<p>人工智能（AI）计算领域正在发生转变。超大规模云计算公司和人工智能公司正从训练大型语言模型转向在真实世界数据上应用这些模型，以释放这项技术所承诺的生产力提升。</p>
<p>大量资金已被投入到优化用于训练大型数据集模型的数据中心基础设施中。另一方面，推理并不需要像训练阶段那样多的并行处理能力，因为它需要对特定输入做出实时响应。然而，德勤估计，推理工作负载的增加将会对更多计算能力产生需求。</p>
<p>这家咨询巨头估计，到 2026 年，推理工作负载将消耗所有人工智能计算能力的三分之二，较去年的 50% 有所上升。这将创造对更适合推理的人工智能芯片的需求。这对 <strong>Intel</strong>（INTC +4.36%）投资者来说是个好消息。事实上，看到 Intel 在人工智能推理时代超越 <strong>Nvidia</strong> 和 <strong>Broadcom</strong>（AVGO 0.52%）并不令人惊讶。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-dXJsPWh0dHBzOi8vZy5mb29sY2RuLmNvbS9lZGl0b3JpYWwvaW1hZ2VzLzg2ODY4OS9pbnNpZGUtaW50ZWwtYnVpbGRpbmctd2l0aC1pbnRlbC1sb2dvLWluLXZpZXdfaW50ZWwuanBnAjZ3PTM4NDACNm9wPXJlc2l6ZQ?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="Intel logo inside a building." original-src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-dXJsPWh0dHBzOi8vZy5mb29sY2RuLmNvbS9lZGl0b3JpYWwvaW1hZ2VzLzg2ODY4OS9pbnNpZGUtaW50ZWwtYnVpbGRpbmctd2l0aC1pbnRlbC1sb2dvLWluLXZpZXdfaW50ZWwuanBnAjZ3PTM4NDACNm9wPXJlc2l6ZQ"/></p>
<p>图片来源：Intel</p>
<h2>人工智能推理需求为 Intel 注入强劲动力</h2>
<p>人工智能推理的不同计算需求解释了专门为这些任务设计的芯片的强劲需求。例如，应用特定集成电路（ASIC）因其相较于通用并行处理器（如图形处理单元）的优势，受到超大规模云计算公司和人工智能公司的健康需求。</p>
<p>展开</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-dXJsPWh0dHBzOi8vZy5mb29sY2RuLmNvbS9hcnQvY29tcGFueWxvZ29zL21hcmsvSU5UQy5wbmcCNnc9MTI4AjZvcD1yZXNpemU?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="Intel Stock Quote" original-src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-dXJsPWh0dHBzOi8vZy5mb29sY2RuLmNvbS9hcnQvY29tcGFueWxvZ29zL21hcmsvSU5UQy5wbmcCNnc9MTI4AjZvcD1yZXNpemU"/></p>
<h2>NASDAQ: INTC</h2>
<p>Intel</p>
<p>今日变动</p>
<p>(4.36%) $4.71</p>
<p>当前价格</p>
<p>$112.86</p>
<h3>关键数据点</h3>
<p>市值</p>
<p>$5680 亿</p>
<p>日内区间</p>
<p>$106.59 - $113.50</p>
<p>52 周区间</p>
<p>$18.96 - $113.50</p>
<p>成交量</p>
<p>570 万</p>
<p>平均成交量</p>
<p>1.07 亿</p>
<p>毛利率</p>
<p>35.90%</p>
<p>ASIC 的能效更高，当它们处理特定任务时，有助于降低整体计算成本。这解释了为什么在 ASIC 市场中占据 60% 市场份额的 Broadcom 预计其人工智能收入将大幅增长。</p>
<p>该公司有信心到 2027 年其人工智能芯片收入将超过 1000 亿美元，较 去年的 200 亿美元 大幅跃升。即使是 Nvidia 也在为推理任务准备即将推出的 Vera Rubin 处理器，承诺 显著降低推理成本。然而，我相信投资 Intel 可能是抓住人工智能推理芯片日益增长需求的最佳方式之一。</p>
<p>这是因为该公司在服务器中央处理器（CPU）市场占据主导地位，市场份额超过 71%。此外，Intel 在 ASIC 市场也在不断发展。其 ASIC 收入在第一季度同比几乎翻了一番，而环比增长为 30%。</p>
<p>相比之下，Intel 的定制 ASIC 收入在 2025 年第四季度同比增长 50%，而环比增长为 26%。显然，Intel 最近在 ASIC 销售方面经历了显著加速，该业务现在的年化收入超过 10 亿美元。看到这一业务进一步加速并不令人惊讶，因为超大规模云计算公司和芯片公司最近与 Intel 的定制人工智能芯片和服务器 CPU 签署了多份合同。</p>
<p>例如，<strong>Alphabet</strong> 的 Google 与 Intel 签订了多年合同，以部署 Intel 的 ASIC 和 Xeon 服务器 CPU。即使是 Nvidia 也在其 Rubin 机架规模服务器系统中集成了 Intel 的 Xeon 服务器 CPU。总的来说，人工智能推理似乎预示着 Intel 一个新的增长时代的到来。</p>
<p>首席执行官 Lip-Bu Tan 在最新的 财报电话会议 上的讲话清楚地表明了公司在推理时代进一步增长的良好前景。他指出：</p>
<p>�3&gt; 人工智能现在正向现实世界转变，朝着更分布式的推理和强化学习工作负载发展，比如代理性、物理 AI、机器人和边缘 AI。
�3&gt;
�3&gt; 这种转变现在开始在我们的业绩中显现，我想花一些时间来讨论这个。过去几年，高性能计算的故事几乎完全围绕 GPU 和其他加速器。在最近几个月，我们看到 CPU 重新插入作为人工智能时代不可或缺的基础的明确迹象。</p>
<p>市场研究公司 Futurum 估计，人工智能数据中心的 CPU 市场到 2028 年可能以 28% 的年化率增长，尽管由于人工智能推理需求的激增，其销售以更快的速度增长也并不令人惊讶。</p>
<h2>Chipzilla 的增长前景改善指向光明的未来</h2>
<p>Intel去年开始扭转业务，这些努力在 2026 年开始结出成果。尽管该公司的收入在 2025 年保持在 530 亿美元几乎持平，分析师预计未来三年将实现持续的两位数百分比增长。</p>
<p></p>
<p>INTC 当前财政年度收入估计 数据来源于 YCharts.</p>
<p>然而，英特尔有可能超越这些预期，这得益于公司专注于提升其制造能力、提高先进工艺节点的生产良率，以及在人工智能芯片领域达成的 丰厚合作伙伴关系。强于预期的增长可能会导致投资者给予 英特尔股票 溢价估值，而目前其交易的销售倍数为 8.7 倍。</p>
<p>这个销售倍数略高于美国科技行业平均的 7 倍，但并不算太贵。假设英特尔的销售倍数为 10 倍，并在 2028 年实现 710 亿美元的收入，其市值将达到 7100 亿美元。这将比目前水平上涨 48%。然而，考虑到对其芯片的需求超过供应，不要惊讶于这只 人工智能股票 带来更大的收益，这表明它的增长速度可能确实快于分析师的预期。</p>

Anthropic发布AI代理新能力，GPT-5.5 Instant两天内全量推出，AI主线再次共振爆发，谁是这波行情的最大赢家？

<p style="margin-left:0;">Anthropic发布AI代理新能力，GPT-5.5 Instant两天内全量推出，AI主线再次共振爆发，谁是这波行情的最大赢家？</p>

AI主线风云再起！Anthropic+OpenAI军备赛加速

04335

英特尔

英伟达

<p>英特尔有望成为人工智能推理市场的重要参与者，因为对专用芯片的需求不断增长。德勤预测，到 2026 年，推理工作负载将消耗三分之二的人工智能计算能力。英特尔在服务器 CPU 市场的份额达到 71%，其 ASIC 收入也显著增长，同比几乎翻倍。与谷歌和英伟达等公司的合作进一步为英特尔在这一不断发展的市场中奠定了成功的基础。分析师预计，英特尔的收入将在未来三年内持续增长，这得益于制造能力的提升和战略合作伙伴关系的建立</p>

<p>A shift is happening in the artificial intelligence (AI) computing space. Hyperscalers and AI companies are transitioning from training large language models to putting those models to work on real-world data to unlock the productivity gains that this technology promises.</p>
<div class="lb-trans"><p>人工智能（AI）计算领域正在发生转变。超大规模云计算公司和人工智能公司正从训练大型语言模型转向在真实世界数据上应用这些模型，以释放这项技术所承诺的生产力提升。</p>
</div><p>A lot of money has been spent on data center infrastructure that's optimized to train models using large datasets. Inference, on the other hand, doesn't require as much parallel-processing power as the training phase, as it needs to act on specific inputs to respond to queries in real time. However, Deloitte estimates that rising inference workloads will create demand for more computing power.</p>
<div class="lb-trans"><p>大量资金已被投入到优化用于训练大型数据集模型的数据中心基础设施中。另一方面，推理并不需要像训练阶段那样多的并行处理能力，因为它需要对特定输入做出实时响应。然而，德勤估计，推理工作负载的增加将会对更多计算能力产生需求。</p>
</div><p>The consulting giant estimates that inference workloads will consume two-thirds of all AI computing power in 2026, up from 50% last year. This will create a need for more AI chips of the varieties that are best for inference. That's great news for <strong>Intel</strong>( INTC +4.36%) investors. In fact, it won't be surprising to see Intel eclipsing the likes of <strong>Nvidia</strong> and <strong>Broadcom</strong>( AVGO 0.52%) in the AI inference era.</p>
<div class="lb-trans"><p>这家咨询巨头估计，到 2026 年，推理工作负载将消耗所有人工智能计算能力的三分之二，较去年的 50% 有所上升。这将创造对更适合推理的人工智能芯片的需求。这对 <strong>Intel</strong>（INTC +4.36%）投资者来说是个好消息。事实上，看到 Intel 在人工智能推理时代超越 <strong>Nvidia</strong> 和 <strong>Broadcom</strong>（AVGO 0.52%）并不令人惊讶。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84Njg2ODkvaW5zaWRlLWludGVsLWJ1aWxkaW5nLXdpdGgtaW50ZWwtbG9nby1pbi12aWV3X2ludGVsLmpwZyZ3PTM4NDA" alt="Intel logo inside a building." original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F868689%2Finside-intel-building-with-intel-logo-in-view_intel.jpg&amp;w=3840&amp;op=resize"/></p>
<p>Image source: Intel</p>
<div class="lb-trans"><p>图片来源：Intel</p>
</div><h2>AI inference demand has given Intel a big shot in the arm</h2>
<div class="lb-trans"><h2>人工智能推理需求为 Intel 注入强劲动力</h2>
</div><p>The different computing demands of AI inference explain the strong demand for chips designed specifically for these tasks. Application-specific integrated circuits (ASICs), for instance, have been in healthy demand from hyperscalers and AI companies because of the advantages they can offer over general-purpose parallel processors such as graphics processing units.</p>
<div class="lb-trans"><p>人工智能推理的不同计算需求解释了专门为这些任务设计的芯片的强劲需求。例如，应用特定集成电路（ASIC）因其相较于通用并行处理器（如图形处理单元）的优势，受到超大规模云计算公司和人工智能公司的健康需求。</p>
</div><p>Expand</p>
<div class="lb-trans"><p>展开</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vYXJ0L2NvbXBhbnlsb2dvcy9tYXJrL0lOVEMucG5nJnc9MTI4" alt="Intel Stock Quote" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Fart%2Fcompanylogos%2Fmark%2FINTC.png&amp;w=128&amp;op=resize"/></p>
<h2>NASDAQ: INTC</h2>
<div class="lb-trans"><h2>NASDAQ: INTC</h2>
</div><p>Intel</p>
<div class="lb-trans"><p>Intel</p>
</div><p>Today's Change</p>
<div class="lb-trans"><p>今日变动</p>
</div><p>(4.36%) $4.71</p>
<div class="lb-trans"><p>(4.36%) $4.71</p>
</div><p>Current Price</p>
<div class="lb-trans"><p>当前价格</p>
</div><p>$112.86</p>
<div class="lb-trans"><p>$112.86</p>
</div><h3>Key Data Points</h3>
<div class="lb-trans"><h3>关键数据点</h3>
</div><p>Market Cap</p>
<div class="lb-trans"><p>市值</p>
</div><p>$568B</p>
<div class="lb-trans"><p>$5680 亿</p>
</div><p>Day's Range</p>
<div class="lb-trans"><p>日内区间</p>
</div><p>$106.59 - $113.50</p>
<div class="lb-trans"><p>$106.59 - $113.50</p>
</div><p>52wk Range</p>
<div class="lb-trans"><p>52 周区间</p>
</div><p>$18.96 - $113.50</p>
<div class="lb-trans"><p>$18.96 - $113.50</p>
</div><p>Volume</p>
<div class="lb-trans"><p>成交量</p>
</div><p>5.7M</p>
<div class="lb-trans"><p>570 万</p>
</div><p>Avg Vol</p>
<div class="lb-trans"><p>平均成交量</p>
</div><p>107M</p>
<div class="lb-trans"><p>1.07 亿</p>
</div><p>Gross Margin</p>
<div class="lb-trans"><p>毛利率</p>
</div><p>35.90%</p>
<div class="lb-trans"><p>35.90%</p>
</div><p>ASICs are more power efficient, and they help reduce overall computational costs when they are handling the highly specific tasks that they are designed to. This explains why Broadcom, the biggest player in the ASIC market (with an estimated 60% market share) is forecasting a huge increase in its AI revenue.</p>
<div class="lb-trans"><p>ASIC 的能效更高，当它们处理特定任务时，有助于降低整体计算成本。这解释了为什么在 ASIC 市场中占据 60% 市场份额的 Broadcom 预计其人工智能收入将大幅增长。</p>
</div><p>The company is confident it will exceed $100 billion in AI chip revenue in 2027, a huge jump from last year's $20 billion. Even Nvidia is gearing up its upcoming Vera Rubin processors for inference-focused tasks, promising significant reductions in inference costs. However, I believe that investing in Intel could be one of the best ways to capitalize on the growing demand for AI inference chips.</p>
<div class="lb-trans"><p>该公司有信心到 2027 年其人工智能芯片收入将超过 1000 亿美元，较 去年的 200 亿美元 大幅跃升。即使是 Nvidia 也在为推理任务准备即将推出的 Vera Rubin 处理器，承诺 显著降低推理成本。然而，我相信投资 Intel 可能是抓住人工智能推理芯片日益增长需求的最佳方式之一。</p>
</div><p>That's because the company is the dominant player in the server central processing unit (CPU) market, with a share of just over 71%. Moreover, Intel is cutting its teeth in the ASIC market as well. Its ASIC revenuealmost doubled year over year in the first quarter, while sequential growth was 30%.</p>
<div class="lb-trans"><p>这是因为该公司在服务器中央处理器（CPU）市场占据主导地位，市场份额超过 71%。此外，Intel 在 ASIC 市场也在不断发展。其 ASIC 收入在第一季度同比几乎翻了一番，而环比增长为 30%。</p>
</div><p>For comparison, Intel's custom ASIC revenue was up by 50% year over year in the fourth quarter of 2025, while sequential growth was 26%. Clearly, Intel has seen a significant acceleration in its ASIC sales lately, and the business now has an annual run rate of more than $1 billion. It won't be surprising to see further acceleration in this business, as hyperscalers and chip companies have lately signed multiple contracts for Intel's custom AI chips and server CPUs.</p>
<div class="lb-trans"><p>相比之下，Intel 的定制 ASIC 收入在 2025 年第四季度同比增长 50%，而环比增长为 26%。显然，Intel 最近在 ASIC 销售方面经历了显著加速，该业务现在的年化收入超过 10 亿美元。看到这一业务进一步加速并不令人惊讶，因为超大规模云计算公司和芯片公司最近与 Intel 的定制人工智能芯片和服务器 CPU 签署了多份合同。</p>
</div><p><strong>Alphabet</strong>'s Google, for instance, has a multiyear contract with Intel to deploy Intel's ASICs and Xeon server CPUs. Even Nvidia is integrating Intel's Xeon server CPUs into its Rubin rack-scale server systems. In all, AI inference seems to be heralding a new era of growth for Intel.</p>
<div class="lb-trans"><p>例如，<strong>Alphabet</strong> 的 Google 与 Intel 签订了多年合同，以部署 Intel 的 ASIC 和 Xeon 服务器 CPU。即使是 Nvidia 也在其 Rubin 机架规模服务器系统中集成了 Intel 的 Xeon 服务器 CPU。总的来说，人工智能推理似乎预示着 Intel 一个新的增长时代的到来。</p>
</div><p>CEO Lip-Bu Tan's remarks on the latest earnings call make it clear why the company is well-placed for further growth in the inference era. He noted:</p>
<div class="lb-trans"><p>首席执行官 Lip-Bu Tan 在最新的 财报电话会议 上的讲话清楚地表明了公司在推理时代进一步增长的良好前景。他指出：</p>
</div><blockquote>
<p>Artificial intelligence is now moving into the real world toward more distributed inference and reinforcement learning workloads, like agentic, physical AI, robots, and edge AI.</p>
<p>This shift is now beginning to show up in our results, and I want to spend some time on this today. For the last few years, the story around high performance computing was almost exclusively about GPUs and other accelerators. In recent months, we have seen clear signs that the CPU is reinserting itself as the indispensable foundation of the AI era.</p>
</blockquote>
<div class="lb-trans"><p>�3&gt; 人工智能现在正向现实世界转变，朝着更分布式的推理和强化学习工作负载发展，比如代理性、物理 AI、机器人和边缘 AI。
�3&gt;
�3&gt; 这种转变现在开始在我们的业绩中显现，我想花一些时间来讨论这个。过去几年，高性能计算的故事几乎完全围绕 GPU 和其他加速器。在最近几个月，我们看到 CPU 重新插入作为人工智能时代不可或缺的基础的明确迹象。</p>
</div><p>Market research company Futurum estimates that the market for CPUs in AI data centers could increase at an annualized rate of 28% through 2028, though it wouldn't be surprising to see their sales increase at a faster pace because of the boom in AI inference demand.</p>
<div class="lb-trans"><p>市场研究公司 Futurum 估计，人工智能数据中心的 CPU 市场到 2028 年可能以 28% 的年化率增长，尽管由于人工智能推理需求的激增，其销售以更快的速度增长也并不令人惊讶。</p>
</div><h2>Chipzilla's improving growth prospects point toward a bright future</h2>
<div class="lb-trans"><h2>Chipzilla 的增长前景改善指向光明的未来</h2>
</div><p>Intel started turning its business around last year, and those efforts are bearing fruit in 2026. While the company's revenue remained almost flat at $53 billion in 2025, analysts anticipate consistent double-digit percentage growth over the next three years.</p>
<div class="lb-trans"><p>Intel去年开始扭转业务，这些努力在 2026 年开始结出成果。尽管该公司的收入在 2025 年保持在 530 亿美元几乎持平，分析师预计未来三年将实现持续的两位数百分比增长。</p>
</div><p></p>
<div class="lb-trans"><p></p>
</div><p>INTC Revenue Estimates for Current Fiscal Year data by YCharts.</p>
<div class="lb-trans"><p>INTC 当前财政年度收入估计 数据来源于 YCharts.</p>
</div><p>However, Intel can surpass those expectations, driven by the company's focus on enhancing its manufacturing capacity, boosting production yields of its advanced process nodes, and the lucrative partnerships it is striking in the AI chip space. Stronger-than-expected growth could lead investors to reward Intel stock with a premium valuation, as compared to the 8.7 times sales it trades at right now.</p>
<div class="lb-trans"><p>然而，英特尔有可能超越这些预期，这得益于公司专注于提升其制造能力、提高先进工艺节点的生产良率，以及在人工智能芯片领域达成的 丰厚合作伙伴关系。强于预期的增长可能会导致投资者给予 英特尔股票 溢价估值，而目前其交易的销售倍数为 8.7 倍。</p>
</div><p>That sales multiple is slightly higher than the U.S. tech sector's average of 7, but it isn't all that expensive. Assuming Intel trades at 10 times sales and achieves $71 billion in revenue in 2028, its market cap would reach $710 billion. That would be an upside of 48% from current levels. Don't be surprised to see this AI stock deliver bigger gains, however, as the demand for its chips is exceeding supply, suggesting that it could indeed grow at a faster pace than analysts are anticipating.</p>
<div class="lb-trans"><p>这个销售倍数略高于美国科技行业平均的 7 倍，但并不算太贵。假设英特尔的销售倍数为 10 倍，并在 2028 年实现 710 亿美元的收入，其市值将达到 7100 亿美元。这将比目前水平上涨 48%。然而，考虑到对其芯片的需求超过供应，不要惊讶于这只 人工智能股票 带来更大的收益，这表明它的增长速度可能确实快于分析师的预期。</p>
</div>

Not Nvidia. Not Broadcom. Intel Is Going to Be the Biggest Winner of the Artificial Intelligence (AI) Inference Era.

- AI 领域正在发生转变，推向实时数据应用和推理计算。  
- 曹煦估计到 2026 年推理工作负载将占 AI 计算能力的三分之二。  
- 英特尔在 ASIC 市场和 CPU 领域的需求上升，预计带动未来增长。

Motley Fool

INTEL-T

博通

南方中证半导体行业精选ETF

国泰中证半导体材料设备主题ETF

动态半导体 ETF - Invesco

费城交易所 半导体 ETF - iShares

2 倍做多 INTC ETF - GraniteShares

2 倍做多 NVDA ETF - T-Rex

华夏国证半导体芯片ETF

国联安半导体ETF

标普半导体 ETF - SPDR

国泰CES半导体芯片行业ETF

半导体 ETF - VanEck Vectors

国联安科创芯片设计ETF

半导体 3 倍做多 - Direxion

华夏上证科创板半导体材料设备主题ETF

谷歌-A

谷歌-C

NVIDIA Corporation

不是 NVIDIA。不是博通。英特尔将成为人工智能推理时代的最大赢家