<p>當 <strong>英偉達</strong> 首席執行官黃仁勳談論人工智能（AI）的未來時，投資者都會傾聽。黃仁勳最近預測，全球在 AI 基礎設施上的支出將從今年的 6000 億美元激增至到本世紀末的 3 萬億到 4 萬億美元。這是一個巨大的增長，且將需要大量先進的芯片。</p>
<p>黃仁勳最近還特別讚揚了 <strong>台積電</strong>（TSMC -1.59%），稱其為 “人類歷史上最偉大的公司之一”，並表示 “任何想要購買台積電股票的人都是非常聰明的人。” 作為全球最有價值公司的負責人，這種認可並不是每天都能聽到的。</p>
<p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzI5NzIvZ2V0dHlpbWFnZXMtMTgwMjI4NTEwOS5qcGcmdz03MDA?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" alt="" width="700" height="420" original-src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzI5NzIvZ2V0dHlpbWFnZXMtMTgwMjI4NTEwOS5qcGcmdz03MDA"/></p>
<p>圖片來源：Getty Images。</p>
<h2>台積電的特別之處</h2>
<p>製造半導體並不容易。這需要技術專長，製造設施（fabs）需要接近滿負荷運轉才能盈利。製造這些複雜芯片的過程可能涉及 700 多個步驟，每個步驟都有可能產生過量廢料。這種複雜性意味着大規模生產這些芯片需要真正的專業知識。這就是為什麼如今大多數芯片製造商不自己生產芯片，而是將設計發送給合同製造商。</p>
<p>這就是台積電的用武之地。它是全球領先的半導體合同製造商，先進的芯片設計師在需要大規模生產其設計時都會選擇它。它運營着全球最先進的芯片製造設施，稱為晶圓廠，並在生產先進芯片（如圖形處理單元（GPU））方面處於明顯領先地位，採用最小的工藝節點。更小的工藝節點在每個芯片上集成更多的晶體管，使其更快、更高效。這並不容易做到，像 <strong>英特爾</strong> 和 <strong>三星</strong> 這樣的競爭對手在追趕台積電的良率方面遇到了困難。</p>
<p>這就是為什麼每個主要的 AI 芯片製造商，從英偉達到 <strong>博通</strong> 再到 <strong>AMD</strong>，都依賴於台積電。無論誰在 AI 芯片設計競賽中脱穎而出，台積電都能獲勝，因為他們都需要其製造專業知識。</p>
<p>台積電的優勢不僅僅在於其規模。它在縮小工藝節點和快速轉向高產量方面始終領先於行業。現在近四分之三的收入來自於 7 納米或更小節點製造的芯片，其中 3 納米芯片已經佔據了近四分之一。該公司已經在準備推出 2 納米生產，保持其在競爭對手前面的領先地位，並鎖定客户多年。</p>
<p>這種技術領先使台積電擁有強大的定價能力。它能夠穩步提高價格，同時仍然讓客户滿意，這就是為什麼儘管在新制造設施上投入巨大，其毛利率在過去幾年中仍然上升。</p>
<p>AI 的機會巨大，台積電正處於捕捉這一機會的完美位置。該公司表示，AI 芯片的需求預計將在 2028 年前以超過 40% 的複合年增長率（CAGR）增長。這是少數公司能夠匹配的增長空間，併為台積電的擴張提供了多年的可見性。它已經在與其最大客户緊密合作，以確保產能並滿足他們日益增長的需求。</p>
<p>AI 並不是唯一的主要增長驅動力。如果機器人出租車真的在本世紀末開始在城市街道上行駛，它們每輛都需要多個強大的芯片來安全處理實時駕駛數據。這可能成為台積電的另一個巨大市場。新興技術如機器人技術和量子計算也同樣需要尖端芯片。無論計算能力走向何方，台積電很可能都會處於其中。</p>
<h2>台積電股票值得購買嗎？</h2>
<p>儘管在半導體生態系統中扮演着關鍵角色，台積電仍然有些低調。它目前也是半導體領域中最具價值的公司之一，基於分析師對 2026 年的預估，其前瞻市盈率（P/E）為 23 倍，儘管面前仍有巨大的增長空間。</p>
<p>尋找參與 AI 基礎設施建設的投資者應該認真考慮黃仁勳的建議。台積電是全球最有利於從即將到來的 AI 支出繁榮中受益的公司之一，可能是未來十年最聰明的買入和持有的股票。</p>

台積電

半導體 3 倍做多 - Direxion

<p>Nvidia 首席執行官黃仁勳預測，到 2030 年，全球人工智能基礎設施支出將從 6000 億美元上升至 3-4 萬億美元，強調了對先進芯片的需求。他稱讚台積電為領先公司，表示投資台積電股票是明智的選擇。台積電在半導體制造方面表現出色，擁有強大的技術領先優勢和定價能力，預計到 2028 年，人工智能芯片需求將以 40% 的年複合增長率增長。台積電在人工智能和其他新興技術領域具備良好的增長前景，使其成為一個引人注目的投資機會</p>

<p>When <strong>Nvidia</strong> CEO Jensen Huang talks about the future of artificial intelligence (AI), investors listen. Huang recently predicted that global spending on AI infrastructure will surge from $600 billion this year to as much as $3 trillion to $4 trillion by the end of the decade. That is a massive increase, and it is going to require a lot of advanced chips.</p>
<div class="lb-trans"><p>當 <strong>英偉達</strong> 首席執行官黃仁勳談論人工智能（AI）的未來時，投資者都會傾聽。黃仁勳最近預測，全球在 AI 基礎設施上的支出將從今年的 6000 億美元激增至到本世紀末的 3 萬億到 4 萬億美元。這是一個巨大的增長，且將需要大量先進的芯片。</p>
</div><p>Huang also recently went out of his way to praise <strong>Taiwan Semiconductor Manufacturing</strong>(TSM -1.59%), calling it &#34;one of the greatest companies in the history of humanity&#34; and saying that &#34;anybody who wants to buy TSMC stock is a very smart person.&#34; Coming from the head of the world's most valuable company, that is not the kind of endorsement you hear every day.</p>
<div class="lb-trans"><p>黃仁勳最近還特別讚揚了 <strong>台積電</strong>（TSMC -1.59%），稱其為 “人類歷史上最偉大的公司之一”，並表示 “任何想要購買台積電股票的人都是非常聰明的人。” 作為全球最有價值公司的負責人，這種認可並不是每天都能聽到的。</p>
</div><p><img src="https://imageproxy.pbkrs.com/https://g.foolcdn.com/image//query-b3A9cmVzaXplJnVybD1odHRwczovL2cuZm9vbGNkbi5jb20vZWRpdG9yaWFsL2ltYWdlcy84MzI5NzIvZ2V0dHlpbWFnZXMtMTgwMjI4NTEwOS5qcGcmdz03MDA" alt="" original-src="https://g.foolcdn.com/image/?url=https%3A%2F%2Fg.foolcdn.com%2Feditorial%2Fimages%2F832972%2Fgettyimages-1802285109.jpg&amp;op=resize&amp;w=700"/></p>
<p>Image source: Getty Images.</p>
<div class="lb-trans"><p>圖片來源：Getty Images。</p>
</div><h2>What makes TSMC special</h2>
<div class="lb-trans"><h2>台積電的特別之處</h2>
</div><p>Manufacturing semiconductors isn't easy. It requires technological expertise, and fabs (manufacturing facilities) need to be running near full capacity to be profitable. The process for making these complex chips can involve 700-plus steps, with the potential for excess waste overshadowing each step. Such complexity means producing these chips at scale requires real expertise. That is why today most chipmakers don't make their own chips, but instead send their designs to contract manufacturers.</p>
<div class="lb-trans"><p>製造半導體並不容易。這需要技術專長，製造設施（fabs）需要接近滿負荷運轉才能盈利。製造這些複雜芯片的過程可能涉及 700 多個步驟，每個步驟都有可能產生過量廢料。這種複雜性意味着大規模生產這些芯片需要真正的專業知識。這就是為什麼如今大多數芯片製造商不自己生產芯片，而是將設計發送給合同製造商。</p>
</div><p>That is where TSMC comes in. It is the world's leading semiconductor contract manufacturer and the company that advanced chip designers turn to when they need their designs manufactured at scale. It runs the most advanced chipmaking facilities on the planet, known as foundries, and is the clear leader in producing advanced chips, like graphics processing units (GPUs), with the smallest nodes. Smaller nodes pack more transistors on each chip, which makes them faster and more efficient. This is not easy to do, and rivals like <strong>Intel</strong> and <strong>Samsung</strong> have struggled to match TSMC's yields.</p>
<div class="lb-trans"><p>這就是台積電的用武之地。它是全球領先的半導體合同製造商，先進的芯片設計師在需要大規模生產其設計時都會選擇它。它運營着全球最先進的芯片製造設施，稱為晶圓廠，並在生產先進芯片（如圖形處理單元（GPU））方面處於明顯領先地位，採用最小的工藝節點。更小的工藝節點在每個芯片上集成更多的晶體管，使其更快、更高效。這並不容易做到，像 <strong>英特爾</strong> 和 <strong>三星</strong> 這樣的競爭對手在追趕台積電的良率方面遇到了困難。</p>
</div><p>That is why every major AI chipmaker, from Nvidia to <strong>Broadcom</strong> to <strong>AMD</strong>, relies on TSMC. It wins no matter who comes out on top in the AI chip design race, because they all need its manufacturing expertise.</p>
<div class="lb-trans"><p>這就是為什麼每個主要的 AI 芯片製造商，從英偉達到 <strong>博通</strong> 再到 <strong>AMD</strong>，都依賴於台積電。無論誰在 AI 芯片設計競賽中脱穎而出，台積電都能獲勝，因為他們都需要其製造專業知識。</p>
</div><p>TSMC's advantage is more than just its scale. It has consistently led the industry in shrinking node sizes and moving production into high volume quickly. Nearly three-quarters of its revenue now comes from chips built on nodes of 7 nanometers or smaller, with 3nm chips already making up nearly a quarter. The company is already preparing to roll out 2nm production, keeping it ahead of competitors and locking in its customers for years to come.</p>
<div class="lb-trans"><p>台積電的優勢不僅僅在於其規模。它在縮小工藝節點和快速轉向高產量方面始終領先於行業。現在近四分之三的收入來自於 7 納米或更小節點製造的芯片，其中 3 納米芯片已經佔據了近四分之一。該公司已經在準備推出 2 納米生產，保持其在競爭對手前面的領先地位，並鎖定客户多年。</p>
</div><p>That technological lead has given TSMC strong pricing power. It has been able to steadily raise prices while still keeping its customers happy, which is why its gross margin has climbed over the last several years even as it spends heavily on new fabs.</p>
<div class="lb-trans"><p>這種技術領先使台積電擁有強大的定價能力。它能夠穩步提高價格，同時仍然讓客户滿意，這就是為什麼儘管在新制造設施上投入巨大，其毛利率在過去幾年中仍然上升。</p>
</div><p>The AI opportunity is massive, and TSMC is in the perfect spot to capture it. The company has said that AI chip demand should grow at more than a 40% compound annual growth rate (CAGR) through 2028. That is the kind of growth runway that few companies can match, and it gives TSMC years of visibility into its own expansion. It is already working closely with its largest customers to secure capacity and meet their growing demand.</p>
<div class="lb-trans"><p>AI 的機會巨大，台積電正處於捕捉這一機會的完美位置。該公司表示，AI 芯片的需求預計將在 2028 年前以超過 40% 的複合年增長率（CAGR）增長。這是少數公司能夠匹配的增長空間，併為台積電的擴張提供了多年的可見性。它已經在與其最大客户緊密合作，以確保產能並滿足他們日益增長的需求。</p>
</div><p>AI is not the only major growth driver either. If robotaxis really do start lining city streets later this decade, they will each need multiple powerful chips to process real-time driving data safely. That could become another huge market for TSMC. The same goes for emerging technologies like robotics and quantum computing, which will require cutting-edge chips. Wherever computing power is headed, TSMC is likely to be right in the middle of it.</p>
<div class="lb-trans"><p>AI 並不是唯一的主要增長驅動力。如果機器人出租車真的在本世紀末開始在城市街道上行駛，它們每輛都需要多個強大的芯片來安全處理實時駕駛數據。這可能成為台積電的另一個巨大市場。新興技術如機器人技術和量子計算也同樣需要尖端芯片。無論計算能力走向何方，台積電很可能都會處於其中。</p>
</div><h2>Is TSMC stock a buy?</h2>
<div class="lb-trans"><h2>台積電股票值得購買嗎？</h2>
</div><p>Despite its critical role in the semiconductor ecosystem, TSMC still flies a bit under the radar. It is also currently one of the best values in the semiconductor space, trading at a forward price-to-earnings (P/E) ratio of 23 times based on analysts' 2026 estimates, despite the huge growth still in front of it.</p>
<div class="lb-trans"><p>儘管在半導體生態系統中扮演着關鍵角色，台積電仍然有些低調。它目前也是半導體領域中最具價值的公司之一，基於分析師對 2026 年的預估，其前瞻市盈率（P/E）為 23 倍，儘管面前仍有巨大的增長空間。</p>
</div><p>Investors looking for a way to play the AI infrastructure buildout should take Huang's advice seriously. TSMC is one of the best-positioned companies in the world to benefit from the coming AI spending boom, and it might just be the smartest stock to buy and hold through the next decade.</p>
<div class="lb-trans"><p>尋找參與 AI 基礎設施建設的投資者應該認真考慮黃仁勳的建議。台積電是全球最有利於從即將到來的 AI 支出繁榮中受益的公司之一，可能是未來十年最聰明的買入和持有的股票。</p>
</div>

Nvidia's CEO Just Delivered Incredible News for Taiwan Semiconductor Manufacturing Stock Investors

- 黃仁勳預測全球 AI 基礎設施支出將從 6000 億美元增至 3 至 4 萬億美元。  
- TSMC 是全球領先的半導體代工廠，主要客户包括 Nvidia 和 AMD。  
- TSMC 在 AI 芯片需求增長中處於有利位置，預計年複合增長率超過 40%。

英偉達的 CEO 剛剛為台積電的股票投資者傳遞了振奮人心的消息