<p>大约 30 年前，互联网开始走向主流，彻底改变了美国企业的增长轨迹。</p>
<p>自 1990 年代中期互联网普及以来，华尔街一直在等待下一个颠覆性创新，以推动美国企业的飞跃发展。尽管许多热门趋势如 3D 打印和元宇宙相继出现又消失，但人工智能（AI）似乎回应了这一期待已久的呼声。</p>
<p>虽然估计数据自然会有所波动，但普华永道的分析师预计，AI 革命将在 2030 年使全球国内生产总值（GDP）增长 26%。这一估计的 15.7 万亿美元 GDP 增长得益于生产力的提升和各种消费侧效应。</p>
<p>到本世纪末，15.7 万亿美元的可寻址市场表明，许多企业将从中受益。然而，没有哪家公司比半导体巨头 <strong>Nvidia</strong>（NVDA 3.14%）更直接地受益于 AI 的崛起。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtYWktbWFjaGluZS1sZWFybmluZy1mYWNpYWwtcmVjb2duaWl0aW9uLWdldHR5LmpwZyZ3PTcwMA?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-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtYWktbWFjaGluZS1sZWFybmluZy1mYWNpYWwtcmVjb2duaWl0aW9uLWdldHR5LmpwZyZ3PTcwMA"/></p>
<p>图片来源：Getty Images。</p>
<h2>Nvidia 的运营扩展堪称教科书式</h2>
<p>自 2023 年初以来，Nvidia 的市值从 3600 亿美元飙升至 3.4 万亿美元。虽然有几家市场领导者曾达到 3 万亿美元的估值高峰，但从未有人在不到两年的时间内见证市值增加 3 万亿美元。</p>
<p>Nvidia 的成功可以归因于其图形处理单元（GPU）在企业数据中心中无可争议的首选地位。TechInsights 的一项研究发现，Nvidia 在 2022 年向数据中心发货了 264 万块 GPU，而在 2023 年则为 376 万块，这两年中向高计算数据中心发货的 GPU 市场份额达到了 98%。</p>
<p>Nvidia 的 H100（&#34;Hopper&#34;）芯片及其后续的 Blackwell GPU 架构正在为 AI 加速的数据中心提供核心处理能力。由于这些芯片的订单积压，该公司在为其产品定价时几乎没有遇到任何困难。</p>
<p>今年早些时候，Hopper 芯片的价格一度超过 4 万美元，约为 <strong>超微半导体</strong>Insight MI300X GPU 售价的四倍。由于 AI-GPU 的稀缺，Nvidia 的毛利率已提升至中 70% 的水平。</p>
<p>此外，Nvidia 的 CUDA 软件平台在保持客户对其产品和服务生态系统的忠诚方面表现出色。CUDA 是开发者用于构建大型语言模型并最大化其 AI-GPU 计算潜力的平台。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvaW52ZXN0b3ItcHJlc3Npbmctc2VsbC1idXR0b24tZ2V0dHkuanBnJnc9NzAw?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-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvaW52ZXN0b3ItcHJlc3Npbmctc2VsbC1idXR0b24tZ2V0dHkuanBnJnc9NzAw"/></p>
<p>图片来源：Getty Images。</p>
<h2>内部人士提供 157 个理由让投资者对 Nvidia 的上升持怀疑态度</h2>
<p>从表面上看，似乎一切对 Nvidia 来说都进行得非常顺利。但当华尔街的事情看起来过于美好时，通常都不是如此。</p>
<p>投资者可以用来评估对股票的情绪和价值的众多工具之一是内部人士活动。公司内部人士，包括董事会成员和高层管理人员，必须在购买或出售自己公司股票后不久向证券交易委员会提交 4 号表格。</p>
<p>在 2020 年 12 月 3 日，Nvidia 的首席财务官 Colette Kress 在公开市场上购买了 200 股公司股票，花费约 107,400 美元。这是 Nvidia 董事会成员或高管最后一次购买股票。</p>
<p>相比之下，自 2020 年 12 月 3 日以来，Nvidia 的内部人士已提交了 157 份 4 号表格，出售了公司股票。</p>
<p>可以理解，并非所有内部人士的出售都是坏消息或预示着麻烦。一些内部人士出售股票是为了支付税款——高管的薪酬大部分来自于基于股票的补偿，包括股票期权——而其他人则只是为了多样化投资组合或锁定一些收益。这些通常是出售的良性理由。</p>
<p>但在距离上次内部人士购买股票已经超过四年的情况下，这引发了疑问。虽然出售股票有多种原因，但内部人士购买股票只有一个原因：期待股票价值上升。在过去四年中，出售数量以 157 比 0 的比例超过购买，这清楚地表明内部人士并不认为 Nvidia 的股票具有良好的价值。</p>
<h2>历史可能是 Nvidia 无法逾越的阻力</h2>
<p>除了 Nvidia 董事会和高管在过去四年中持续的内部人士出售外，历史也是当前和潜在股东需要关注的另一个潜在痛点。</p>
<p>虽然不可否认 AI 为大多数行业带来了长期潜力，但历史清楚表明，下一代技术需要时间来成熟。在过去的三十年中，包括互联网的出现，每一个热门趋势或创新都经历了早期阶段的泡沫——没有任何迹象表明人工智能会避免这种命运。</p>
<p>更糟糕的是，大多数企业尚未提供明确的计划，说明它们将如何利用 AI 来获得其人工智能硬件和软件投资的正回报。这进一步证明投资者再次高估了新技术的早期实用性和采用率。</p>
<p>如果历史重演，人工智能泡沫破裂，我无法想象有哪只股票会比 Nvidia 受到更大的冲击，因为它几乎完全依赖于 GPU 需求的激增来实现超乎寻常的销售增长。</p>
<p>对于拥有五年以上投资时间线的投资者来说，这些担忧不太可能引起太多波动。如前所述，人工智能前景光明，市场潜力诱人。但在接下来的几个季度和几年中，英伟达的近乎抛物线的攀升突然停止也并不令人意外。</p>

英伟达

<p>英伟达的市值显著增长，主要得益于其在人工智能相关 GPU 领域的主导地位。然而，内部人士的抛售引发了怀疑，自 2020 年 12 月以来，管理层进行了 157 笔出售而没有任何购买，这表明对股票价值的信心不足。此外，历史趋势表明，新技术往往会经历泡沫，这引发了人们对英伟达在人工智能热潮中长期可持续性的担忧</p>

<p>Roughly 30 years ago, the internet began going mainstream and completely changed the growth trajectory for corporate America.</p>
<div class="lb-trans"><p>大约 30 年前，互联网开始走向主流，彻底改变了美国企业的增长轨迹。</p>
</div><p>Since the proliferation of the internet in the mid-1990s, Wall Street has been waiting for the next game-changing innovation to provide a leap forward for American businesses. While numerous buzzy trends have come and gone, including 3D printing and the metaverse, artificial intelligence (AI) appears to have answered this long-awaited call.</p>
<div class="lb-trans"><p>自 1990 年代中期互联网普及以来，华尔街一直在等待下一个颠覆性创新，以推动美国企业的飞跃发展。尽管许多热门趋势如 3D 打印和元宇宙相继出现又消失，但人工智能（AI）似乎回应了这一期待已久的呼声。</p>
</div><p>While estimates are, understandably, fluid, the analysts at PwC are looking for the AI revolution to increase global gross domestic product (GDP) by 26% in 2030. This estimated $15.7 trillion GDP boost comes courtesy of increased productivity and various consumption-side effects.</p>
<div class="lb-trans"><p>虽然估计数据自然会有所波动，但普华永道的分析师预计，AI 革命将在 2030 年使全球国内生产总值（GDP）增长 26%。这一估计的 15.7 万亿美元 GDP 增长得益于生产力的提升和各种消费侧效应。</p>
</div><p>A $15.7 trillion addressable market by the turn of the decade suggests plenty of businesses are going to benefit. However, no company has been a more direct beneficiary of the rise of AI than semiconductor titan <strong>Nvidia</strong>(NVDA 3.14%).</p>
<div class="lb-trans"><p>到本世纪末，15.7 万亿美元的可寻址市场表明，许多企业将从中受益。然而，没有哪家公司比半导体巨头 <strong>Nvidia</strong>（NVDA 3.14%）更直接地受益于 AI 的崛起。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvYXJ0aWZpY2lhbC1pbnRlbGxpZ2VuY2UtYWktbWFjaGluZS1sZWFybmluZy1mYWNpYWwtcmVjb2duaWl0aW9uLWdldHR5LmpwZyZ3PTcwMA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F800504%2Fartificial-intelligence-ai-machine-learning-facial-recogniition-getty.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>Nvidia's operating expansion has been textbook</h2>
<div class="lb-trans"><h2>Nvidia 的运营扩展堪称教科书式</h2>
</div><p>Since the start of 2023, Nvidia's market cap has skyrocketed from $360 billion to $3.4 trillion. While a couple of market leaders have reached the $3 trillion valuation plateau before, no one has ever witnessed a $3 trillion increase in market cap in less than two years.</p>
<div class="lb-trans"><p>自 2023 年初以来，Nvidia 的市值从 3600 亿美元飙升至 3.4 万亿美元。虽然有几家市场领导者曾达到 3 万亿美元的估值高峰，但从未有人在不到两年的时间内见证市值增加 3 万亿美元。</p>
</div><p>Nvidia's success can be explained by its graphics processing units (GPUs) being the undisputed top choice in enterprise data centers. A study from TechInsights found that Nvidia shipped 2.64 million GPUs to data centers in 2022 and 3.76 million in 2023, which represents a 98% share of GPUs shipped to high-compute data centers over these two years.</p>
<div class="lb-trans"><p>Nvidia 的成功可以归因于其图形处理单元（GPU）在企业数据中心中无可争议的首选地位。TechInsights 的一项研究发现，Nvidia 在 2022 年向数据中心发货了 264 万块 GPU，而在 2023 年则为 376 万块，这两年中向高计算数据中心发货的 GPU 市场份额达到了 98%。</p>
</div><p>Nvidia's H100 (&#34;Hopper&#34;) chip and successor Blackwell GPU architecture are serving as the brains of AI-accelerated data centers. With orders for these chips backlogged, the company has had little trouble commanding a premium price for its products.</p>
<div class="lb-trans"><p>Nvidia 的 H100（&#34;Hopper&#34;）芯片及其后续的 Blackwell GPU 架构正在为 AI 加速的数据中心提供核心处理能力。由于这些芯片的订单积压，该公司在为其产品定价时几乎没有遇到任何困难。</p>
</div><p>Earlier this year, the Hopper chip was briefly fetching north of $40,000, which is roughly quadruple what <strong>Advanced Micro Devices</strong> is netting for its Insight MI300X GPUs. A substantially higher price point fueled by AI-GPU scarcity has lifted Nvidia's gross margin to the mid-70% range.</p>
<div class="lb-trans"><p>今年早些时候，Hopper 芯片的价格一度超过 4 万美元，约为 <strong>超微半导体</strong>Insight MI300X GPU 售价的四倍。由于 AI-GPU 的稀缺，Nvidia 的毛利率已提升至中 70% 的水平。</p>
</div><p>Additionally, Nvidia's CUDA software platform has been masterful in keeping customers loyal to its ecosystem of products and services. CUDA is the platform used by developers to build large language models and maximize the computing potential of their AI-GPUs.</p>
<div class="lb-trans"><p>此外，Nvidia 的 CUDA 软件平台在保持客户对其产品和服务生态系统的忠诚方面表现出色。CUDA 是开发者用于构建大型语言模型并最大化其 AI-GPU 计算潜力的平台。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MDA1MDQvaW52ZXN0b3ItcHJlc3Npbmctc2VsbC1idXR0b24tZ2V0dHkuanBnJnc9NzAw" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F800504%2Finvestor-pressing-sell-button-getty.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>Insiders offer 157 reasons for investors to be skeptical of Nvidia's ascent</h2>
<div class="lb-trans"><h2>内部人士提供 157 个理由让投资者对 Nvidia 的上升持怀疑态度</h2>
</div><p>On paper, seemingly everything has gone perfectly for Nvidia. But when things seem too good to be true on Wall Street, they usually are.</p>
<div class="lb-trans"><p>从表面上看，似乎一切对 Nvidia 来说都进行得非常顺利。但当华尔街的事情看起来过于美好时，通常都不是如此。</p>
</div><p>One of the many tools investors can use to gauge both sentiment and value toward a stock is insider activity. Company insiders, which includes board directors and high-ranking members of the executive team, are required to file Form 4 with the Securities and Exchange Commission shortly after they purchase or sell shares of their own company's stock.</p>
<div class="lb-trans"><p>投资者可以用来评估对股票的情绪和价值的众多工具之一是内部人士活动。公司内部人士，包括董事会成员和高层管理人员，必须在购买或出售自己公司股票后不久向证券交易委员会提交 4 号表格。</p>
</div><p>On Dec. 3, 2020, Nvidia's Chief Financial Officer Colette Kress purchased 200 shares of her company's stock on the open market for approximately $107,400. This marks the last time an Nvidia board member or executive has purchased shares.</p>
<div class="lb-trans"><p>在 2020 年 12 月 3 日，Nvidia 的首席财务官 Colette Kress 在公开市场上购买了 200 股公司股票，花费约 107,400 美元。这是 Nvidia 董事会成员或高管最后一次购买股票。</p>
</div><p>Comparatively, Nvidia insiders have filed 157 separate Form 4s since Dec. 3, 2020, selling shares of their company's stock.</p>
<div class="lb-trans"><p>相比之下，自 2020 年 12 月 3 日以来，Nvidia 的内部人士已提交了 157 份 4 号表格，出售了公司股票。</p>
</div><p>Understandably, not all insider selling is necessarily bad news or portends trouble. Some insiders sell to cover their tax bill -- executives may generate the lion's share of their pay from stock-based compensation, including stock options -- while others are simply diversifying their portfolios or locking in some gains. These are generally benign reasons to sell.</p>
<div class="lb-trans"><p>可以理解，并非所有内部人士的出售都是坏消息或预示着麻烦。一些内部人士出售股票是为了支付税款——高管的薪酬大部分来自于基于股票的补偿，包括股票期权——而其他人则只是为了多样化投资组合或锁定一些收益。这些通常是出售的良性理由。</p>
</div><p>But when it's been over four years since the last insider purchase, it raises questions. Whereas there are multiple reasons to sell stock, there's only one reason for insiders to buy: the expectation of shares increasing in value. With sells outnumbering buys 157 to zero over the last four years, it paints a pretty clear picture that insiders don't believe Nvidia stock is a good value.</p>
<div class="lb-trans"><p>但在距离上次内部人士购买股票已经超过四年的情况下，这引发了疑问。虽然出售股票有多种原因，但内部人士购买股票只有一个原因：期待股票价值上升。在过去四年中，出售数量以 157 比 0 的比例超过购买，这清楚地表明内部人士并不认为 Nvidia 的股票具有良好的价值。</p>
</div><h2>History may be an insurmountable headwind for Nvidia</h2>
<div class="lb-trans"><h2>历史可能是 Nvidia 无法逾越的阻力</h2>
</div><p>In addition to a steady dose of insider selling by Nvidia's board and executives over the last four years, history is another potential sore spot for current and prospective shareholders to monitor.</p>
<div class="lb-trans"><p>除了 Nvidia 董事会和高管在过去四年中持续的内部人士出售外，历史也是当前和潜在股东需要关注的另一个潜在痛点。</p>
</div><p>While there's no denying the long-term potential AI brings to the table for most sectors and industries, history is crystal clear that next-big-thing technologies need time to mature. For three decades, which includes the advent of the internet, every hot trend or innovation has navigated its way through an early stage bubble -- and there's nothing that suggests artificial intelligence will avoid this fate.</p>
<div class="lb-trans"><p>虽然不可否认 AI 为大多数行业带来了长期潜力，但历史清楚表明，下一代技术需要时间来成熟。在过去的三十年中，包括互联网的出现，每一个热门趋势或创新都经历了早期阶段的泡沫——没有任何迹象表明人工智能会避免这种命运。</p>
</div><p>To add fuel to the fire, most businesses have yet to provide clear plans as to how they'll use AI to generate a positive return on their artificial intelligence hardware and software investments. This serves as further evidence that investors have, once again, overestimated the early stage utility and adoption rate of a new technology.</p>
<div class="lb-trans"><p>更糟糕的是，大多数企业尚未提供明确的计划，说明它们将如何利用 AI 来获得其人工智能硬件和软件投资的正回报。这进一步证明投资者再次高估了新技术的早期实用性和采用率。</p>
</div><p>If history rhymes and the artificial intelligence bubble does burst, I can't imagine a stock being hit harder than Nvidia, which has relied almost exclusively on an influx of GPU demand to deliver otherworldly sales growth.</p>
<div class="lb-trans"><p>如果历史重演，人工智能泡沫破裂，我无法想象有哪只股票会比 Nvidia 受到更大的冲击，因为它几乎完全依赖于 GPU 需求的激增来实现超乎寻常的销售增长。</p>
</div><p>For investors with a five-plus-year timeline, these concerns are unlikely to ruffle too many feathers. As noted, AI has a promising outlook and a tantalizing addressable market. But in the quarters and few years that lie ahead, it wouldn't be a surprise if Nvidia's near-parabolic climb came to an abrupt halt.</p>
<div class="lb-trans"><p>对于拥有五年以上投资时间线的投资者来说，这些担忧不太可能引起太多波动。如前所述，人工智能前景光明，市场潜力诱人。但在接下来的几个季度和几年中，英伟达的近乎抛物线的攀升突然停止也并不令人意外。</p>
</div>

157 Reasons for Investors to Be Skeptical of Nvidia's Near-Parabolic Gains

- 人工智能预计到 2030 年将推动全球 GDP 增长 26%。  
- Nvidia 在 AI 领域的成功使其市值在 2023 年大幅上涨。  
- 然而，内部人士的 157 次卖出引发对其未来价值的质疑。

投资者对英伟达近乎抛物线式增长持怀疑态度的 157 个原因