better AI models, they will need more computing power, which tends to be supplied by GPUs. In fact, as models have advanced, they have tended to need exponentially more AI chips to be trained on. For example, the newest models from both <strong>Meta Platforms</strong> and xAI have been trained on about 10 times as many GPUs as their prior versions.</p>
<div class="lb-trans"><p>只要公司继续竞相构建更好的 AI 模型，它们就需要更多的计算能力，而这通常由 GPU 提供。事实上，随着模型的进步，它们往往需要指数级增加的 AI 芯片进行训练。例如，<strong>Meta Platforms</strong> 和 xAI 的最新模型所需的 GPU 数量是其之前版本的约 10 倍。</p>
</div><p>Palantir, meanwhile, faces a potential risk related to the current budget cuts from the U.S. government, which is its largest customer, accounting for more than 40% of its revenue last year. The company is particularly tied to the Department of Defense (DOD) and military-related spending. As part of Department of Government Efficiency (DOGE), the Trump administration has asked the DOD to reduce its budget by 8% annually over the next five years.</p>
<div class="lb-trans"><p>与此同时，Palantir 面临着与美国政府当前预算削减相关的潜在风险，该政府是其最大客户，去年占其收入的 40% 以上。该公司与国防部（DOD）及军事相关支出特别紧密相关。作为政府效率部（DOGE）的一部分，特朗普政府已要求国防部在未来五年内每年削减 8% 的预算。</p>
</div><p>That's a huge cut to the DOD's budget, which will impact a lot of programs. How much or how little it impacts Palantir or its growth opportunities is still unknown. Palantir CEO Alex Karp has publicly stated that he supports DOGE and has hinted that the company could benefit. However, he and other company insiders have also been dumping Palantir stock. Still, it is certainly possible that if Palantir's AI platform can show that it improves efficiency and helps lower costs, it could be a DOGE winner.</p>
<div class="lb-trans"><p>这对国防部的预算是一个巨大的削减，将影响许多项目。它对 Palantir 或其增长机会的影响程度仍然未知。Palantir 首席执行官 Alex Karp 公开表示支持 DOGE，并暗示公司可能会受益。然而，他和其他公司内部人士也在抛售 Palantir 的股票。不过，如果 Palantir 的 AI 平台能够证明其提高了效率并帮助降低成本，确实有可能成为 DOGE 的赢家。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MTI4NzAvZ2V0dHlpbWFnZXMtMTU4MjY1MzQ4OC5qcGcmdz03MDA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F812870%2Fgettyimages-1582653488.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>Valuations</h2>
<div class="lb-trans"><h2>估值</h2>
</div><p>One of the big differences between Nvidia's and Palantir's stocks is their valuations. At the time of this writing, Nvidia's shares are quite cheap, with the stock trading at a forward price-to-earnings (P/E) ratio of around 24 times based on this year's analyst estimates, with a price/earnings-to-growth (PEG) ratio of just over 0.4. Stocks with PEG ratios below 1 are typically considered undervalued, making Nvidia a nice bargain according to this metric.</p>
<div class="lb-trans"><p>Nvidia 和 Palantir 股票之间的一个重大区别是它们的估值。在撰写本文时，Nvidia 的股票相对便宜，基于今年的分析师预估，其股票的前瞻市盈率（P/E）约为 24 倍，市盈率与增长比（PEG）比率略高于 0.4。PEG 比率低于 1 的股票通常被认为被低估，因此根据这一指标，Nvidia 是一笔不错的交易。</p>
</div><p>Palantir's stock, on the other hand, is quite expensive. The stock trades at a forward price-to-sales (P/S) multiple of 53, which is more than double the peak multiples that software-as-a-service (SaaS) stocks traded at in 2021 with similar growth rates. Note that with Nvidia's stock, we are looking at its valuation based on earnings, while with Palantir's stock, we are looking at its valuation based on revenue, so the difference is pretty huge.</p>
<div class="lb-trans"><p>而 Palantir 的股票则相对昂贵。该股票的前瞻市销率（P/S）为 53，是 2021 年类似增长率的服务型软件（SaaS）股票的峰值市销率的两倍多。请注意，对于 Nvidia 的股票，我们关注的是基于收益的估值，而对于 Palantir 的股票，我们关注的是基于收入的估值，因此两者之间的差异相当巨大。</p>
</div><p>Given their recurring revenue models, software companies should trade at much higher valuations than semiconductor companies. However, the valuation difference between the two companies is still quite stark. Both companies have potential growth drivers and potential risks. As such, I prefer Nvidia at the moment, which is just the much bigger bargain if AI infrastructure spending continues to grow.</p>
<div class="lb-trans"><p>鉴于它们的经常性收入模型，软件公司应该比半导体公司有更高的估值。然而，这两家公司之间的估值差异仍然相当明显。两家公司都有潜在的增长动力和潜在的风险。因此，我目前更倾向于英伟达，如果人工智能基础设施支出继续增长，它将是一个更大的投资机会。</p>
</div>

Better Artificial Intelligence Stock: Palantir vs. Nvidia

- 2024 年，Palantir 和 Nvidia 是热门的人工智能股票。  
- Nvidia 的市盈率约为 24 倍，Palantir 的市销率为 53 倍。  
- Nvidia 在 AI 基础设施支出增长中更具投资价值。  

更好的人工智能股票：Palantir Tech vs. 英伟达