<p>毫无疑问，<strong>Nvidia</strong>（NVDA -4.84%）已经赢得了作为全球最大上市公司的地位。它的技术是大多数人工智能（AI）平台的核心，而人工智能本身正努力成为有史以来最具革命性的发明之一。这就是为什么如此多的国内外公司继续购买 Nvidia 能够销售给他们的尽可能多的加速器。</p>
<p>不过，正如可以预见的那样，该公司在 AI 硬件市场的近乎垄断地位不会永远不受制约。像 <strong>Intel</strong> 和 <strong>Advanced Micro Devices</strong> 这样的竞争对手也希望抓住这个机会，而各国政府则对其业务过于依赖单一关键技术供应商感到担忧。</p>
<p>在中国，这种情况尤其明显。虽然在行业初期，北京对使用美国制造的 AI 处理器持宽容态度，但贸易紧张局势和国家安全担忧很快促使了本土的应对措施。这一点并不令人惊讶。</p>
<p>令人惊讶的是，领导中国 AI 硬件开发工作的组织是 <strong>Alibaba Group</strong>（BABA -8.60%）。没错，就是中国的电子商务巨头。更令人惊讶的是，这家电子商务巨头在 AI 开发方面表现出色，甚至可能成为中国的 Nvidia。</p>
<p>如果这真的是未来的发展方向，投资者将会获得丰厚的回报。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzU5MTkvY29tcHV0ZXItY2lyY3VpdC1ib2FyZC5qcGcmdz03MDA?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-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzU5MTkvY29tcHV0ZXItY2lyY3VpdC1ib2FyZC5qcGcmdz03MDA"/></p>
<p>图片来源：Getty Images。</p>
<h2>阿里巴巴正在引领中国的 AI 发展</h2>
<p>阿里巴巴的大部分收入仍来自电子商务。2025 年第二季度，其销售额的超过一半来自国内电子商务业务，另有 15% 来自国际电子商务业务。</p>
<p>不过，阿里巴巴不仅仅是电子商务。虽然相对较小，但公司的云计算部门——包括其新兴的 AI 业务——占其总销售额的 15%，并且是第二季度增长最快的部门，同比增长了 26%。如果继续以当前的速度扩张，它最终可能成为公司的最大业务。</p>
<p>考虑到公司在这一领域的进展，这种可能性不仅存在，而且甚至很可能。</p>
<p>虽然阿里巴巴已经在开发其 T-Head 并行处理单元一段时间，但上个月发布的处理器芯片版本值得注意。具体来说，它几乎完美匹配 Nvidia 旗舰 H20 GPU 的性能规格，后者是为海外用户出口的。从理论上讲，没有理由让中国公司选择 H20 而不是阿里巴巴的硅，特别是后者的成本低约 40%。</p>
<p>不过，这不仅仅是硬件。阿里巴巴通常与其 AI 解决方案配套的软件至少与 Nvidia 的 CUDA 生态系统一样具有市场竞争力，甚至可能更具优势。这是因为阿里巴巴在构建其 AI 平台时，主要是以便于使用开源软件的方式进行，这通常比像 Nvidia 的 CUDA 这样的封闭系统更灵活。</p>
<p>公司还向所有潜在客户明确表示，其云解决方案和人工智能解决方案是一起购买的，许多工具旨在服务于现有的云客户。</p>
<p>鉴于阿里巴巴在其本国及邻近国家所拥有的机会，如果该公司在这些地区取得与 Nvidia 相当的成功，也不要感到惊讶。</p>
<h2>机会在等待</h2>
<p><strong>摩根大通</strong> 最近的一项分析说明了一切。正如其研究人员在回应中国最近的 AI 发展工作时直言不讳地指出的那样，“沉睡的巨人觉醒了。” 量化这一评论，摩根大通的分析师认为，中国的 AI 产业到 2030 年可能价值 1.4 万亿美元，这意味着在此期间对该技术的投资回报率可达 52%。</p>
<p>当然，这项技术将在中国以外作为服务出售，但它也将在国内用于提高国家自身的商业效率。<strong>高盛</strong> 预计，国家到 2030 年将经历的任何国内生产总值（GDP）增长率，现在将因中国对 AI 的日益采用而提高 20 到 30 个基点。</p>
<p>阿里巴巴的处理器并不会对这一扩张负全部责任。<strong>百度</strong>（BIDU -8.09%）和华为也在开发自己的 AI 处理器。事实上，百度的昆仑处理器在技术上是中国首个国产 AI 硅。</p>
<p>不过，这并不构成对阿里巴巴未来主导地位的直接威胁。百度的昆仑是一款主要面向移动市场的云到边缘 AI 芯片。它没有阿里巴巴的 T-Head 那种原始计算能力，后者预计能够处理更重的推理工作，比如在没有特定数据或明确历史可供参考时，如何回应新的查询。实际上，中国国有电信公司中国联通在这些芯片正式进入大规模生产之前，已经为青海的新数据中心订购了一批阿里巴巴的 T-Head 处理器。</p>
<p>这可能只是未来需求的一个缩影。再次强调，摩根大通预计中国的 AI 市场到 2030 年将增长到 1.4 万亿美元，而该国目前的总 GDP 略低于 20 万亿美元。</p>
<h2>虽然不是 Nvidia，但仍然值得购买，或者很快就会值得购买</h2>
<p>阿里巴巴会否复制 Nvidia 在过去三年中所取得的所有成功，尤其是在 ChatGPT 引发 AI 军备竞赛之后？更进一步，阿里巴巴的股票会否在这一期间实现与 Nvidia 股票相同的 1400% 的涨幅？</p>
<p>这两个问题的答案可能都是否定的。</p>
<p>不过，不要失去对比较的视角。英伟达在正确的时间、正确的地点推出了合适的产品，进一步得益于没有人预见到在 COVID-19 大流行后会出现这样的现象。这其中有不少运气成分。即使阿里巴巴从现在起的表现只有英伟达的一半，它仍然会是一个极好的投资。</p>
<p>是的，阿里巴巴确实看起来有可能成为中国的英伟达，即使它的净机会明显较小。分析师认为，该公司的收入不仅将在 2027 年之前持续增长，而且整个过程中还会加速。</p>
<p>如果你感兴趣，就不要拖延太久。尽管阿里巴巴的股票目前有些过热，且可以说是时候调整了，但这种调整不太可能将该股票拖回到最近的主要低点。尽量抓住任何合理的折扣。</p>

阿里巴巴

阿里巴巴-W

2 倍做多阿里巴巴 ETF - KraneShares

BABO

2 倍做多 BABA ETF - GraniteShares

<p>阿里巴巴正在成为中国人工智能硬件发展的关键参与者，可能会与英伟达形成竞争。尽管电子商务仍然是其主要收入来源，但阿里巴巴的云计算和人工智能业务正在快速增长，占销售额的 15%，同比增长 26%。其 T-Head 处理器在成本更低的情况下与英伟达的性能相匹配，公司正在将人工智能解决方案与其云服务整合。分析师预测，到 2030 年，中国的人工智能产业可能达到 1.4 万亿美元，推动 GDP 增长 20-30 个基点，阿里巴巴有望抓住这一机会</p>

<p>There's no denying <strong>Nvidia</strong>(NVDA -4.84%) has earned its place as the world's biggest publicly traded company. Its technology is the heart and soul of most of the planet's artificial intelligence (AI) platforms, and AI itself is working to become one of the most revolutionary inventions of all time. That's why so many companies here and abroad continue buying as many of Nvidia's accelerators as Nvidia can sell them.</p>
<div class="lb-trans"><p>毫无疑问，<strong>Nvidia</strong>（NVDA -4.84%）已经赢得了作为全球最大上市公司的地位。它的技术是大多数人工智能（AI）平台的核心，而人工智能本身正努力成为有史以来最具革命性的发明之一。这就是为什么如此多的国内外公司继续购买 Nvidia 能够销售给他们的尽可能多的加速器。</p>
</div><p>As could have been predicted, though, the company's near monopoly of the AI hardware market wasn't going to remain unchecked forever. Competitors like <strong>Intel</strong> and <strong>Advanced Micro Devices</strong> want to capitalize on the opportunity as well, while governments grow leery of their businesses becoming too dependent on a single supplier of <em>any</em> crucial technology.</p>
<div class="lb-trans"><p>不过，正如可以预见的那样，该公司在 AI 硬件市场的近乎垄断地位不会永远不受制约。像 <strong>Intel</strong> 和 <strong>Advanced Micro Devices</strong> 这样的竞争对手也希望抓住这个机会，而各国政府则对其业务过于依赖单一关键技术供应商感到担忧。</p>
</div><p>That's certainly been the case in China, anyway. While Beijing was tolerant of the use of American-made AI processors when the industry was in its infancy, trade tensions and national security concerns soon prompted a homegrown response. No surprises there.</p>
<div class="lb-trans"><p>在中国，这种情况尤其明显。虽然在行业初期，北京对使用美国制造的 AI 处理器持宽容态度，但贸易紧张局势和国家安全担忧很快促使了本土的应对措施。这一点并不令人惊讶。</p>
</div><p>What's surprising is the organization leading China's AI hardware development effort. That's <strong>Alibaba Group</strong>(BABA -8.60%). Yes, China's e-commerce giant. Even more surprising is that the e-commerce powerhouse is doing a great job on the AI development front. It may even become the Nvidia of China.</p>
<div class="lb-trans"><p>令人惊讶的是，领导中国 AI 硬件开发工作的组织是 <strong>Alibaba Group</strong>（BABA -8.60%）。没错，就是中国的电子商务巨头。更令人惊讶的是，这家电子商务巨头在 AI 开发方面表现出色，甚至可能成为中国的 Nvidia。</p>
</div><p>And if that's what's in the cards, investors have much to gain.</p>
<div class="lb-trans"><p>如果这真的是未来的发展方向，投资者将会获得丰厚的回报。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzU5MTkvY29tcHV0ZXItY2lyY3VpdC1ib2FyZC5qcGcmdz03MDA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F835919%2Fcomputer-circuit-board.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>Alibaba is leading China's AI charge</h2>
<div class="lb-trans"><h2>阿里巴巴正在引领中国的 AI 发展</h2>
</div><p>The bulk of Alibaba's revenue still comes from e-commerce. More than half of its Q2 2025 sales came from domestic e-commerce operations, while another 15% was driven by its international e-commerce efforts.</p>
<div class="lb-trans"><p>阿里巴巴的大部分收入仍来自电子商务。2025 年第二季度，其销售额的超过一半来自国内电子商务业务，另有 15% 来自国际电子商务业务。</p>
</div><p>Alibaba is more than e-commerce, though. While still relatively small, the company's cloud computing arm -- which includes its nascent AI business -- accounts for another 15% of its total sales, and it was the second quarter's fastest-growing arm with a year-over-year improvement of 26%. If it continues expanding at its current pace, it could eventually become the company's biggest business.</p>
<div class="lb-trans"><p>不过，阿里巴巴不仅仅是电子商务。虽然相对较小，但公司的云计算部门——包括其新兴的 AI 业务——占其总销售额的 15%，并且是第二季度增长最快的部门，同比增长了 26%。如果继续以当前的速度扩张，它最终可能成为公司的最大业务。</p>
</div><p>This is possible, too, and even likely, given how far along the company now is on this front.</p>
<div class="lb-trans"><p>考虑到公司在这一领域的进展，这种可能性不仅存在，而且甚至很可能。</p>
</div><p>While Alibaba's been working on its T-Head parallel processing unit for a while now, the version of the processor chip unveiled last month is noteworthy. Namely, it almost perfectly matches the performance specs of Nvidia's flagship H20 GPU meant for export to overseas users. In theory, there is no reason for a Chinese firm to choose the H20 over Alibaba's silicon, particularly given that the latter costs about 40% less.</p>
<div class="lb-trans"><p>虽然阿里巴巴已经在开发其 T-Head 并行处理单元一段时间，但上个月发布的处理器芯片版本值得注意。具体来说，它几乎完美匹配 Nvidia 旗舰 H20 GPU 的性能规格，后者是为海外用户出口的。从理论上讲，没有理由让中国公司选择 H20 而不是阿里巴巴的硅，特别是后者的成本低约 40%。</p>
</div><p>It's not just hardware, though. The software Alibaba is typically pairing with its AI solutions is at least as marketable as Nvidia's CUDA ecosystem, if not more so. That's because Alibaba is largely building its AI platforms in a way that makes it easy to use open-source software, which tends to be more flexible than closed-off systems like Nvidia's CUDA.</p>
<div class="lb-trans"><p>不过，这不仅仅是硬件。阿里巴巴通常与其 AI 解决方案配套的软件至少与 Nvidia 的 CUDA 生态系统一样具有市场竞争力，甚至可能更具优势。这是因为阿里巴巴在构建其 AI 平台时，主要是以便于使用开源软件的方式进行，这通常比像 Nvidia 的 CUDA 这样的封闭系统更灵活。</p>
</div><p>The company has also made clear to all potential customers that its cloud solutions and artificial intelligence solutions are meant to be purchased in tandem, with many of its tools being aimed at its existing cloud customers.</p>
<div class="lb-trans"><p>公司还向所有潜在客户明确表示，其云解决方案和人工智能解决方案是一起购买的，许多工具旨在服务于现有的云客户。</p>
</div><p>And given the opportunity that Alibaba has just within its home country and its near neighbors, don't be surprised if the company makes an Nvidia-size splash there.</p>
<div class="lb-trans"><p>鉴于阿里巴巴在其本国及邻近国家所拥有的机会，如果该公司在这些地区取得与 Nvidia 相当的成功，也不要感到惊讶。</p>
</div><h2>Opportunity awaits</h2>
<div class="lb-trans"><h2>机会在等待</h2>
</div><p>A recent analysis done by <strong>JPMorgan</strong> speaks volumes. As its researchers bluntly noted in response to China's recent AI development work, &#34;A sleeping giant awakens.&#34; Quantifying the comment, JPMorgan's analysts believe China's AI industry could be worth $1.4 trillion by 2030, translating into a 52% return on investments made in this tech in the meantime.</p>
<div class="lb-trans"><p><strong>摩根大通</strong> 最近的一项分析说明了一切。正如其研究人员在回应中国最近的 AI 发展工作时直言不讳地指出的那样，“沉睡的巨人觉醒了。” 量化这一评论，摩根大通的分析师认为，中国的 AI 产业到 2030 年可能价值 1.4 万亿美元，这意味着在此期间对该技术的投资回报率可达 52%。</p>
</div><p>This technology will, of course, be sold as a service outside China, but it will also be used within its borders to improve the nation's own business efficiency. <strong>Goldman Sachs</strong> expects that whatever gross domestic product (GDP) growth rate the nation was going to experience through 2030 will now be improved by 20 to 30 basis points thanks to China's growing adoption of AI.</p>
<div class="lb-trans"><p>当然，这项技术将在中国以外作为服务出售，但它也将在国内用于提高国家自身的商业效率。<strong>高盛</strong> 预计，国家到 2030 年将经历的任何国内生产总值（GDP）增长率，现在将因中国对 AI 的日益采用而提高 20 到 30 个基点。</p>
</div><p>Alibaba's processors won't be responsible for <em>all</em> of this expansion. <strong>Baidu</strong>(BIDU -8.09%) is working on its own AI processors, as is Huawei. Baidu's Kunlun processor was technically China's first domestically made AI silicon, in fact.</p>
<div class="lb-trans"><p>阿里巴巴的处理器并不会对这一扩张负全部责任。<strong>百度</strong>（BIDU -8.09%）和华为也在开发自己的 AI 处理器。事实上，百度的昆仑处理器在技术上是中国首个国产 AI 硅。</p>
</div><p>It's not quite a direct threat to Alibaba's eventual dominance, though. Baidu's Kunlun is a cloud-to-edge AI chip largely meant for the mobile market. It doesn't have the same sort of raw computing power of Alibaba's T-Head, which is expected to be able to handle heavier-duty inference work like figuring out how respond to a new query when there's no specific data or clear history to draw on. China's state-owned telecom outfit, China Unicom, already ordered a bunch of Alibaba's T-Head processors for a new data center in Qinghai, in fact, even before these chips officially entered mass production.</p>
<div class="lb-trans"><p>不过，这并不构成对阿里巴巴未来主导地位的直接威胁。百度的昆仑是一款主要面向移动市场的云到边缘 AI 芯片。它没有阿里巴巴的 T-Head 那种原始计算能力，后者预计能够处理更重的推理工作，比如在没有特定数据或明确历史可供参考时，如何回应新的查询。实际上，中国国有电信公司中国联通在这些芯片正式进入大规模生产之前，已经为青海的新数据中心订购了一批阿里巴巴的 T-Head 处理器。</p>
</div><p>And this may be just a taste of the demand that awaits. Again, JPMorgan expects China's AI market to grow to $1.4 trillion by 2030, while the country's total GDP is a little less than $20 trillion right now.</p>
<div class="lb-trans"><p>这可能只是未来需求的一个缩影。再次强调，摩根大通预计中国的 AI 市场到 2030 年将增长到 1.4 万亿美元，而该国目前的总 GDP 略低于 20 万亿美元。</p>
</div><h2>It's no Nvidia, but it's still a buy, or will be soon enough</h2>
<div class="lb-trans"><h2>虽然不是 Nvidia，但仍然值得购买，或者很快就会值得购买</h2>
</div><p>Will Alibaba replicate all the success that Nvidia has dished out over the course of the past three years, after the launch of ChatGPT set off an AI arms race? For that matter, will Alibaba stock reproduce the 1,400% gain that Nvidia's stock has made during this stretch?</p>
<div class="lb-trans"><p>阿里巴巴会否复制 Nvidia 在过去三年中所取得的所有成功，尤其是在 ChatGPT 引发 AI 军备竞赛之后？更进一步，阿里巴巴的股票会否在这一期间实现与 Nvidia 股票相同的 1400% 的涨幅？</p>
</div><p>The answer to both questions is probably not.</p>
<div class="lb-trans"><p>这两个问题的答案可能都是否定的。</p>
</div><p>Don't lose perspective on the comparison, though. Nvidia was in the right place at the perfect time with the right product, further assisted by the fact that nobody saw such a phenomenon taking shape coming out of the COVID-19 pandemic. There was more than a little bit of luck involved. Even if Alibaba does only half as well from here as Nvidia has, it will still be a fantastic investment.</p>
<div class="lb-trans"><p>不过，不要失去对比较的视角。英伟达在正确的时间、正确的地点推出了合适的产品，进一步得益于没有人预见到在 COVID-19 大流行后会出现这样的现象。这其中有不少运气成分。即使阿里巴巴从现在起的表现只有英伟达的一半，它仍然会是一个极好的投资。</p>
</div><p>And yes, it certainly looks like Alibaba could become the Nvidia of China, even if its net opportunity is measurably smaller. Analysts believe the company's top line is not only going to grow all the way through 2027, but also accelerate the whole time.</p>
<div class="lb-trans"><p>是的，阿里巴巴确实看起来有可能成为中国的英伟达，即使它的净机会明显较小。分析师认为，该公司的收入不仅将在 2027 年之前持续增长，而且整个过程中还会加速。</p>
</div><p>Just don't tarry too long if you're interested. Although Alibaba stock is a bit overheated right now and arguably due for a dip, it's unlikely that that dip is going to drag this ticker back down to its most recent major low. Take whatever decent-size discount you can get.</p>
<div class="lb-trans"><p>如果你感兴趣，就不要拖延太久。尽管阿里巴巴的股票目前有些过热，且可以说是时候调整了，但这种调整不太可能将该股票拖回到最近的主要低点。尽量抓住任何合理的折扣。</p>
</div>

Alibaba's Artificial Intelligence (AI) Push: Could This Be China's Best Answer to Nvidia?

- Nvidia 在全球 AI 硬件市场占据主导地位，竞争者如阿里巴巴正在崛起。  
- 阿里巴巴的 T-Head 处理器性能接近 Nvidia 的 H20，且成本低 40%。  
- JPMorgan 预计中国 AI 市场到 2030 年将达 1.4 万亿美元，推动 GDP 增长。  

阿里巴巴在人工智能领域的推进：这会是中国对英伟达的最佳回应吗？