恩智浦 (NXPI) 2026 Q3 法說會逐字稿

Good day, and thank you for standing by. Welcome to the NXP third quarter 2026 earnings conference call. (Operator Instructions) Please be advised that today's conference is being recorded.

大家好，感謝您稍候。歡迎參加恩智浦（NXP）2026 年第三季財報電話會議。(接線員指示) 請注意，今天的會議將被錄音。

I would now like to turn the conference over to Mike Lucarelli, Senior Vice President, Vice -- excuse me, Senior Vice President, Investor Relations. Please go ahead.

現在我想把會議交給 Mike Lucarelli，資深副總裁，副總裁——抱歉，資深副總裁，投資人關係。請開始。

Thank you, Lisa, and good afternoon, everyone. Welcome to NXP's third quarter earnings call. With me on the call today are Rafael Sotomayor, NXP's President and CEO; and Bill Betz, our CFO. The call today is being recorded and will be available for replay from our corporate website. Today's call will include forward-looking statements that involve risks and uncertainties that could cause NXP's results to differ materially from management's current expectations.

謝謝你，Lisa，各位下午好。歡迎參加恩智浦第三季財報電話會議。今天與我一同在線上的有恩智浦總裁暨執行長 Rafael Sotomayor，以及我們的財務長 Bill Betz。本次電話會議將被錄音，並可於公司官網回放。今天的電話會議將包含前瞻性陳述，涉及風險與不確定性，可能導致恩智浦的結果與管理層目前的預期出現重大差異。

These risks and uncertainties include, but are not limited to, statements regarding the macroeconomic impact on the specific end markets in which we operate, the sale of new and existing products, and our expectations for the financial results for the fourth quarter of 2026.

這些風險與不確定性包括但不限於：有關總體經濟對我們所營運之特定終端市場的影響、新舊產品的銷售，以及我們對 2026 年第四季財務結果的預期等陳述。

NXP undertakes no obligation to revise or update publicly any forward-looking statements. For a full disclosure of forward-looking statements, please refer to our press release. Additionally, we will refer to certain non-GAAP financial measures, which are driven primarily by discrete events that management does not consider to be directly related to NXP's underlying core operating performance.

恩智浦不承擔公開修訂或更新任何前瞻性陳述的義務。如需前瞻性陳述的完整揭露，請參閱我們的新聞稿。此外，我們將提及若干非 GAAP 財務衡量指標，這些指標主要受一次性事件驅動，管理層認為其與恩智浦核心基礎營運表現並無直接關聯。

Pursuant to Regulation G, NXP has provided reconciliations of the non-GAAP financial measures to the most directly comparable GAAP measures in our third quarter 2026 earnings release, which we furnished to the SEC on Form 8-K and is available on NXP's website in the Investor Relations section at nxp.com. I'll now turn the call over to Rafael.

依據 Regulation G，恩智浦已在 2026 年第三季財報新聞稿中提供非 GAAP 財務衡量指標與最直接可比 GAAP 指標之調節表；該新聞稿已以 Form 8-K 提交至美國證券交易委員會（SEC），並可於 nxp.com 投資人關係專區取得。現在我把電話交給 Rafael。

Thank you, Mike, and good afternoon. In Q3, we exceeded our outlook across all end markets and every region grew double digits year-on-year. Growth was led by our company-specific growth drivers in Automotive and Industrial, which grew 30% year-over-year and represent roughly one third of our total revenue as well as continued strength in data center infrastructure. Revenue was $3.79 billion, up 19% year-over-year or up 22%, excluding the MEMS sensor business. Non-GAAP operating margin was 37%, up 330 basis points, and non-GAAP EPS was $4.20, up 35% year-over-year.

謝謝你，Mike，各位下午好。第三季我們在所有終端市場的表現均超越展望，且各地區年增率皆達雙位數成長。成長主要由我們在汽車與工業領域的公司特定成長動能所帶動，兩者年增 30%，合計約占總營收三分之一；同時資料中心基礎設施亦持續強勁。營收為 37.9 億美元，年增 19%；若排除 MEMS 感測器業務，則年增 22%。非 GAAP 營業利益率為 37%，提升 330 個基點；非 GAAP 每股盈餘（EPS）為 4.20 美元，年增 35%。

These results are translating into strong free cash flow generation, enabling us to invest in the business and return cash to shareholders. To that end, our Board has authorized an incremental $5 billion for share repurchases, bringing total authorization to around $6.3 billion. Now turning to end market performance. Automotive revenue crossed $2 billion for the first time, up 17%, excluding sensors.

這些成果正轉化為強勁的自由現金流，使我們得以投資業務並回饋股東。為此，董事會已核准額外 50 億美元的庫藏股回購額度，使總核准額度約達 63 億美元。接著談終端市場表現。汽車營收首次突破 20 億美元；若排除感測器，年增 17%。

This growth is driven by content gains led by our SDV portfolio in a flattish unit production environment. SDV revenue in Q3 is running above $1.5 billion annualized and remains well on track to reach our $2 billion 2027 target.

在整體出貨量持平的生產環境下，這項成長由內容量提升所驅動，並以我們的 SDV 產品組合為主。第三季 SDV 營收按年化計算已超過 15 億美元，且仍穩步朝我們 2027 年 20 億美元的目標前進。

We view the transition from domain-based architectures to zonal and centralized compute as a multiyear upgrade cycle that will enable physical AI and autonomous capabilities in the vehicle. These new architectures create a need for system solutions that drive higher content across NXP's portfolio, including additional processing end nodes, connectivity and power management. This results in 2 to 3 times higher content per vehicle compared to today's architecture.

我們認為，從以功能域（domain）為基礎的架構轉向分區（zonal）與集中式運算，是一個為期多年的升級循環，將使車內的實體 AI 與自動化能力成為可能。這些新架構帶來對系統解決方案的需求，推動恩智浦產品組合中更高的內容量，包括更多處理端點、連線能力與電源管理。相較於現行架構，這將使每輛車的內容量提高 2 到 3 倍。

This evolution is still in its early stages. In 2027, we will begin shipping our next-generation SDV processors, including the S32K5 zonal and S32N central compute platforms. We are starting to see material proof points of the SDV transition. RV Tech, the joint venture of Volkswagen and Rivian, announced yesterday it has standardized its zonal platform on NXP's S32K5 across both companies.

這項演進仍處於早期階段。在 2027 年，我們將開始出貨下一代 SDV 處理器，包括 S32K5 分區平台與 S32N 中央運算平台。我們開始看到 SDV 轉型的實質佐證。Volkswagen 與 Rivian 的合資企業 RV Tech 昨日宣布，已在兩家公司之間將其分區平台標準化採用恩智浦的 S32K5。

This is a clear signal of the leadership of NXP's SDV portfolio and where the industry is heading. Engagements with our new platform continue to expand globally, and we now expect our first S32K5 shipments in China as early as the first quarter of 2027.

這清楚顯示恩智浦 SDV 產品組合的領先地位，以及產業發展方向。我們與新平台的合作案持續在全球擴大，且我們現在預期最早可於 2027 年第一季在中國開始首批 S32K5 出貨。

Software-defined vehicles are one of the first deployments of physical AI and the opportunity reaches well beyond the car. The same building blocks of central compute, sensing, networking, and secure control are moving into robotics and other intelligent systems at the edge. This quarter, we secured our first S32K5 design win in humanoid robotics with a leading US global technology company.

軟體定義車輛是實體 AI 的首批落地應用之一，而機會遠不止於汽車。同樣的中央運算、感測、網路連結與安全控制等基礎模組，正延伸至機器人與其他邊緣端智慧系統。本季，我們與一家領先的美國全球科技公司在類人型機器人領域取得首個 S32K5 設計導入（design win）。

Now moving to Industrial & IoT. Revenue was $802 million, up 39% year-over-year. The Industrial portion of this end market grew nearly 50%, led by data center, factory automation and health care and now represents about 70% of Industrial and IoT. As we mentioned in our Q2 call, AI-enabled processors will represent approximately 15% of Industrial & IoT processor revenue in 2026, more than doubling from last year.

接著談工業與 IoT。營收為 8.02 億美元，年增 39%。該終端市場中的工業部分年增近 50%，主要由資料中心、工廠自動化與醫療保健帶動，且目前約占工業與 IoT 的 70%。如同我們在第二季電話會議中提到的，具 AI 功能的處理器在 2026 年將約占工業與 IoT 處理器營收的 15%，較去年增加一倍以上。

Given market adoption of intelligent systems, we expect growth to continue in 2027. Ara, our Discrete NPU is a key component of our AI portfolio. It continues to gain traction across industrial automation and automotive AI applications, and we expect to be in volume production in 2028.

鑑於智慧系統的市場採用，我們預期 2027 年成長將持續。Ara（我們的離散式 NPU）是我們 AI 產品組合的關鍵元件。它在工業自動化與汽車 AI 應用上持續獲得採用動能，我們預期將於 2028 年進入量產。

Now let me give you an update on our data center business. Given the strength we've seen year-to-date, we now expect our revenue to exceed the $500 million outlook provided earlier this year. Given the new products we're developing, combined with expanding customer engagements, we now expect this business to be over $1 billion in a couple of years. Going to Communication Infrastructure and Other. Revenue was $503 million, up 54% year-over-year, driven by data center and secure cards. And lastly, mobile revenue was $411 million, down 4% year-over-year.

接下來更新我們的資料中心業務。鑑於今年迄今所見的強勁表現，我們現在預期營收將超過今年稍早提供的 5 億美元展望。考量我們正在開發的新產品並結合客戶合作的擴大，我們現在預期該業務在未來幾年內將超過 10 億美元。再談通訊基礎設施與其他。營收為 5.03 億美元，年增 54%，主要由資料中心與安全卡（secure cards）帶動。最後，行動裝置營收為 4.11 億美元，年減 4%。

Turning to our Q4 outlook. We're guiding revenue to $3.85 billion, up 2% sequentially and up 15% year-over-year or 18% excluding sensors. This would represent our fourth consecutive quarter of double-digit annual growth. At the midpoint, we expect the following trends in our business during Q4. Automotive is expected to be up in the low double-digit percent range year-over-year and up low single digit sequentially. Adjusted for the sale of the MEMS sensor business, our guidance implies a high teens percentage growth year-over-year.

接著談我們對第四季的展望。我們預估營收為 38.5 億美元，季增 2%，年增 15%；若排除感測器，則年增 18%。這將是我們連續第四個季度達成年增雙位數成長。以區間中點來看，我們預期第四季各業務趨勢如下：汽車業務預期年增低雙位數百分比，季增低個位數。若就出售 MEMS 感測器業務進行調整，我們的指引意味著年增為高十幾個百分點。

Industrial and IoT is expected to be up in high 20% range year-over-year and up in the low single-digit range sequentially. Mobile is expected to be down in the mid-teens percent range year-over-year and up slightly on a sequential basis. And finally, Communication Infrastructure and Other is expected to be up over 50% range year-over-year and up low single digits sequentially.

工業與 IoT 預期年增約高 20% 區間，季增低個位數區間。行動裝置預期年減約中十幾個百分點，季比則小幅成長。最後，通訊基礎設施與其他預期年增超過 50% 區間，季增低個位數。

And now I'd like to pass the call to Bill for a review of our financial performance.

現在我想把電話交給 Bill，請他回顧我們的財務表現。

Thank you, Rafael, and good afternoon to everyone on today's call. As Rafael has already covered the revenue drivers, I will turn to the financial highlights. Q3 was another strong quarter with record revenue, operating margin and non-GAAP EPS. Revenue grew 19% year-over-year and 8% sequentially to $3.79 billion.

謝謝你，Rafael，也向今天參與電話會議的各位致意、下午好。由於 Rafael 已經說明了營收驅動因素，我將轉而說明財務重點。第三季再度表現強勁，營收、營業利益率與非 GAAP 每股盈餘皆創新高。營收年增 19%、季增 8%，達到 37.9 億美元。

The growth was led by our company-specific growth drivers across software-defined vehicles, electrification, connectivity, radar and industrial edge processing as well as continued momentum in our emerging data center business.

成長主要由我們在軟體定義車輛（SDV）、電氣化、連網、雷達與工業邊緣運算等公司特定成長動能所帶動，並且我們新興的資料中心業務也持續展現動能。

Non-GAAP gross margin was 58.6%, up 160 basis points year-on-year. Our gross margin performance reflects improved factory utilization and solid fall-through on higher revenues. Non-GAAP operating expenses were $810 million or 21.4% of revenue, reflecting strong operating leverage. This led to a non-GAAP operating margin of 37.1%, up 330 basis points year-on-year.

非 GAAP 毛利率為 58.6%，年增 160 個基點。我們的毛利率表現反映工廠稼動率提升，以及在營收增加下良好的利潤落袋（fall-through）。非 GAAP 營業費用為 8.10 億美元，占營收 21.4%，展現強勁的營運槓桿。因此非 GAAP 營業利益率達 37.1%，年增 330 個基點。

Below the line, non-GAAP interest expense was $85 million. Tax rate was steady at 18%. Noncontrolling interest was $15 million, and results from our joint venture equity-accounted investees were a $4 million loss. This resulted in non-GAAP earnings per share of $4.20, increasing 35% year-over-year or about 2 times our revenue growth rate.

在營業利益以下（below the line），非 GAAP 利息費用為 8,500 萬美元。稅率維持在 18%。非控制權益為 1,500 萬美元，而我們以權益法認列之合資企業被投資公司（equity-accounted investees）貢獻為 400 萬美元虧損。因此非 GAAP 每股盈餘為 4.20 美元，年增 35%，約為我們營收成長率的 2 倍。

Turning now to the balance sheet, cash flow and capital returns. Our balance sheet remains strong, providing flexibility to invest in our strategic growth priorities, advance our hybrid manufacturing model and return capital to shareholders. We ended Q3 with $3.94 billion of cash and $10.98 billion of debt, resulting in a net leverage ratio of 1.3 times.

接著談資產負債表、現金流與資本回饋。我們的資產負債表仍然穩健，提供彈性以投資策略性成長優先事項、推進混合製造模式，並回饋資本給股東。第三季末我們持有現金 39.4 億美元、負債 109.8 億美元，淨槓桿比率為 1.3 倍。

Days of inventory declined to 148 days from 156 days in Q2, including approximately 13 days of prebuilt inventory related to our planned front-end factory consolidations. We expect to have approximately 15 days of prebuilt inventory at the end of 2026.

庫存天數由第二季的 156 天降至 148 天，其中包含約 13 天與我們規劃中的前段晶圓廠整併相關的預先建置庫存。我們預期在 2026 年底將有約 15 天的預先建置庫存。

For the quarter, operating cash flow was strong at $1.26 billion, and net CapEx was $50 million, resulting in free cash flow of $1.21 billion or approximately 32% of revenue. During Q3, we invested $218 million between equity and access fees in VSMC and ESMC. A detailed breakdown of our joint venture investments is available in our historical financial model on our Investor Relations website. We expect to be over 75% complete with our JV investments by year-end.

本季營運現金流強勁，達 12.6 億美元；淨資本支出（CapEx）為 5,000 萬美元，因此自由現金流為 12.1 億美元，約占營收 32%。第三季期間，我們在 VSMC 與 ESMC 之間合計投入 2.18 億美元（包含股權與取得權利費用）。關於我們合資投資的詳細拆分，可於投資人關係網站上的歷史財務模型中查閱。我們預期在年底前完成超過 75% 的合資投資。

On a trailing 12-month basis, free cash flow was approximately $3.5 billion or 25% of revenue, in line with our long-term target. Over that same period, we returned $1.7 billion to shareholders through dividends and share repurchases. Most recently, shareholder returns have been below our long-term target as we prioritize deleveraging, strategic acquisitions, and investments in our manufacturing joint ventures to support future growth and gross margin expansion.

以過去 12 個月（TTM）計，自由現金流約 35 億美元，約占營收 25%，符合我們的長期目標。在同一期間，我們透過股利與庫藏股回購向股東回饋 17 億美元。最近股東回饋低於我們的長期目標，因為我們優先進行去槓桿、策略性併購，以及投資製造合資企業，以支持未來成長與毛利率擴張。

Moving forward, we plan to accelerate our buyback activity in Q4 and 2027. Consistent with that commitment, the Board authorized a new $5 billion share repurchase program, bringing total authorization to $6.3 billion, representing around 10% of our current market capitalization. Overall, we remain committed to returning 100% of our excess free cash flow to shareholders over the long term.

展望未來，我們計畫在第四季與 2027 年加速回購活動。為落實此承諾，董事會核准新的 50 億美元股票回購計畫，使總核准額度達 63 億美元，約相當於我們目前市值的 10%。整體而言，長期來看我們仍致力於將 100% 的超額自由現金流回饋給股東。

Now turning to our expectations for Q4. We expect revenue of $3.85 billion, plus or minus $100 million, which is up 15% year-on-year or up 18% adjusted for the sale of our sensors business and up 2% sequentially. We expect non-GAAP gross margin of 59%, plus or minus 50 basis points, driven by higher revenue and manufacturing utilizations. We expect operating expenses of $830 million, plus or minus $10 million. The increase sequentially is primarily driven by our new product tape-outs to support our trusted physical AI strategy.

接著談我們對第四季的預期。我們預期營收為 38.5 億美元，上下浮動 1 億美元，年增 15%；若調整出售感測器業務的影響則年增 18%，並較前一季成長 2%。我們預期非 GAAP 毛利率為 59%，上下浮動 50 個基點，主要由較高營收與製造稼動率所帶動。我們預期營業費用為 8.30 億美元，上下浮動 1,000 萬美元。費用的季增主要來自我們為支持可信任實體 AI（trusted physical AI）策略而進行的新產品流片（tape-out）。

At the midpoint, it results in a non-GAAP operating margin of 37.5%. Below the line, we expect non-GAAP financial expense to be $79 million, and our non-GAAP tax rate to be approximately 17.5%. We expect noncontrolling interest to be $14 million and $5 million of losses in our joint venture equity-accounted investees. This implies Q4 non-GAAP earnings per share of $4.36 at the midpoint, an increase of 30% year-over-year, demonstrating continued earnings power.

以中位數計算，非 GAAP 營業利益率為 37.5%。在營業利益以下，我們預期非 GAAP 財務費用為 7,900 萬美元，非 GAAP 稅率約為 17.5%。我們預期非控制權益為 1,400 萬美元，且以權益法認列之合資企業被投資公司將有 500 萬美元虧損。這意味著第四季非 GAAP 每股盈餘中位數為 4.36 美元，年增 30%，展現持續的獲利能力。

Turning to Q4 uses of cash. We expect capital expenditures to be approximately 3% of revenue and equity payments to VSMC of $86 million and ESMC of $60 million. Looking ahead into 2027 and for your modeling considerations, we will continue to drive double-digit growth in revenue. We plan to expand our gross margins to 60% during 2027.

接著談第四季的現金用途。我們預期資本支出約占營收 3%，並預期對 VSMC 的股權付款為 8,600 萬美元、對 ESMC 為 6,000 萬美元。展望 2027 年並供各位建模參考，我們將持續推動營收雙位數成長。我們計畫在 2027 年將毛利率擴張至 60%。

We expect operating expenses for the first half to be around 23% of revenue, driven by new product introductions and second half be below 23%. Assuming a non-GAAP annual effective tax rate of 17.5%, we now expect joint venture investee losses to be around about $80 million for the full year and suggest using about a $10 million loss for Q1 and a $20 million loss for Q2.

我們預期上半年營業費用約占營收 23%，主要受新產品導入帶動；下半年則將低於 23%。在假設非 GAAP 全年有效稅率為 17.5% 的情況下，我們目前預期全年合資企業被投資公司虧損約為 8,000 萬美元，並建議第一季採用約 1,000 萬美元虧損、第二季採用約 2,000 萬美元虧損。

Capital expenditures to be around 5% as we plan to bring on more back-end capacity online in 2027 and build out our new assembly and test facility in Malaysia to be ready for production in 2028. Lastly, we will continue to generate healthy levels of cash, and we plan to accelerate capital returns to our owners. Next, to better understand our business, we plan to realign our end market revenues beginning in 2027.

資本支出約為 5%，因為我們計畫在 2027 年啟用更多後段產能，並擴建我們在馬來西亞的新封裝與測試廠，以於 2028 年具備量產準備。最後，我們將持續產生健康水準的現金，並計畫加速對股東的資本回饋。接下來，為了更好地理解我們的業務，我們計畫自 2027 年起重新調整終端市場營收的分類方式。

In addition, we plan to host our Investor Day on March 3rd in Boston, where we will update our strategic and long-term financial model. And finally, I'd like to thank Jeff Palmer, who has served us as the Head of Investor Relations for the past 16 years, and thank him for his many contributions to NXP. I would like to now turn the call back to the operator for your questions.

此外，我們計畫於 3 月 3 日在波士頓舉辦投資人日，屆時將更新我們的策略與長期財務模型。最後，我想感謝 Jeff Palmer，他在過去 16 年擔任投資人關係主管，並感謝他對 NXP 的諸多貢獻。現在我把電話交回給接線員，進行提問。

(Operator Instructions)

（接線員指示）

Jim Schneider, Goldman Sachs.

高盛（Goldman Sachs）的 Jim Schneider。

I was wondering if you can maybe talk a little bit about the customer behavior you're seeing across your OEM customers, specifically, are you seeing signs of them restocking at all in a more aggressive way than they had in past quarters? And maybe give us a sense of the customer commentary in terms of what's happening with lead times and how they're responding to that.

我想請教您是否能談談您在 OEM 客戶端所觀察到的客戶行為；具體而言，您是否看到他們相較於過去幾季，以更積極的方式進行補庫存？另外也請您分享一下客戶對交期（lead times）變化的評論，以及他們如何因應。

Yes. So, Jim, thank you for the question. I assume this is an Auto question. So in the context of auto, no, we're not seeing anything noticeable that would suggest a broad-based restocking. Now from a visibility perspective, I think visibility has greatly improved. And so lead times have extended, but also has resulted in better visibility.

是的。Jim，謝謝你的問題。我假設這是汽車相關的問題。就汽車領域而言，沒有，我們沒有看到任何明顯跡象顯示會出現廣泛性的補庫存。不過從能見度的角度來看，我認為能見度已大幅改善。因此交期拉長，但也帶來更好的能見度。

The other kind of, I would say, the other trend in the market in Automotive is OEMs are getting more involved in providing that guidance directly to companies like us. And so there's a lot more intimacy from a customer relationship perspective with respect to their needs throughout the year. So I feel like we have good visibility on what that supply is going to be required and the type of products we're going to need to deploy.

另外一個趨勢，我會說，在汽車市場中，OEM 正更積極地直接向像我們這樣的公司提供指引。因此，從客戶關係角度來看，針對他們全年需求的互動更為緊密。所以我覺得我們對所需供應量以及我們需要部署的產品類型，都有良好的能見度。

Yes. And then as a follow-up, I just wanted to sort of ask, it's very helpful to see the solid guidance for 2027. Can you maybe give us a sense about by end market, some of your -- the places where you have greatest underlying confidence in sort of outperforming that overall double-digit outlook you provided and other areas where you see there's a little bit more uncertainty?

好的。接著追問一下，我覺得看到你們對 2027 年提出穩健的指引非常有幫助。你能否按終端市場（end market）分享一下：哪些領域讓你們對於能夠超越你們所提供的整體雙位數展望最有信心；以及哪些領域你們認為不確定性相對更高？

Sure. Well, I think what is working in 2026, that will continue in 2027. I mean what your question is what's driving our confidence in the growth in double digits in 2027 and continues to be the areas that we're focused on. SDV adoption continues. As a matter of fact, it's just barely starting, and we see several ramps, especially in the second half of 2027.

當然。嗯，我認為在 2026 年行之有效的做法，將會延續到 2027 年。也就是說，你的問題是：是什麼在驅動我們對 2027 年雙位數成長的信心；答案仍然是我們所聚焦的那些領域。SDV 的採用持續推進。事實上，這才剛剛開始，我們看到多個爬坡機會，特別是在 2027 年下半年。

We have very important ramps in our SDV portfolio. The same thing goes for Industrial & IoT. The adoption of AI continues to add content gains in Industrial & IoT and it's quite strong. And the deployment that we're barely starting on the control plane and data center continues to play out for us.

我們在 SDV 產品組合上有非常重要的爬坡。工業與 IoT 也是同樣的情況。AI 的採用持續在工業與 IoT 帶來內容量（content）提升，而且相當強勁。此外，我們才剛開始在控制平面與資料中心的部署，這也會持續為我們帶來成果。

Thomas O'Malley, Barclays.

Thomas O'Malley，巴克萊。

Maybe this is for Bill. When I look at December guidance, I think you guys are guiding to 59% gross margins. It's a pretty big step-up in incrementals from September. I think it's 85% quarter-over-quarter, close to 70% year-over-year. Obviously, volume is some component of that, but maybe walk through some other things that are driving that just as that's a big step-up from what you've seen over the last three, four quarters.

也許這題給 Bill。當我看 12 月的指引時，我認為你們指引毛利率為 59%。相較於 9 月，增量毛利（incrementals）有相當大的跳升。我想季對季大約是 85%，年對年接近 70%。顯然，出貨量是其中一個因素，但能否也請你說明一下，還有哪些因素在驅動這個表現？因為這相較於過去三、四個季度看到的水準，是一個很大的躍升。

Absolutely, Thomas. A couple of points on gross margin. First, they're performing to our expectations. And as we shared, they continue to expand into Q4. And I would say our gross margin improvements are very structural in nature.

當然可以，Thomas。關於毛利率有幾點。第一，它們的表現符合我們的預期。而且如同我們分享的，毛利率在第四季仍會持續擴張。我會說，我們的毛利率改善在本質上是非常結構性的。

They're driven by a combination of the higher revenues, the higher utilizations. As I mentioned, we're running last -- I gave guidance running in the mid-80s. They are running in the mid-80s now. We expect them to continue in Q4 in the mid-80s. And we're actually in 2027, expect them to run in the high 80s.

這是由更高的營收、更高的產能利用率等多項因素共同驅動。如我提到的，我們目前的運轉——我之前給的指引是落在 80% 中段。現在確實是在 80% 中段運轉。我們預期第四季仍會維持在 80% 中段。而且我們其實預期在 2027 年會提升到接近 80% 高段。

Clearly, product mix plays a role and then obviously, improved operational efficiencies. I would also say our gross margins have nothing to do with the price increases that we're seeing at the moment because all we're doing with the price increases is offsetting the higher inflationary costs that we're passing on to our customers.

很明顯，產品組合也會扮演角色，另外當然還有營運效率的提升。我也要說，我們目前看到的漲價與毛利率無關，因為我們所做的漲價只是為了抵銷更高的通膨成本，並將其轉嫁給客戶。

And then I think overall, we expect gross margins to continue to expand, achieving that 60% during 2027. And furthermore, I think they'll expand above 60% in 2028 with the ramp of our VSMC joint venture that we shared about in the past.

整體而言，我們預期毛利率會持續擴張，在 2027 年達到 60%。此外，隨著我們先前提到的 VSMC 合資案進入爬坡，我認為 2028 年毛利率將會擴張到 60% 以上。

And if you have follow-up--

如果你有追問——

I do, thanks Mike. Just on seasonality, so I'm sure everyone will have slightly different numbers, but by my numbers, Auto was a little ahead of seasonal in September. And then your guide is kind of low single digits sequentially.

有的，謝謝 Mike。就季節性來看，我知道每個人的估算可能略有不同，但按我的數字，汽車業務在 9 月其實略高於季節性。而你們的指引看起來是環比低個位數成長。

My December traditional seasonality is close to like mid- to high single digits. Any reason why the delta there? Is it because of some of the stronger quarters that you've got in June, September? Or are you calling out anything in particular that's worth noting?

我對 12 月的傳統季節性估算大概接近中到高個位數。為什麼會有這個差異？是因為你們 6 月、9 月的季度表現比較強，所以基期不同嗎？還是你們有特別指出什麼值得注意的因素？

Tom, I think you're talking about Auto specifically. Is that correct? If that's so?

Tom，我想你是在特別談汽車（Auto）對吧？是這樣嗎？如果是的話？

Auto specifically into December, yes.

是的，特別是汽車到 12 月的走勢。

So you're right. Auto has been quite strong this year and above seasonal, I would say, almost every quarter this year. Looking at 4Q, I'm not sure there's a lot of range on seasonality, a lot of different years people use 3, 5, 7, 10 years. But I think it's seasonal in Auto is probably flat to up slightly.

你說得沒錯。今年汽車業務一直相當強勁，而且我會說幾乎每一季都高於季節性。看第四季，我不確定季節性的區間有多大，因為不同人會用 3 年、5 年、7 年、10 年等不同期間來看。但我認為汽車的季節性大概是持平到小幅上升。

So I think that 2% growth is above seasonal. That's how I would view that. So I would say the strong trends continue in auto. So that's a straightforward answer on the Automotive market specifically. Next question, please.

所以我認為 2% 的成長其實是高於季節性的。我會這樣解讀。因此我會說，汽車的強勁趨勢仍在延續。這就是針對汽車市場本身的直接回答。下一題，謝謝。

Joshua Buchalter, TD Cowen.

Joshua Buchalter，TD Cowen。

Maybe following up on Tom's one. Any help you can give us on what you're seeing in the overall pricing environment now? And then as we look into 2027, I know in some prior years, there's been sort of a step down in gross margins due to the annual pricing resetting. Should we be expecting a step down in 1Q again looking into 2027 this year?

也許延續 Tom 的問題。你們能否多分享一些目前看到的整體定價環境？另外，展望 2027 年，我知道在過去某些年度，因為年度定價重置，第一季毛利率會出現某種程度的下滑。那麼在今年展望 2027 年時，我們是否也應該預期 2027 年第一季會再出現一次下滑？

Yes. Let me start off with the pricing. Going into 2026, we expected a low single-digit price decline. Now as we plan to exit 2026, we expect pricing to be neutral for the year for '26. Clearly, for 2027, we continue to see higher input costs in many areas.

好的。我先從定價談起。進入 2026 年時，我們原本預期價格會有低個位數的下滑。但現在在規劃 2026 年結束時，我們預期 2026 全年價格將大致持平。很明顯地，對於 2027 年，我們仍看到許多領域的投入成本上升。

And unfortunately, we will pass those to our customers to ensure our gross margins are not impacted. I think we expect revenue to also be supported by the mix of our products, as Rafael shared, especially in the SDV portfolio with our new solutions to drive higher average selling prices in 2027 because of the portfolio where SDVs carry 2 or 3 times more value than our traditional portfolio.

而且很遺憾，我們會把這些成本轉嫁給客戶，以確保我們的毛利率不受影響。我認為營收也會受到產品組合的支撐，正如 Rafael 分享的，特別是在 SDV 產品組合中，透過我們的新解決方案在 2027 年推動更高的平均售價，因為在 SDV 的產品組合中，其價值量大約是我們傳統產品組合的 2 到 3 倍。

But in the short term, I would say pricing is doing what we expected. We're passing them along, neutral, I'd say Q3, maybe plus 1%. 1% was pricing in Q3, another 1% in Q4. So it keeps it neutral for the whole year for 2026.

但就短期而言，我會說定價的走勢符合我們的預期。我們把它們轉嫁出去，整體持平；我會說第三季可能是加 1%。第三季定價貢獻 1%，第四季再 1%。因此 2026 全年就能維持大致持平。

Josh, maybe let me add -- this is Rafael. Let me add a high-level perspective here. And put aside the variable input cost for a second here. The way we see this is we see a structural shift in the industry with respect to pricing, right? Unlike the past decade, where annual price erosion was common. Today, I think semiconductors account for a larger share of the value in the innovation in the systems.

Josh，我補充一下——我是 Rafael。我想在這裡補充一個更高層次的觀點。先暫時把可變投入成本放一邊。我們的看法是，產業在定價方面正在發生結構性轉變，對吧？不同於過去十年，年度價格侵蝕（price erosion）是常態。今天，我認為半導體在系統創新中的價值占比更高。

And I think that creates opportunities for much better value capture, right? This has nothing to do with input costs, right? It's just a shift in value with respect to semiconductors and ultimately on the systems. Sure, competitive pressures will always remain. But I think what we expect is the historical pattern of annual price erosion to actually moderate over time. And I think it's going to -- over time, I think it's going to make pricing less of a headwind year-over-year.

而我認為這會帶來更好的價值擷取機會，對吧？這與投入成本無關，對吧？這只是半導體相對於系統的價值移轉。當然，競爭壓力永遠都會存在。但我認為我們預期，歷史上那種年度價格侵蝕的模式，會隨時間推移而趨於緩和。而且我認為，長期來看，這會讓定價在年對年的比較中，成為較小的逆風。

As a follow-up, you mentioned that you're not seeing any signs of widespread restocking. It seems like there's been pretty mixed commentary from the industry over the last couple of quarters about whether restocking is going on or not.

追問一下，你提到你們沒有看到廣泛補庫存（restocking）的跡象。過去幾個季度，產業對於是否正在補庫存的評論似乎相當分歧。

How much of, I guess, what you're seeing is proactive conservatism on your part, similar to how you handled the previous cycle? Or is there anything different about your portfolio or products that -- or exposure that would give you a different cadence of restocking than some of your peers that are saying that they're seeing it already?

你們看到的情況，有多少是你們主動採取的保守作法，類似你們在上一個循環的處理方式？或者，你們的產品組合、產品或曝險結構是否有任何不同之處——使得你們的補庫存節奏會與一些同業不同；那些同業表示他們已經看到補庫存正在發生？

Josh, we're not seeing it, right? I can only tell you what we see. We're just not seeing it. And the fact that our core products are at the core of the software-defined vehicle, I don't see how we -- if there was restocking, we would see it, right? What it has, has been conversations by some OEMs, right, to be able to actually have some safety stock put in place given the tightness in the supply right now, but it has yet to occur, right?

Josh，我們沒有看到，對吧？我只能告訴你我們看到的情況。我們就是沒有看到。而且考量到我們的核心產品正是軟體定義車輛（SDV）的核心，我不明白如果真的有補庫存——我們應該會看得到，對吧？目前有的是一些 OEM 的討論，對吧，鑑於目前供應緊張，想要實際建立一些安全庫存，但到目前為止仍未發生，對吧？

So I mean, I acknowledge some of the headlines by some Tier 1 OEMs, but we have not seen the restocking. And given the fact that our products are core to the core function of the vehicle, if there were any restocking, we will see it.

所以我的意思是，我承認一些一級供應商（Tier 1）OEM 的新聞標題，但我們並沒有看到補庫存。而且鑑於我們的產品是車輛核心功能中的核心，如果有任何補庫存，我們一定會看得到。

Yes. Maybe just to add, remember, for automotive distribution, we service that by about 40%, and we have a really good control on that.

是的。也許補充一下，記得在汽車通路（distribution）方面，我們大約服務其中 40%，而且我們對那部分有非常好的掌控。

Blayne Curtis, Jefferies.

Blayne Curtis，Jefferies。

I just want to start on the Auto segment. Obviously, you have a lot of structural drivers with SDV. I was curious on the electrification side, what are you seeing? I think Texas saw an uptick. Curious with gas prices. Maybe start with that.

我想先從汽車（Auto）部門開始。顯然，你們在 SDV 上有很多結構性驅動因素。我想了解在電動化這一側，你們看到了什麼？我記得德州那邊有看到回升。也想知道在油價的影響下。也許先從這部分談起。

Yes. This is Bill. Blayne, we are seeing strength across all our company-specific growth drivers in Automotive. And last year, electrification was going through inventory digestion, and our new products were delayed a bit. But those are performing quite well this year and catching up, and I would say, growing above that growth rate that we provided during Investor Day, specifically in electrification.

是的。我是 Bill。Blayne，我們在汽車領域所有公司特定的成長驅動因素上都看到強勁表現。去年電動化正處於庫存去化階段，而我們的新產品也稍微延後了一些。但今年這些產品表現相當不錯、正在追趕進度；我會說，特別是在電動化方面，成長速度高於我們在投資人日所提供的成長率。

If I step back and look at total SAAR, total SAAR is slightly down. But the penetration of xEVs, including hybrids, you combine that, that's actually growing. I mean I think total production this year is forecasted to be about 47.5% roughly of xEVs. And next year, it's expected to be across the 50% mark to about 55%. So the 47% this year compared to last year was 41%, growing 12%.

如果我退一步看整體 SAAR（季調後年化銷售率），整體 SAAR 略為下滑。但 xEV（含油電混合車）滲透率，把混合動力也算進去，其實是在成長。我的意思是，今年總產量預估約有 47.5% 左右是 xEV。而明年預期會跨過 50% 的門檻，約到 55%。所以今年 47% 相較去年 41%，滲透率成長 12%。

The forecast is to grow another 15% of that penetration rate. So from a macro standpoint, we do see xEVs continuing to grow globally. And as you know, it carries 2 to 3 times more semi content in the industry compared to a regular ICE vehicle.

預測滲透率還會再成長 15%。所以從總體角度來看，我們確實看到 xEV 在全球持續成長。而如你所知，相較於一般燃油車（ICE），xEV 在產業中的半導體含量高出 2 到 3 倍。

And then Bill, don't kill me for this, but I wanted to ask you any early thoughts on March, either kind of by segment or overall?

另外 Bill，別因為我問這個就生氣，但我想請教你對 3 月有沒有任何初步想法，不管是按部門或整體來看？

So you're saying for -- for the first quarter, you're saying--

所以你是說——第一季，你是說——

Sure, yes.

對，沒錯。

So I knew someone asked the question, Blayne. I thought. But it was you, Blayne, you were ahead of the queue. So again, this is not guidance. There's a wide range on these expectations. I'll give my normal caveats there.

我就知道會有人問這個問題，Blayne。我想也是。但原來是你，Blayne，你排在隊伍前面。再次強調，這不是財測指引（guidance）。這些預期的區間很大。我先給出我一貫的但書。

But typically, 1Q for us is down about mid-single digits. What drives that mid-single-digit sequential decline is Mobile down mid-teens sequentially and Automotive, Industrial & IoT, and Comms Infra is down kind of low to mid-single digits, call it, 3%, 4%, 5% sequentially. Again, this is not guidance. This is just overall trends over the past 5, 10 years in our business.

但通常我們的第一季會季減約中個位數百分比。造成這個中個位數季減的原因是：行動（Mobile）季減約中十位數百分比；而汽車、工業與 IoT、以及通訊基礎設施（Comms Infra）則大概是低到中個位數季減，約 3%、4%、5% 的季減。再次強調，這不是指引。這只是過去 5 到 10 年我們業務的整體趨勢。

Harlan Sur, JPMorgan.

Harlan Sur，摩根大通（JPMorgan）。

And thanks for the near-term trends you're seeing in your Auto franchise. Obviously, you guys have many other company-specific drivers in your other segments that are firing, edge AI, physical AI, data center to name a few. So if you look across your overall business, was book-to-bill greater than one in Q3? And if so, did backlog grow across all your segments? And did overall customer escalations continue to rise just given the tight supply environment?

謝謝你們分享汽車業務（Auto franchise）近期看到的趨勢。顯然，你們在其他部門也有許多公司特定的驅動因素正在發力，例如邊緣 AI、實體 AI、資料中心等等。所以如果看你們整體業務，第三季的接單出貨比（book-to-bill）是否大於 1？如果是的話，各部門的在手訂單（backlog）是否都有成長？另外在供應緊張的環境下，整體客戶的加急/升級需求（customer escalations）是否持續上升？

Yes. Let me take that question. I would just say all signals continue to improve from 90 days ago. So the book-to-bill remains well above one across all end markets and regions. Our backlog continues to grow in dollars and extend further into 2027. Lead times, I would say, continue to extend, just to give you a feel.

是的。我來回答這個問題。我會說，相較 90 天前，所有訊號都持續改善。所以在所有終端市場與各地區，接單出貨比仍然遠高於 1。我們的在手訂單以金額計持續成長，並且延伸到 2027 年以後。我會說交期（lead times）也持續拉長，讓你有個概念。

I would say 80% of our lead times are above the 16 weeks compared to a year ago when it was only 25%. And customer escalations continue to increase, and they're up sizable quarter-over-quarter.

我會說，我們有 80% 的交期超過 16 週；而一年前這個比例只有 25%。而客戶加急/升級需求持續增加，且季對季大幅上升。

Do you have a follow-up, Harlan?

你有後續問題嗎，Harlan？

Yes. On the supply side, on the 300-millimeter expansion initiatives, primarily driven to your ESMC and VSMC 300-millimeter JVs, I believe VSMC, I think that's targeted for analog power, mixed-signal products. I think VSMC just started risk production/qualification with production shipments targeted for the first half of next year.

有。在供應端，關於 300 毫米擴產計畫，主要是由你們的 ESMC 與 VSMC 300 毫米合資（JV）所驅動。我相信 VSMC 主要是針對類比電源、混合訊號產品。我想 VSMC 剛開始風險試產/認證（risk production/qualification），量產出貨目標是在明年上半年。

And if I recall, ESMC for your MCU and embedded products, those were targeted for production sort of end of next year. Can you guys just give us an update on these initiatives? In a tight supply environment, it seems like your 300-millimeter programs are layering in just like right at the perfect time.

如果我沒記錯，ESMC 是針對你們的 MCU 與嵌入式產品，目標大概是在明年年底左右開始量產。你們能否更新一下這些計畫的進度？在供應緊張的環境下，看起來你們的 300 毫米計畫正好在非常完美的時間點逐步到位。

Yes, absolutely. As you all know, it's happening both in the front end and the back end. So front end, recently, we just celebrate the opening of VSMC, and that is ahead of schedule. And the first thing that's going to ramp is actually our partner side, which is ramping the interposers. In our side, we are in the process of qualifying our first set of technologies, which we will start that material in the first half and expect out in the second half.

是的，當然可以。如各位所知，這同時發生在前段與後段。以前段來說，最近我們剛慶祝 VSMC 開幕，而且進度超前。第一個要爬坡（ramp）的其實是我們合作夥伴那一側，他們正在提升中介層（interposers）的產能。在我們這一側，我們正在進行第一批技術的認證流程；我們會在上半年開始投料，並預期在下半年產出。

So we do see some outs coming ahead of schedule, but really, the big bulk where we see a lot of volume pickup is in 2028 for us, obviously, also helping to expand our gross margins. As you know, you get the 2 benefits from that. You get the benefit from moving 200 millimeter to 300 millimeter, but you also get a benefit of not paying that margin stacking like you buy from an outside foundry partner.

所以我們確實看到有些產出會早於原先時程，但真正的大量放量（volume pickup）對我們而言是在 2028 年，當然也有助於擴大毛利率。如你所知，這有兩個好處。一是從 200 毫米轉到 300 毫米的效益；二是不用支付像向外部晶圓代工夥伴採購時那種層層加價的毛利堆疊（margin stacking）。

Related to the back end, as I mentioned in my prepared remarks, we are increasing our CapEx to about 5% versus, I would say, we've been spending around 2% this year. And there's two factors there. There's -- one is we are adding capacity into Q1 and Q2 to support our revenue in the back end as we have a little space, and we're doing that.

至於後段（back end），如我在事先準備的發言中提到，我們正把資本支出（CapEx）提高到約 5%，相較之下，我會說今年我們大約花在 2% 左右。這裡有兩個因素。其中之一是，我們在第一季與第二季增加產能，以支援後段的營收，因為我們還有一些空間，所以正在這麼做。

And so that will help drive the double-digit growth. But more importantly, we're building out and getting production ready of a whole new facility in KL that we broke ground in August, and that will be up and running in the beginning of 2028. And that will also service a lot of revenue and also help support our gross margin expansion as well.

因此這將有助於推動雙位數成長。但更重要的是，我們正在擴建並讓一座全新的 KL 廠房達到可量產狀態；我們在 8 月已動土，該廠將在 2028 年初開始運轉。這也將支撐大量營收，並同樣有助於我們的毛利率擴張。

Matthew Prisco, Cantor.

Matthew Prisco，Cantor。

I guess, first, on the SDV side. With these increased engagements, can you kind of give more color on what's driving those wins today? And how do we think about the ramp timing and potential inflection given those wins? Is that more of a '27, '28 story?

我想先從 SDV 這邊問起。隨著這些合作案（engagement）增加，你能否更具體說明一下，現在推動這些贏單（wins）的因素是什麼？以及我們該如何看待這些贏單帶來的放量時程與可能的拐點？這會比較像是 2027、2028 年的故事嗎？

And then any ancillary opportunities you're seeing that Volkswagen partnership seems centered on the MCU side, but obviously, you're selling these systems with other components as well. So any opportunities beyond that core?

另外，你們看到的任何周邊機會（ancillary opportunities）？你們與 Volkswagen 的合作看起來是以 MCU 端為主，但顯然你們在銷售這些系統時也會搭配其他元件。所以除了核心之外，還有沒有其他機會？

Yes. So what is driving some of the design win is the massive adoption that we've been discussing about software-defined vehicle, right? And the whole concept is just revolutionary. It's transforming the way a vehicle is made. And the two architectures are the zonal architecture and the central compute architecture, and I'm simplifying it a little bit.

是的。推動部分設計贏單（design win）的因素，是我們一直在討論的軟體定義車輛（software-defined vehicle）的巨大採用潮，對吧？而整個概念確實是革命性的。它正在改變車輛的製造方式。兩種架構分別是區域式（zonal）架構與中央運算（central compute）架構，我這裡稍微簡化說明。

The Volkswagen press announcement was surrounded about their zonal architecture, which uses our high-performance S32K5. Having said that, we also have plenty of engagements in our S32N central compute architecture as well. I would say that we already have design wins in North America. We have a recently designed win in South Pacific and Asia.

Volkswagen 的新聞稿主要圍繞其區域式架構，該架構採用我們的高效能 S32K5。話雖如此，我們在 S32N 中央運算架構方面也有大量合作案（engagement）。我會說我們在北美已經拿到設計贏單。我們最近也在南太平洋與亞洲拿到一個設計贏單。

And we have different -- several POCs in China and in Europe. The early ramps of the central compute S32N are likely to happen at late 2027. And the full volume ramps for both the K5 and the S32N, you should expect -- we should expect that in 2028.

此外，我們在中國與歐洲也有不同的——多個概念驗證（POC）。中央運算 S32N 的早期放量（ramp）很可能會在 2027 年底發生。而 K5 與 S32N 兩者的全面量產放量（full volume ramps），你應該預期——我們也預期會在 2028 年。

So we're quite bullish on what is happening there because it's not only content gain that is happening through SDV, but the design wins and as we provide more and more transparency between now and the next year, it will show that the compounds -- this content gain compounds with share gain.

所以我們對那邊正在發生的事情相當樂觀，因為透過 SDV 發生的不只是內容量提升（content gain），還有設計贏單；而且當我們在現在到明年之間提供越來越多透明度時，會顯示這些——內容量提升會與市占提升（share gain）相互疊加、形成複利效應。

And you have a follow-up?

你還有追問嗎？

Helpful. Yes. And then maybe just on physical AI quick. Last time -- last call, I think you started talking about a pipeline you were seeing there and talked about your system-level approach. So maybe update on how you're thinking about that physical AI opportunity kind of sizing that into the future, maybe a percentage of processor revenues as you think about '27 and beyond, like how quickly this market can grow?

很有幫助。有的。再快速問一下實體 AI（physical AI）。上一次——上一通電話會議，我記得你開始談到你在那邊看到的一個管線（pipeline），也談到你們的系統級方法。所以能否更新一下你們如何看待這個實體 AI 機會、以及未來的規模化？例如你們在思考 2027 年及以後時，它可能占處理器營收的百分比？以及這個市場能以多快速度成長？

Well, that is the core of what we're discussing the physical AI. AI is the main driver for content gain in Industrial & IoT. I think we provided last -- I think the last call that AI-enabled processors without the Discrete NPU was approximating 15% of total processor revenue, and we expect that to -- from doubling from a year ago, right?

嗯，這正是我們在討論的實體 AI 的核心。AI 是工業與 IoT（Industrial & IoT）內容量提升（content gain）的主要驅動力。

So we expect that to continue to go very strongly into 2027. The NPU -- the Discrete NPU, Ara, we're now getting significant traction with different POCs, both in Industrial & IoT and Automotive, and we now expect early production in 2028.

因此我們預期這個趨勢會一路強勁延續到 2027 年。至於 NPU——獨立 NPU Ara，我們目前在工業與 IoT 以及車用領域都透過不同的概念驗證（POC）取得顯著進展（traction），並且我們現在預期在 2028 年開始早期量產（early production）。

Vivek Arya, Bank of America Securities.

Vivek Arya，美國銀行證券（Bank of America Securities）。

Rafael, one of the pushbacks against NXP has been that the company doesn't have enough exposure to the data center. And I think you mentioned this year's target of $500 million plus, that's like low single digit. Even the $1 billion would be less than 10% versus many of your peers that have anywhere between 10% to 20% data center exposure.

Rafael，外界對 NXP 的一個質疑是，公司在資料中心的曝險不夠。而我想你提到今年的目標是 5 億美元以上，那大概是低個位數百分比。即使到 10 億美元，相較許多同業在資料中心曝險介於 10% 到 20%，你們也會低於 10%。

So the question is, is data center not an imperative for NXP? And if it is, then do you think you can go after it organically? Or -- because you're suggesting that the use of cash will be more towards buybacks rather than something else. So just strategically, how do you think about the data center business? And how important is it? And how big of an exposure should NXP have for that end market?

所以問題是：資料中心對 NXP 來說不是一個必須（imperative）嗎？如果是，那你認為你們能以有機方式切入嗎？或者——因為你暗示現金的使用會更偏向庫藏股回購，而不是其他用途。所以從策略上，你們如何看待資料中心業務？它有多重要？NXP 在這個終端市場的曝險應該要到多大？

Yes. Maybe, Vivek, thanks for the question. Maybe summarize it a little bit, right? So today, we provided an outlook that said data center was going to be about $500 million. And we just said that it's going to -- we're going to probably close the year exceeding that number.

是的。Vivek，也許先謝謝你的問題。我稍微總結一下，對吧？今天我們給的展望是資料中心大約 5 億美元。而我們也剛說過，今年結束時我們可能會超過這個數字。

And now with the new products that we're developing and the engagements that we have, we now expect that revenue to go to $1 billion in 2028. Now the thesis and the answer to the question, the thesis of our playbook and data plane is that -- sorry, the control plane is the data plane is going up in performance.

而現在，隨著我們正在開發的新產品以及我們已有的合作案（engagement），我們預期該營收在 2028 年會達到 10 億美元。至於我們的打法（playbook）與對這個問題的核心論點：在資料中心裡——抱歉，是控制平面（control plane）——資料平面（data plane）的效能正在提升。

And therefore, the control plane needs also more performance. And that means not just performance and processing, but more networking, more security, more safety. And that's where the core competency of NXP is.

因此，控制平面也需要更高效能。這不只是效能與運算，還需要更多網路連接（networking）、更多資安（security）、更多安全（safety）。而這正是 NXP 的核心能力所在。

And so because there's zero tolerance for downtime, right? Reliable industrial-grade control processors are required. So the way you got to look at it is our portfolio is top-of-rack switches, right? And that's starting to actually kind of roll. And then controllers and processors for board management, where there are secure elements, there are processors and microcontrollers, and also interfaces.

而且因為對停機是零容忍，對吧？因此需要可靠的工業等級控制處理器。所以你應該這樣看：我們的產品組合包含機架頂端（top-of-rack）交換器，對吧？而那部分其實已經開始逐步放量。另外還有用於板級管理（board management）的控制器與處理器，其中包含安全元件（secure elements）、處理器與微控制器，以及各種介面（interfaces）。

And so we're developing right now more and more products to be able to expand the SAM. We are using the core competence of NXP to address the control plane and data center. And I think the growth that we're going to see between $500 million and $1 billion in the next two years with actually new ramps and products is quite impressive.

因此我們目前正在開發越來越多產品，以擴大可服務市場（SAM）。我們正運用 NXP 的核心能力來切入控制平面與資料中心。我認為未來兩年在新放量與新產品帶動下，從 5 億美元成長到 10 億美元的增幅其實相當令人印象深刻。

Yes. And Vivek, I wouldn't read into the buyback of saying we're not investing in data centers. We're generating, as you can see, a lot of cash, and we are going to invest in the data center. And you'll probably hear more about that during our Investor Day in March.

是的。Vivek，我不會把回購解讀成我們不投資資料中心。如你所見，我們正在產生大量現金，我們也會投資資料中心。你可能會在 3 月的投資人日（Investor Day）聽到更多相關內容。

I'll add one thing. I think you should look at it, and I think it's a very negative the question you asked, in my opinion. If you look at it, I think data center is going to become another leg of the stool of our growth story. We have Auto as a core franchise. We talk about SDV, that's winning.

我再補充一點。我認為你問的這個問題——以我的看法——是非常負面的。如果你從整體來看，我認為資料中心將成為我們成長故事的另一隻腳。車用（Auto）是我們的核心特許經營（core franchise）。我們談 SDV，而那正在贏。

That's fantastic. We have data centers, a new opportunity, emerging opportunity. Rafael outlined some things there that we're doing to drive that growth faster in the future. Then you have the Industrial & IoT market and physical AI is another opportunity. So we really have three legs to this growth stool for NXP. We're not hooked on one thing. We have multiple growth drivers, which are net positive for the company.

那非常棒。我們還有資料中心，這是一個新的機會、正在浮現的機會。Rafael 剛才概述了一些我們正在做的事情，以在未來更快推動那部分成長。接著還有工業與 IoT 市場，而實體 AI 又是另一個機會。所以 NXP 的成長凳（growth stool）其實有三隻腳。我們不會只押注單一件事。我們有多個成長驅動因素，對公司而言是淨正面（net positive）。

And for my follow-up, on your Comms Infrastructure and Other, what is the right way to think about growth there as we look into next year? Because this year, it was, I think, the fastest-growing segment or has been the fastest growing segment after declining for the last two years. So what are kind of the puts and takes, Rafael, as we look at that specific segment for next year?

我的追問是，關於你們的通訊基礎設施與其他（Comms Infrastructure and Other）業務，當我們展望明年時，應該用什麼方式來看待那邊的成長？因為今年它我想是成長最快的部門，或是在過去兩年下滑之後成為成長最快的部門。所以 Rafael，當我們看明年這個特定部門時，主要的利多與利空（puts and takes）會是什麼？

Yes. Let me take you and just remind you how that segment is broken up. And clearly, as I mentioned in my prepared remarks, we are going to realign the segment starting in 2027. So we are more transparent in how to understand this business and the right components. At 2025 -- exiting 2025, roughly 50% was secure card business, 25% was DN, digital networking, and 25% was RF power.

是的。讓我帶各位回顧一下這個事業部是如何拆分的。而且很明顯，如同我在事先準備的發言中提到的，我們將從2027年開始重新調整該事業部的架構。因此，我們在理解這項業務及其正確組成部分方面會更透明。在2025年——也就是2025年結束時，大約50%是安全卡業務，25%是DN（數位網路），另外25%是RF功率。

As we look at the end of 2026, obviously, and we shared this, is that the data center or half of it is coming from DN. So expect the Comms and Infra for DN to be above 30%. And then we've exited next-generation RF power type of business and expect that to be below 20% as we exit 2026.

當我們看2026年年底時，很明顯，而且我們也分享過，資料中心（業務）或其中一半是來自DN。因此，預期DN的通訊與基礎設施占比將高於30%。然後我們已退出下一代RF功率類型的業務，並預期在2026年結束時其占比將低於20%。

And I think what you'll see that continued trend along with nice growth continuing in our secured card business. But we're going to break that out and make it much more transparent in the future in our Q4 earnings when we release.

我想你們會看到這個趨勢持續，同時我們的安全卡業務也會持續有不錯的成長。但我們會把它拆分出來，並在未來（特別是我們發布第四季財報時）讓資訊更透明。

Can we go to our last question please.

我們可以進入最後一個問題嗎？

Francois Bouvignies, UBS.

Francois Bouvignies，瑞銀（UBS）。

So I have two quick questions, maybe a follow-up of the data center part. I mean you mentioned $1 billion for 2028 from the $500 million this year. I mean if we look at your peers, and I mean, most of them are looking at doubling their revenues into next year already. When you look at the gigawatt additions in different data, you can see that it's going to double in '27, maybe potentially versus '26.

我有兩個簡短問題，可能先追問一下資料中心的部分。你提到2028年是10億美元，而今年是5億美元。如果我們看同業，我的意思是，多數同業已經在看明年營收就要翻倍。當你看不同資料中心的GW新增量，你可以看到在27年可能會相較26年翻倍。

So all the data suggest that the data center market should double almost next year already. So I was just wondering for your $1 billion on '28, that would suggest a slower growth in '27. Is it you being conservative here? Or can you reach $1 billion as early as next year? Or should we be aware about any drag into next year?

所以所有數據都顯示資料中心市場幾乎在明年就應該翻倍。因此我想問，你們2028年10億美元的目標，似乎意味著2027年的成長會比較慢。這是你們比較保守嗎？還是你們最快明年就能達到10億美元？或者我們需要留意明年是否會有任何拖累因素？

Francois, I think the first thing when we look at our data center business is the type of products that we are deploying. We're not in discrete, we're not power. We are in the control plane. Anything that is an MCU, a microprocessor in the rack, we consider that our market, okay? We started this business almost this year. And so right now, the growth relies on product ramps. The growth relies on acceptance and design wins, and also relies on customers deploying software into our products.

Francois，我想第一點是，當我們看資料中心業務時，要看我們正在部署的產品類型。我們不做分立元件，我們也不在功率（電源）領域。我們是在控制平面（control plane）。機櫃（rack）裡任何屬於MCU、微處理器的部分，我們都視為我們的市場，好嗎？我們幾乎是今年才開始這項業務。因此目前的成長取決於產品放量爬坡（product ramps）。成長也取決於客戶接受度與設計導入（design wins），同時也取決於客戶把軟體部署到我們的產品上。

And so it's a different type of portfolio. I believe that it's even more sticky than some of the products and some of the companies that you are mentioning. So we feel quite excited about the ramp of our business in data centers because these are processors, these are core competencies that NXP has and networking and security and in processors. And I think this is going to be something that NXP is going to do quite well.

所以這是一個不同類型的產品組合。我相信它甚至比你提到的一些產品與公司更具黏著度。因此我們對資料中心業務的放量爬坡感到相當興奮，因為這些是處理器，這是NXP在網路、安全與處理器方面的核心能力。我認為這會是NXP能做得相當好的領域。

And maybe my follow-up would be on the pricing. I mean you mentioned, Bill, input cost as a driver of pricing. Rafael, you mentioned the mix as well impact -- that will impact pricing into next year. One other area of pricing could be the supply-demand environment and many people are starting to talk about the tightness and that could lead to a COVID-like environment.

我的追問可能是關於定價。你提到，Bill，投入成本是定價的驅動因素。Rafael，你也提到產品組合（mix）的影響——這會影響到明年的定價。定價的另一個面向可能是供需環境，很多人開始談到供應吃緊，這可能導致類似COVID時期的環境。

Now when I look at the utilization rate of your peers, I mean, Renesas at 60%, TI has new fabs ramping up. I mean Infineon has a new fab ramping up. You have a new fab ramping up. And you didn't mention the tightness pricing. Is it fair to say that it's not as tight as people think. Maybe you have a lot of capacity coming. Maybe it's a timing issue. It's a lead times issue, but it's temporary, not really fundamental like COVID was. Is that fair to say?

現在當我看你們同業的產能利用率，例如瑞薩（Renesas）在60%，TI有新晶圓廠在爬坡。英飛凌（Infineon）也有新廠在爬坡。你們也有新廠在爬坡。但你們沒有提到「供應吃緊帶來的定價」。是否可以說其實沒有大家想的那麼緊？也許有很多產能即將到位。也許是時點問題。是交期（lead times）問題，但屬於暫時性，而不是像COVID那樣的根本性問題。這樣說公平嗎？

Yes. Let me take that question, Francois. Obviously, it's very well known that the back-end capacity today is very, very tight. We do see into the future that the front end gets tight. And the folks that are having capacity come online like ourselves, like some of the others will help that, obviously. But we do see capacity, and that's the next phase of any supply and demand imbalance.

是的。Francois，讓我來回答這個問題。很明顯，大家都很清楚目前後段（back-end）產能非常、非常吃緊。我們也看到未來前段（front-end）會變得吃緊。而像我們以及其他一些公司那樣有產能陸續上線，顯然會有所幫助。但我們確實看到產能（緊張）正在形成，而這就是任何供需失衡的下一階段。

We see that happening sometime next year. But we're very fortunate of the investments that we have placed both in the back end and the front end that we'll be able to service it. Right now, the pricing is only related to inflationary costs. Could there be pricing related to demand and supply imbalances there like COVID like we saw? Absolutely, but we'll wait and see how that plays out.

我們預期這會在明年某個時點發生。但我們很幸運，因為我們在後段與前段都已投入資本，因此能夠提供服務、滿足需求。目前定價只與通膨成本相關。是否可能出現因供需失衡而帶來的定價（上行），像我們在COVID時看到的那樣？當然有可能，但我們會再觀察事情如何發展。

And then, Francois, the demand environment is very different, right? So you mentioned COVID, but the demand environment that is happening now is real. It's driven by transformation, SDVs transformation, physical AI transformation, data CapEx investment transformation. And so it's a much different environment. And I think that the demand is going to fuel value creation, value capture and potentially pricing power.

另外，Francois，需求環境非常不同，對吧？你提到COVID，但現在正在發生的需求環境是真實的。它是由轉型所驅動：SDV（軟體定義車輛）轉型、實體AI（physical AI）轉型、資料中心資本支出（CapEx）投資轉型。所以這是完全不同的環境。我認為需求將推動價值創造、價值擷取，並可能帶來定價能力。

And that does conclude today's Q&A session. I would like to turn the call over to Rafael for closing remarks. Please go ahead.

以上就是今天的問答環節。我想把電話會議交回給Rafael做結語。請開始。

Thank you, everyone. I want to leave you with three takeaways. First, our growth is structural. Software-defined vehicles, the intelligent industrial edge, and data center are multiyear drivers and wins like RV Tech show we are positioned for the next leg of growth. Second, this growth is broad-based with every region growing and Automotive, Industrial & IoT, and Comms and Infrastructure all growing double digits year-over-year.

謝謝各位。我想用三個重點作結。第一，我們的成長是結構性的。軟體定義車輛、智慧工業邊緣，以及資料中心，都是跨多年的成長驅動因素；像RV Tech這樣的勝出案（wins）也顯示我們已為下一階段成長做好布局。第二，這個成長是廣泛的：各地區都在成長，而汽車、工業與IoT、以及通訊與基礎設施，年對年都呈現雙位數成長。

And last, our growth results in robust free cash flow, enabling us to invest in the business and return cash to shareholders. As disclosed today, we have a new $5 billion repurchase authorization, positioning us to accelerate our buybacks beginning in Q4. Thank you all for joining us today.

最後，我們的成長帶來強勁的自由現金流，使我們能投資於業務並回饋股東。如同今天揭露的，我們新增50億美元的庫藏股回購授權，使我們能從第四季開始加速回購。感謝各位今天的參與。

Thank you for joining us today. You may now disconnect. This concludes today's program.

感謝各位今天加入。您現在可以掛線。今天的節目到此結束。