恩智浦 (NXPI) 2026 Q2 法說會逐字稿

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

大家好，感謝您撥冗等候。歡迎參加恩智浦（NXP）2026 年第二季財報電話會議。(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 Jeff Palmer, Senior Vice President of Investor Relations. Please go ahead.

現在我想把會議交給投資人關係資深副總裁 Jeff Palmer。請開始。

Thank you, Lisa. Good morning, everyone. Welcome to NXP's second-quarter earnings call. With me on the call today is Rafael Sotomayor, NXP's President and CEO; Bill Betz, our CFO; and Mike Lucarelli, our incoming Head of Investor Relations. The call today is being recorded. It will be available for replay from our corporate website.

謝謝你，Lisa。各位早安。歡迎參加恩智浦第二季財報電話會議。今天與我一同參與電話會議的有：恩智浦總裁暨執行長 Rafael Sotomayor；財務長 Bill Betz；以及即將上任的投資人關係主管 Mike Lucarelli。今天的電話會議將被錄音。稍後可至我們的公司網站收聽重播。

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. 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 financial results for the third quarter of 2026. 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.

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。這些風險與不確定性包括但不限於：宏觀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 年第三季財務結果的預期等相關陳述。恩智浦不承擔公開修訂或更新任何前瞻性陳述之義務。如需前瞻性陳述的完整揭露，請參閱我們的新聞稿。

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. Pursuant to Regulation G, NXP has provided reconciliations of the non-GAAP financial measures to the most directly comparable GAAP measures in our second-quarter 2026 earnings press release, which will be furnished to the SEC on Form 8-K and is available on NXP's website in the Investor Relations section at nxp.com.

此外，我們將提及若干非 GAAP 財務衡量指標，這些指標主要受管理層認為與恩智浦核心營運表現不直接相關之個別事件所驅動。依據 Regulation G，恩智浦已在 2026 年第二季財報新聞稿中，提供非 GAAP 財務衡量指標與最直接可比之 GAAP 指標的調節表；該新聞稿將以 Form 8-K 提交至美國證券交易委員會（SEC），並可於 nxp.com 的投資人關係專區取得。

Now I'll turn the call over to Rafael.

現在我把電話會議交給 Rafael。

Thank you, Jeff, and good afternoon. Our second-quarter performance exceeded expectations once again as the strong momentum we saw in the first quarter continued in Q2, setting the stage for a strong second half. Demand improved across all end markets, highlighted by strength in both our company-specific growth drivers and core business. This combination of secular and cyclical growth is enabling a durable revenue stream that expands margins and drives strong earnings growth.

謝謝你，Jeff，大家午安。我們第二季表現再次超出預期，第一季所見的強勁動能延續至第二季，為下半年強勁表現奠定基礎。所有終端市場的需求皆有所改善，亮點在於我們公司特定的成長驅動因素與核心業務同時展現強勢。此一結合結構性與循環性成長的態勢，正促成更具韌性的營收來源，擴大毛利率並推動強勁的獲利成長。

Second-quarter revenue was $3.5 billion, up 19% year over year. Non-GAAP operating margin of 35%, and non-GAAP EPS of $3.61, all exceeding the midpoint of our guidance. All end markets and regions grew versus the prior year. Our company-specific growth drivers grew in the mid-20% range year over year and represented roughly 1/3 of second-quarter revenue. In addition, our core businesses increased in the high-teens range year over year, and broad-based momentum is also contributing to our growth.

第二季營收為 35 億美元，年增 19%。非 GAAP 營業利益率為 35%，非 GAAP 每股盈餘（EPS）為 3.61 美元，皆高於我們財測區間的中位數。所有終端市場與各地區相較去年同期皆成長。我們公司特定的成長驅動因素年增約 20% 中段，約占第二季營收的三分之一。此外，我們的核心業務年增約 10% 高段，而廣泛的動能也正為我們的成長提供助力。

Now, turning to end market performance. In Automotive, revenue was $1.94 billion, up 12% year over year and slightly above expectations. Adjusted for the sale of the MEMS Sensors business earlier this year, Automotive growth was up 17% year over year. The company-specific growth drivers grew in the low 20% range year over year and represented 47% of the Auto business. Growth was driven primarily by Software-Defined Vehicles, electrification, and connectivity.

接著談終端市場表現。在汽車領域，營收為 19.4 億美元，年增 12%，且略高於預期。若調整今年稍早出售 MEMS 感測器業務的影響，汽車業務年增為 17%。公司特定的成長驅動因素年增約 20% 低段，並占汽車業務的 47%。成長主要由軟體定義車輛（SDV）、電動化與連線所帶動。

SDV processor design wins continue to accelerate, including S32N and S32K series platforms. Additionally, we secured new design awards for our next-generation multi-gigabit Ethernet switches, purpose-built for SDV in-vehicle network architectures. These are multi-year platform commitments that expand NXP's content per vehicle. In Industrial & IoT, revenue was $755 million, up 38% year over year and in line with our guidance.

SDV 處理器的設計導入（design win）持續加速，包括 S32N 與 S32K 系列平台。此外，我們也取得新一代多千兆乙太網路交換器的全新設計採用，該產品為 SDV 車內網路架構量身打造。這些都是多年期的平台承諾，可提升恩智浦在每輛車中的內容價值（content per vehicle）。在工業與物聯網（Industrial & IoT）方面，營收為 7.55 億美元，年增 38%，符合我們的財測。

The company-specific growth drivers, which include our newest processing portfolio of i.MX and RT MCX, grew at 40% year on year and represented 36% of the Industrial & IoT business. Communication Infrastructure revenue was $452 million, up 41% year on year, at the high end of guidance. Growth was driven by digital networking exposure to Data Center and continued ramps of our UCODE RFID products. And lastly, Mobile revenue was $351 million, up 6% year over year and in line with guidance, reflecting normal mid-year seasonal trends in our secure mobile transactions franchise.

公司特定的成長驅動因素（包括我們最新的 i.MX 與 RT MCX 處理產品組合）年增 40%，並占工業與物聯網業務的 36%。通訊基礎設施（Communication Infrastructure）營收為 4.52 億美元，年增 41%，位於財測區間高端。成長動能來自數位網路在資料中心的布局，以及我們 UCODE RFID 產品的持續放量。最後，行動（Mobile）營收為 3.51 億美元，年增 6%，符合財測，反映我們安全行動交易（secure mobile transactions）業務在年中正常的季節性趨勢。

Now, turning to our Data Center exposure. Nine days ago, we quantified this exposure for the first time. To recap, 2025 revenue was approximately $200 million, and we expect to exceed $500 million in 2026. Our position is squarely in the control plane of AI infrastructure, the same domain where NXP has built deep expertise across vehicles and factories for decades, now operating at hyperscale infrastructure specifically in two franchises.

接著談我們在資料中心的布局。九天前，我們首次量化了這項布局。回顧一下，2025 年營收約為 2 億美元，而我們預期 2026 年將超過 5 億美元。我們的定位明確聚焦於 AI 基礎設施的控制平面（control plane），這正是恩智浦數十年來在車輛與工廠領域累積深厚專業的範疇，如今在超大規模（hyperscale）基礎設施中，主要落在兩大產品線。

First, top-of-rack switching and SmartNIC control, anchored by our Layerscape family, which is ramping across leading hyperscalers. With every new data plane switch generation, as speed increases, the control plane performance must also increase, as there's simply more to manage, monitor, and secure. Hence, we are accelerating our Layerscape roadmap to deliver the control plane performance each new generation demands. Customer engagement gives us confidence that these programs materially broaden our addressable content and extend the franchise well into the future.

第一，是機架頂端（top-of-rack）交換與 SmartNIC 控制，以我們的 Layerscape 產品家族為核心，正於主要超大規模客戶中加速放量。隨著每一代資料平面（data plane）交換器速度提升，控制平面的效能也必須同步提升，因為需要管理、監控與防護的事項更多。因此，我們正加速推進 Layerscape 的產品藍圖，以提供每一代新產品所需的控制平面效能。客戶互動讓我們有信心，這些專案將實質擴大我們可服務的內容價值（addressable content），並使該產品線在未來長期延伸。

Second, the processors that control, monitor, cool, and secure every component within a rack. Data Center infrastructure is converging towards industrial-grade principles, where reliability, real-time monitoring, control, and zero tolerance for downtime are critical. NXP is uniquely positioned as these functions thrive on key industrial processing attributes, where our portfolio is differentiated.

第二，是用於控制、監控、散熱與保護機架內每一個元件的處理器。資料中心基礎設施正朝向工業級原則收斂，其中可靠性、即時監控、控制，以及對停機的零容忍至關重要。在這些功能高度依賴關鍵工業處理特性之際，恩智浦具備獨特優勢，因為我們的產品組合具差異化。

Now, I want to address something fundamental. AI is moving from the cloud to the physical world, into vehicles, factories, and robots. It is moving directly into the markets where NXP already has leadership positions. Intelligence deployed at the edge demands real-time performance, ultra-low power, and design in safety and security. True physical intelligence also requires distributing AI workloads across multiple layers of the system, an architecture we call the neural axis which is the foundation for deterministic and safe operation of physical AI. These are capabilities, again, NXP has spent decades building.

現在，我想談一個根本性的議題。AI 正從雲端走向實體世界，進入車輛、工廠與機器人。它正直接進入恩智浦已具領先地位的市場。部署於邊緣端的智慧需要即時效能、超低功耗，以及內建的安全與資安設計。真正的實體智慧也需要將 AI 工作負載分散到系統的多個層級；我們將此架構稱為「神經軸（neural axis）」，它是實體 AI 進行確定性且安全運作的基礎。而這些能力，恩智浦同樣已投入數十年打造。

Our differentiated position rests in three areas. First, NXP has the industry's broadest and most differentiated Edge AI compute platform. Our portfolio places the right intelligence at the right layer, from high-performance reasoning and coordination in our i.MX and S32N processors to real-time reflexive control in our S32K and i.MX RT families, all unified under our eIQ software environment. This is already translating into measurable growth. We estimate AI-enabled processors will represent approximately 15% of Industrial & IoT processor revenue in 2026, more than doubling from last year.

我們的差異化優勢建立在三個面向。第一，恩智浦擁有業界最廣泛且最具差異化的邊緣 AI 運算平台。我們的產品組合能在正確的層級部署適切的智慧：從 i.MX 與 S32N 處理器提供的高效能推理與協調，到 S32K 與 i.MX RT 系列提供的即時反射式控制，並在 eIQ 軟體環境下整合為一體。這已轉化為可衡量的成長。我們估計，具 AI 功能的處理器在 2026 年將約占工業與物聯網處理器營收的 15%，較去年增加一倍以上。

Second, Physical AI is a system problem, not just a compute problem. Intelligent machines must sense, connect, and act in real time. NXP is the only company that delivers all of this in one integrated, trusted platform, something no compute-only competitor can replicate.

第二，實體 AI 是一個系統性問題，而不僅僅是算力問題。智慧機器必須能夠即時感知、連接並採取行動。NXP 是唯一一家能在單一整合且可信賴的平台上交付所有這些能力的公司，這是任何僅提供算力的競爭對手都無法複製的。

Third, winning in Physical AI requires reaching a highly fragmented market at scale. Unlike cloud AI, the Edge spans thousands of applications and customers across Automotive, Industrial, and IoT markets. NXP's ecosystem of distribution partners, reference designs, and field support gives us unmatched reach into this market. Taken together, compute, system, and reach, Physical AI is already showing up in our revenue, and we expect it to accelerate.

第三，在實體 AI 領域勝出，需要以規模觸達高度碎片化的市場。不同於雲端 AI，邊緣端橫跨汽車、工業與物聯網市場中的數千種應用與客戶。NXP 的生態系包含經銷合作夥伴、參考設計與現場支援，使我們在該市場具備無可匹敵的觸達能力。綜合來看，算力、系統與觸達三者相加，實體 AI 已經反映在我們的營收中，我們預期其將加速成長。

Now, turning to the third quarter. The operational metrics we track to assess business health continue to strengthen, and our outlook is better than we anticipated 90 days ago. We are guiding third-quarter revenue to $3.75 billion, up 21% year over year adjusted for the MEMS Sensors sale, and up 7% sequentially. We expect all regions and all end markets to be up sequentially, a reflection of expanded customer adoption of our differentiated portfolio.

現在，轉到第三季。我們用來評估業務健康狀況的營運指標持續改善，而我們的展望也比 90 天前預期的更好。我們對第三季營收的指引為 37.5 億美元，若就 MEMS 感測器出售案進行調整，年增 21%，且季增 7%。我們預期所有地區與所有終端市場都將呈現季增，反映客戶對我們差異化產品組合的採用擴大。

At the midpoint, we expect the following trends in our business during Q3. Automotive is expected to be up in the low double-digit percent range year over year and up in the mid-single-digit range sequentially. Adjusted for the sale of the MEMS Sensors business, our guidance implies a high-teens percentage growth year over year. Industrial & IoT is expected to be up in the high 30% range year over year and up in the mid-single-digit range sequentially, continuing the strength we saw in Q2.

以中位數來看，我們預期第三季業務將呈現以下趨勢。汽車業務預期年增落在低雙位數百分比區間，且季增落在中個位數區間。若就 MEMS 感測器業務出售案進行調整，我們的指引意味著年增為高十位數百分比。工業與物聯網預期年增落在高 30% 區間，且季增落在中個位數區間，延續我們在第二季看到的強勁表現。

Mobile is expected to be down in the mid-single-digit percent range year over year and up in the mid-teens-digit range on a sequential basis. And finally, Communications Infrastructure and Other is expected to be up about 50% year over year and up in the high single-digit range versus Q2 2026. What you saw this quarter, double-digit growth driven by company-specific growth drivers and a 35% operating margin, is the compounding result of staying disciplined on the right priorities.

行動業務預期年減落在中個位數百分比區間，但季增落在十位數中段區間。最後，通訊基礎設施及其他預期年增約 50%，相較 2026 年第二季則成長落在高個位數區間。本季你所看到的，由公司特定成長動能帶動的雙位數成長以及 35% 的營業利益率，是我們在正確優先事項上保持紀律所累積的複利成果。

And now, I would 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. Q2 was a strong quarter with record revenue of $3.5 billion, up 19% year on year and 10% sequentially. All end markets performed above the midpoint of guidance, led by our company-specific growth drivers in Software-Defined Vehicles, electrification, Industrial Edge processing, connectivity, and increasing Data Center content.

謝謝你，Rafael，也向今天參與電話會議的各位致上下午好。第二季表現強勁，營收創新高達 35 億美元，年增 19%，季增 10%。所有終端市場的表現都高於指引中位數，主要由我們公司特定的成長動能所帶動，包括軟體定義車輛、電動化、工業邊緣運算處理、連接性，以及資料中心內容占比提升。

Non-GAAP gross profit was $2.03 billion, an increase of $376 million or 23% year on year. Non-GAAP gross margin was 58.0%, in line with guidance and expanding approximately 150 basis points year on year and 90 basis points sequentially. Our gross margin performance reflects better product mix, improved factory utilization, and solid fall-through on higher revenue. Non-GAAP operating expenses were $794 million, or 22.7% of revenue, within our long-term operating model. Non-GAAP operating profit was $1.23 billion, up 31% year on year. Non-GAAP operating margin was 35.1%, expanding 310 basis points year on year and 40 basis points above the midpoint of guidance.

非 GAAP 毛利為 20.3 億美元，年增 3.76 億美元或 23%。非 GAAP 毛利率為 58.0%，符合指引，年增約 150 個基點，季增 90 個基點。我們的毛利率表現反映了更佳的產品組合、工廠稼動率提升，以及在較高營收下良好的利潤落差（fall-through）。非 GAAP 營業費用為 7.94 億美元，占營收 22.7%，符合我們的長期營運模型。非 GAAP 營業利益為 12.3 億美元，年增 31%。非 GAAP 營業利益率為 35.1%，年增 310 個基點，且較指引中位數高出 40 個基點。

Taken together, our second-quarter results demonstrate that the margin expansion we are delivering is structural, driven by product mix, factory utilization discipline, and operational leverage across our fixed cost base. Below the line, non-GAAP interest expense was $87 million. Taxes were $205 million, non-controlling interest was $15 million, and results from equity-accounted investees were a $3 million loss, collectively in line with guidance. This resulted in non-GAAP earnings per share of $3.61, $0.11 above guidance.

綜合而言，我們第二季的結果顯示，我們所實現的利潤率擴張具有結構性，主要由產品組合、工廠稼動率紀律，以及固定成本基礎上的營運槓桿所驅動。在損益表下方，非 GAAP 利息費用為 8,700 萬美元。稅費為 2.05 億美元，非控制權益為 1,500 萬美元，按權益法認列之被投資公司結果為 300 萬美元損失，整體符合指引。因此非 GAAP 每股盈餘為 3.61 美元，較指引高出 0.11 美元。

Turning to changes in cash, debt, and capital returns. Our balance sheet remains strong and provides flexibility to invest in our strategic priorities and hybrid manufacturing plans. We ended Q2 with $10.98 billion in total debt and $3.2 billion in cash. Net debt was $7.7 billion, or 1.5 times adjusted EBITDA, and our adjusted EBITDA interest coverage ratio was 15 times.

接著談現金、負債與資本回饋的變化。我們的資產負債表仍然強健，並提供彈性以投資於策略優先事項與混合製造計畫。第二季結束時，我們的總負債為 109.8 億美元，現金為 32 億美元。淨負債為 77 億美元，約為調整後 EBITDA 的 1.5 倍，而我們的調整後 EBITDA 利息保障倍數為 15 倍。

In Q2, we returned $360 million to our owners, made up of $256 million in dividends and $104 million in share repurchases. We remain committed to our long-term capital allocation strategy, balancing returns to shareholders with disciplined investments in the business to support long-term profitable growth.

第二季我們向股東回饋 3.6 億美元，其中包含 2.56 億美元股利與 1.04 億美元庫藏股回購。我們仍致力於長期資本配置策略，在回饋股東與對業務進行有紀律的投資之間取得平衡，以支持長期且具獲利性的成長。

Turning to working capital. Our cash conversion cycle improved to 129 days, from 140 days in Q1. Days of inventory improved to 156 from 165 days, inclusive of approximately 9 days of prebuilds for our planned front-end factory consolidations. Receivables were 33 days, and payables were 60 days, slightly better than last quarter. During the quarter, we generated $860 million in operating cash flow, which helped fund the $750 million debt retirement. $360 million of capital returns, $174 million into VSMC, $12 million into ESMC, and $69 million of net CapEx. Taken together, we generated non-GAAP free cash flow of $791 million, or 22.6% of revenue. On a trailing 12-month basis, free cash flow was approximately $2.8 billion, or 21% of revenue.

接著談營運資金。我們的現金轉換週期改善至 129 天，較第一季的 140 天縮短。存貨天數由 165 天改善至 156 天，其中包含約 9 天的預先備貨，用於我們規劃中的前段晶圓廠整併。應收帳款為 33 天，應付帳款為 60 天，較上季略有改善。本季我們產生 8.6 億美元的營運現金流，協助支應 7.5 億美元的債務清償、3.6 億美元的資本回饋、對 VSMC 投入 1.74 億美元、對 ESMC 投入 1,200 萬美元，以及 6,900 萬美元的淨資本支出。綜合而言，我們產生非 GAAP 自由現金流 7.91 億美元，占營收 22.6%。以過去 12 個月計，自由現金流約為 28 億美元，占營收約 21%。

Now turning to our expectations for Q3. We expect revenue of $3.75 billion, plus or minus $100 million, which is up 18% year on year and 7% sequentially. We expect non-GAAP gross margin of 58.5%, plus or minus 50 basis points, which is up 150 basis points year on year and up 50 basis points sequentially, driven by the higher revenue and our manufacturing utilization. We expect operating expenses of $810 million, plus or minus $10 million at the midpoint. This results in a non-GAAP operating margin of 36.9%. Below the line, we expect non-GAAP financial expenses to be approximately $85 million and our non-GAAP tax rate to be 18%. We expect non-controlling interest to be $15 million, including $5 million losses in our equity-accounted investees for VSMC and ESMC. This implies Q3 non-GAAP earnings per share of $4.11 at the midpoint.

現在談我們對第三季的預期。我們預期營收為 37.5 億美元，上下浮動 1 億美元，年增 18%，季增 7%。我們預期非 GAAP 毛利率為 58.5%，上下浮動 50 個基點，年增 150 個基點，季增 50 個基點，主要由較高營收與我們的製造稼動率所帶動。我們預期營業費用在中位數為 8.10 億美元，上下浮動 1,000 萬美元。因此非 GAAP 營業利益率為 36.9%。在損益表下方，我們預期非 GAAP 財務費用約為 8,500 萬美元，非 GAAP 稅率為 18%。我們預期非控制權益為 1,500 萬美元，其中包含按權益法認列之被投資公司 VSMC 與 ESMC 合計 500 萬美元的損失。這意味著第三季非 GAAP 每股盈餘中位數為 4.11 美元。

Turning to Q3 uses of cash. We expect capital expenditures to be approximately 3% of revenue, with VSMC capacity access fee of $70 million and equity investment of $80 million. For ESMC, we expect equity investments of $30 million. This brings our cumulative investment in VSMC and ESMC to approximately $2.4 billion, or about 70% of the total planned commitment across the two joint ventures.

接著談第三季的現金用途。我們預期資本支出約為營收的 3%，其中包含 VSMC 產能使用費 7,000 萬美元，以及 8,000 萬美元的股權投資。就 ESMC 而言，我們預期股權投資為 3,000 萬美元。這使我們對 VSMC 與 ESMC 的累計投資約達 24 億美元，約占兩家合資企業合計規劃承諾總額的 70%。

At the midpoint of our Q3 guidance, year-to-date revenue is $10.4 billion, up 17% versus the same period in 2025 and consistent with a double-digit growth trajectory. Our growth continues to be driven by the compounding effect of our company-specific growth drivers and the return to expansion of our core franchises. We remain confident that we will deliver on our financial commitments.

在我們第三季財測指引的中點，年初至今營收為104億美元，較2025年同期成長17%，並符合雙位數成長軌跡。我們的成長持續由公司特定成長驅動因素的複利效應，以及核心產品線重回擴張所帶動。我們仍有信心能夠達成我們的財務承諾。

I would like to now turn the call back to the operator for your questions. Lisa?

我現在想把電話交回給接線員，開始進行各位的提問。Lisa？

(Operator Instructions) Joe Moore, Morgan Stanley.

(接線員指示) Joe Moore，摩根士丹利。

I was quite interested in the conversation about Physical AI and how it's maybe changing your thinking and the products that you have aligned to that. Can you just talk to how that intersects customer interest? Areas like Software-Defined Vehicles, are they thinking along those lines? And just how quickly do you think that this transforms into something that's more of a Physical AI-driven market?

我對於你們談到「實體AI（Physical AI）」以及它可能如何改變你們的思考與你們對應的產品配置，感到相當有興趣。你能談談這與客戶興趣之間如何交會嗎？像是軟體定義車輛（SDV）這類領域，他們是否也朝這些方向思考？以及你認為這會多快轉變成一個更由實體AI驅動的市場？

The whole notion of cognification of the edge, whether it's Industrial and Auto, the conversation is now. I don't think we will get a design win without having a big part of a very, very strong value proposition with respect to AI. So it's happening now. But everything starts before you put AI in the system.

邊緣端的「認知化」這整個概念，不論是工業或汽車，現在就正在被討論。我不認為如果沒有在AI方面提出非常、非常強的價值主張，我們還能拿到設計導入（design win）。所以這件事正在發生，就在現在。但一切都始於你把AI放進系統之前。

Everything starts with creating a software-defined system, right, whether it's a vehicle or whether it's industrial. And so, I think that's the beginning, right? The beginning is, how do you create a software-defined system, what is the platform, and then what is the overlay that you do with AI? It's already happening. I think in one of the prepared remarks it was said that AI-enabled products in Industrial & IoT already compose, I think, 15% of our revenue around AI-enabled products.

一切都始於打造一個軟體定義的系統，對吧，不論是車輛或工業系統。所以我認為那就是起點，對吧？起點是：你如何建立一個軟體定義系統、平台是什麼，然後你如何用AI做上層疊加（overlay）？這已經在發生。我想在其中一段事先準備的談話中提到，工業與IoT中的AI賦能產品目前已經占我們與AI賦能產品相關營收的大約15%。

I think this is going to become even more material and bigger as we go into 2027. But there is simply no actual activity with a customer today, rarely, without having a real material conversation around how AI is going to get deployed.

我認為隨著我們進入2027年，這會變得更具實質性、規模也更大。但今天與客戶的實際互動中，幾乎很少會沒有一場關於AI將如何部署的、真正具實質性的討論。

And then just on the broader Automotive market, can you talk about some of the dynamics that some of your peers have talked about, maybe the beginnings of restocking from very low levels? I know you've talked about Tier 1 inventories being quite low. Are you seeing anything that's different along those lines?

另外就更廣泛的汽車市場而言，你能談談一些同業提到的動態嗎，例如從非常低的水位開始出現補庫存的跡象？我知道你們談過一級供應商（Tier 1）的庫存相當低。你們在這方面有看到任何不同的情況嗎？

Yeah, great question on the Auto side, since Auto, I think, this is where our secular story shows up very clearly. As we saw in Q2, we're growing in the high teens. This is removing Sensors. And I would say that for us, we actually see no restocking, right? What we see and what's driving our revenue, if you look at it, our accelerated growth drivers grew 22% this year, as we stated.

是的，汽車端是個很好的問題，因為我認為汽車正是我們長期結構性故事非常清楚呈現的地方。如同我們在第二季看到的，我們以高十幾%的幅度成長。這是在扣除感測器（Sensors）之後。而我會說，對我們而言，我們其實沒有看到補庫存，對吧？我們看到、以及推動我們營收的因素是——如果你看我們的數據——我們的加速成長驅動因素今年成長了22%，如我們所述。

They're becoming almost close to half of the revenue. And so, the driver there is around content. It's around content driven by an architecture transformation that SDV is doing with the vehicles, and we have a aligned a roadmap to lead this architecture shift towards SDV, which, by the way, is still in the early phases of adoption. So I don't think necessarily Auto is about restocking or is it about cycle.

它們幾乎接近占到營收的一半。因此，驅動因素在於單車內容價值（content）。這是由SDV對車輛所帶來的架構轉型所驅動的內容提升，而我們已對齊一條產品藍圖（roadmap），以引領這個朝向SDV的架構轉移——順帶一提，這仍處於採用的早期階段。所以我不認為汽車一定是補庫存，或是景氣循環的問題。

I think this is compounding on content, and I think we're very well positioned with our SDV roadmap.

我認為這是內容價值的複利累積，而我認為我們在SDV藍圖上具備非常好的布局。

Matthew Prisco, Cantor.

Matthew Prisco，Cantor。

Starting on the Industrial side, can you maybe break down the trends you're seeing between the core Industrial and the IoT side of the business and also anything to highlight in the segment from a geographic demand perspective?

先從工業端開始，你能否拆解一下你們在核心工業業務與IoT業務之間看到的趨勢，並且從地理區域需求的角度，談談這個部門是否有任何值得強調的重點？

The Industrial, I think, you saw the Industrial growth is quite strong and it's growing in the high 30s into Q2. The Industrial & IoT accelerated growth drivers were in the 40% range year on year for Q2, Matthew. But also, what's happening is the core business is coming back. And you see that our core business also grew in the high 30s.

工業方面，我想你也看到工業成長相當強勁，在第二季成長到高30%區間。Matthew，工業與IoT的加速成長驅動因素在第二季年增約40%區間。但同時正在發生的是核心業務正在回升。你也看到我們的核心業務同樣成長到高30%區間。

And so very strong growth in Industrial & IoT for a total growth for the quarter into the high 30s. And, by the way, this is now the second quarter in the year that we're in the high 30s. This is actually a market that is performing quite well for us.

因此，工業與IoT在本季合計成長到高30%區間，成長非常強勁。而且順帶一提，這已經是今年第二個季度我們維持在高30%區間。這其實是一個對我們而言表現相當好的市場。

And then maybe on the pricing side, how are you seeing that as a benefit today and maybe how much of that is impacting the Q3 guide and how you think about pricing dynamics through the year?

另外在定價方面，你們目前如何看待它帶來的效益？以及其中有多少正在影響第三季指引？你們又如何看待全年定價動態？

Yeah, I think your question highlights something that has been in the press quite a lot, which is this issue that all of us are facing with respect to inflationary pressure. And I will start the answer by saying our first move is always to mitigate the price pressure through operational efficiency. We did make some price adjustments to selected products. These adjustments are not across the board.

是的，我想你的問題凸顯了一個在媒體上被大量討論的議題，也就是我們所有人都面臨的通膨壓力。我先從這點回答：我們的第一步一向是透過營運效率來緩解價格壓力。我們確實對部分產品做了一些價格調整。這些調整並非全面性。

Price for Q2 was essentially neutral. And Q3, our guide already incorporates an estimate, but we won't know the exact impact until much later.

第二季的價格影響基本上是中性。而第三季，我們的指引已經納入一個估計值，但要到更晚之後我們才會知道確切影響。

Francois-Xavier Bouvignies, UBS.

Francois-Xavier Bouvignies，瑞銀（UBS）。

My first question is on SDV. I mean, is it possible to get an update on where you are in terms of revenues? I believe you said that it represented $1 billion revenue in '24, and you expect it to double at $2 billion in '27 to reach your targets. And you said that SDV is doing very well with high growth. Can you maybe help us quantify for this year all the growth rate? Anything on that would be great. That's my first question.

我的第一個問題是關於SDV。也就是說，是否可以更新一下你們目前在營收方面的進展？我記得你們說過，SDV在2024年代表了10億美元營收，而你們預期到2027年為了達成目標會倍增至20億美元。你們也說SDV表現非常好、成長很高。你能否協助我們量化一下今年的成長率？任何相關資訊都會很有幫助。這是我的第一個問題。

Just to state, maybe recap what we said, right, the accelerated growth drivers in Automotive grew 22% in Q2 year on year, and they became close to 50% of the total Automotive revenue of the company. SDV is the highest-growing part of the accelerated growth drivers. Remember, we have Radar, electrification, connectivity, and SDV, and SDV is the one that is performing the best.

先重申並回顧一下我們說過的內容：汽車業務的加速成長驅動因素在第二季年增22%，而且它們已接近占公司汽車總營收的50%。SDV是加速成長驅動因素中成長最快的部分。請記得，我們有雷達（Radar）、電動化（electrification）、連網（connectivity）以及SDV，而SDV是表現最好的那一項。

And obviously, that is the driver for the architecture transformation that is happening in Automotive. And we're very well positioned with respect to our roadmap. What is even more encouraging to think about this is that the current outperformance in Auto, and I think our Auto business is doing quite well, is happening without our latest products.

而顯然，這正是汽車產業正在發生的架構轉型的驅動力。就我們的產品藍圖而言，我們的布局非常到位。更令人鼓舞的是，目前汽車端的超預期表現——我認為我們的汽車業務做得相當好——是在我們最新產品尚未貢獻的情況下發生的。

S32N, our 5-nanometer products, and the S32K5, which is our flagship zonal product in 16-nanometer, they haven't even begun ramping yet. And so they are in design-win phase. And so, I think we're quite encouraged about the performance of our SDV portfolio because there is more to come with the next-generation products.

S32N，也就是我們的5奈米產品，以及S32K5——我們在16奈米的旗艦分區（zonal）產品——甚至都還沒有開始放量爬坡（ramping）。因此它們仍處於設計導入階段。所以，我們對SDV產品組合的表現感到相當振奮，因為下一代產品還會帶來更多動能。

And my follow-up would be again on the Automotive side. I mean, you guided Q3 mid-single-digit quarter on quarter, if I'm not mistaken, which seems to be roughly in line with what we have seen before. I mean, if I look at your nine-year seasonality, it was even a bit higher than 5%, so it doesn't translate so much to the Automotive recovery when we look at the quarter-on-quarter pattern, if that makes sense. So year on year, yes, because maybe you destocked last year, but we don't see a stronger recovery than maybe one would expect when we hear TI, STM. Is there any drag we should be aware of that is limiting your growth, or is it just to come, or it could come later?

我接下來的追問還是關於汽車端。我的意思是，如果我沒記錯的話，你們對第三季的指引是季對季中個位數成長，這看起來大致與我們先前看到的情況一致。我的意思是，如果我看你們九年的季節性，甚至還略高於 5%，所以從季對季的走勢來看，這並不太能反映汽車復甦（如果這樣說得通的話）。所以年對年來看，是的，因為也許你們去年有去庫存，但我們沒有看到比起聽到 TI、STM 時所預期的更強勁復甦。是否有任何我們需要留意、正在限制你們成長的拖累因素？還是只是尚未反映，或可能會在之後才出現？

Well, Francois-Xavier, I mean, the way we look at our business, given the fact that more and more of the revenue in Automotive is driven by the accelerated growth drivers, is year on year because year on year, this is where true content growth shows up. You will always have sequentially a blur of product ramps. And so, I think the better way to look at it is to look at the growth year on year.

嗯，Francois-Xavier，我的意思是，我們看待業務的方式是：鑑於汽車營收愈來愈多是由加速成長的驅動因素所帶動，因此更應該看年對年，因為年對年才是真正內容（content）成長顯現的地方。你在序列（季對季）上總是會看到產品爬坡帶來的模糊效應。所以，我認為更好的看法是看年對年的成長。

And if you look at what we're guiding into Q3, it's high teens, it's mid-teens growth year on year, with Sensors. That is quite strong. And it's above our model.

而如果你看我們對第三季的指引，年對年是高十幾%、中十幾%的成長，包含感測器（Sensors）。這相當強勁。而且高於我們的模型。

Makes sense.

了解。

Vivek Arya, Bank of America Securities.

Vivek Arya，美國銀行證券。

Rafael, historically, NXP visibility has been a quarter or quarter-plus or so, and I'm curious, how far does your visibility extend right now? Which areas would you say you have better visibility, and where do you see lead times stretching out?

Rafael，歷史上 NXP 的能見度大概是一季或一季多一點，我很好奇，你們目前的能見度延伸到多遠？你會說哪些領域的能見度更好？又在哪些地方你看到交期正在拉長？

Well, lead times are stretching out. And visibility has improved across all end segments. We have better visibility into Q4. We have visibility into Q1. And we've seen that really happen. Basically, visibility has improved throughout the year.

嗯，交期正在拉長。而且各個終端市場的能見度都有改善。我們對第四季有更好的能見度。我們對第一季也有能見度。而我們確實看到這樣的情況發生。基本上，能見度在今年一路都有改善。

Yeah, maybe I'll just add to what Rafael said. Versus 90 days ago, some of those health signals that we measure internally, Rafael is absolutely correct. Our backlog continues to grow quarter plus 1, quarter plus 2, quarter plus 3. So we have a signal of 18 months out. Our distribution backlogs, when we look into their books, continue to follow similar patterns as ours.

是的，我也許補充一下 Rafael 說的。相較於 90 天前，我們內部衡量的一些健康訊號，Rafael 說得完全正確。我們的在手訂單（backlog）持續成長，延伸到「本季加 1」、「本季加 2」、「本季加 3」。所以我們有一個可看到 18 個月之後的訊號。我們看通路的在手訂單（distribution backlogs），從他們的訂單簿來看，也持續呈現與我們類似的型態。

Clearly, our book-to-bill ratio is above 1, and it's above last quarter again. Lead times will continue to extend, and we started to extend them, if you remember, maybe three or four quarters ago. Customer escalations, what we track in the quarter, have doubled since last quarter. So everything, all the signals continue to show.

很明顯地，我們的接單出貨比（book-to-bill）高於 1，而且又再次高於上一季。交期會持續延長，而我們也已經開始延長交期了，如果你記得的話，大概是三到四季前。我們在季度內追蹤的客戶升級（escalations）自上一季以來已經翻倍。所以所有訊號都持續顯示（同樣的方向）。

Even the return parts that we get, the late orders in the quarter continue to climb as well. So everything is much better, or continues to improve, versus 90 days ago.

甚至我們收到的退回零件、以及季度內的晚下單（late orders）也都持續攀升。所以相較於 90 天前，一切都好很多，或說持續在改善。

And for my follow-up, I was hoping you could remind us of how you view your typical seasonal pattern in Q4. And given this visibility, if you could care to give us some directional indication of how it might shape up. And if I zoom out, I think on the last call, Rafael, you mentioned you're still comfortable with the '27 outlook which had a midpoint of $16 billion or so. But at that time when you had given that outlook, data center was not expected to be a big driver, and now data center is a bigger driver. So is there a different way we should think about the NXP prospects for '27 including the upside from data center?

我的追問是，希望你能提醒我們，你們如何看待第四季典型的季節性型態。並且在目前這樣的能見度之下，如果你願意的話，能否給我們一些方向性的指引，說明可能會如何發展。如果我把視角拉遠，我記得在上次電話會議上，Rafael，你提到你仍然對 2027 年展望感到安心，其中中位數大約是 160 億美元左右。但在你提出那個展望時，資料中心並不被預期會是主要驅動因素，而現在資料中心成為更大的驅動因素。那麼，我們是否應該用不同方式來看待 NXP 在 2027 年的前景，把資料中心帶來的上行也納入考量？

You're essentially asking me to guide Q4 and '27, so I'm going to pick a Q4 number real quick. Listen, I think what you're asking me with Q4 is one way to ask me to guide. We guide one quarter at a time. But I'll leave it with the following.

你基本上是在要我對第四季和 2027 年給指引，所以我先很快挑一個第四季的數字。聽著，我想你用第四季來問我，其實是一種要我給指引的方式。我們一次只對一個季度給指引。但我先用以下幾點作結。

I think we feel very good about what's happening right now with our business, right? The signals that we track, they're all pointing in the right way. And I think Bill mentioned some of these signals, right? The orders continue to strengthen. Backlog continues to build. And very importantly for 2027, I think the time when ramps are going to climb.

我認為我們對目前業務的進展感覺非常好，對吧？我們追蹤的訊號都朝著正確方向。而我想 Bill 提到其中一些訊號，對吧？訂單持續轉強。在手訂單持續累積。而且對 2027 年非常重要的是，我認為爬坡（ramps）將會開始攀升的時間點。

So we like the momentum. And I think that -- we're not going to give you a Q4 number, but I'll tell you that the momentum that we have into Q3 continues into Q4. And then with respect to '27, I think the strength of the business and the strength that we have into 2026 can only translate to a strong 2027.

所以我們喜歡這個動能。我想——我們不會給你第四季的數字，但我可以告訴你，我們進入第三季的動能會延續到第四季。至於 2027 年，我認為業務的強勁，以及我們在 2026 年的強勢，只會轉化為一個強勁的 2027 年。

And things have improved, continue to improve, and I think our confidence in our long-term growth rate has only increased, and the portfolio behind this is even better.

而且情況已經改善、並持續改善；我認為我們對長期成長率的信心只增不減，而支撐這一點的產品組合也更好了。

James Schneider, Goldman Sachs.

James Schneider，高盛。

I was wondering if you could maybe, following up on the prior question, just sort of reiterate the double-digit growth outlook you expressed last quarter on the call in terms of 2027 and specifically the 60% gross margin level that you expressed. Anything that changes your confidence there, or has your confidence in fact increased on those targets for next year?

我想追問前一個問題，你是否可以重申一下你們在上季電話會議中對 2027 年所表達的雙位數成長展望，特別是你提到的 60% 毛利率水準。有任何因素會改變你對此的信心嗎？或者你對明年這些目標的信心其實提高了？

Hey, James, this is Bill. There's no change to what we previously said. As you hear, we continue to see things improve on the revenue side, so that's all intact. Related to gross margins, again, we feel very confident of hitting what we shared using a rule of thumb. You could see that play out in our results from a year-over-year perspective. We actually did a bit better. So everything is intact, linked to the higher revenues we plan for, and it scales very nicely into our model.

嗨，James，我是 Bill。我們先前所說的沒有任何改變。如你所聽到的，我們持續看到營收端在改善，所以這部分都維持不變。關於毛利率，我們同樣非常有信心，能用我們分享的經驗法則（rule of thumb）達成目標。你可以從我們年對年的結果看到這點。我們實際上還做得更好一些。所以一切都維持不變，並且與我們規劃的更高營收相連結，而它在我們的模型中也能非常漂亮地擴張。

James, let me add, with respect to '27, I think the right way to think about 2027 is the way we think about it and we're excited about the strength that we have in 2026 is the runway that we have ahead, the opportunity that we have ahead, and our ability to compound into it. I think I mentioned it before. One of the perspectives that we take is we have the strength, and our newest products in Automotive have not even launched.

James，我補充一下，關於 2027 年，我認為正確的思考方式是：我們看待 2027 年的方式，是我們對 2026 年的強勁感到興奮，因為那代表我們前方的跑道、前方的機會，以及我們把它複利化（compound）成長的能力。我之前提過。我們採取的一個觀點是：我們具備這個強度，而我們在汽車領域最新的產品甚至都還沒上市。

And Physical AI is in the early stages of basically design-ins, and the heavy deployments have not even started. And so, I think 2027, right now, we're looking at it in a very constructive manner, and it really underpins the long-term growth rates that we have established.

而實體 AI（Physical AI）仍處於早期的設計導入（design-ins）階段，大規模部署甚至都還沒開始。因此，我認為就 2027 年而言，我們目前是以非常建設性的方式來看待它，而這也確實支撐了我們所建立的長期成長率。

Yeah, maybe I'll just build on that because of what Rafael said about Physical AI. I know he shared about the 15% enablement more than doubling. There's another metric we do track. As he mentioned, early innings. As you all know, we acquired an asset called Kinara, and the design-win funnel that we shared with you, the engagement, the excitement, the amount of input we're getting from our customers, that grew last quarter to over $1 billion.

是的，也許我就接著拉斐爾（Rafael）剛才談到的 Physical AI 再補充一下。我知道他分享過 15% 的賦能（enablement）提升已經不只翻倍。我們還有另一個會追蹤的指標。如他所提到的，現在還在很早期的階段。各位都知道，我們收購了一項名為 Kinara 的資產，而我們先前與各位分享的 design-win 漏斗、客戶互動、興奮程度，以及我們從客戶那裡收到的投入量，在上一季成長到超過 10 億美元。

I would say this quarter it's sitting at $1.5 billion in the funnel. And obviously, we'll have to convert those into design wins, but that's a leading indicator, again, of the early innings stage for Physical AI, and we're excited about it.

我會說這一季它在漏斗中是 15 億美元。當然，我們必須把這些轉換成 design wins，但這再次是一個領先指標，顯示 Physical AI 仍處於早期階段，而我們對此感到很振奮。

Just to add to that, Bill, we have that $1.5 billion pipeline. It represents over 200 unique and distinct customers, so it's very broad-based, James.

再補充一下，Bill，我們有那條 15 億美元的 pipeline。它代表超過 200 個獨特且不同的客戶，所以覆蓋面非常廣，James。

Thanks. And then as just a quick follow-up, can you maybe comment on what channel inventories did in the quarter, what you're expecting for next quarter in terms of weeks, et cetera?

謝謝。接著快速追問一下，你能否評論一下本季通路庫存的變化，以及你們對下季以週數等指標的預期？

Yeah, no. Like we said in the past, we want to run in that target with 11 weeks. It was 11 weeks last quarter, so we feel good about it and make sure we service our customers and get our fair share of the market share.

是的，不。就像我們過去說的，我們希望把它維持在目標的 11 週。上一季是 11 週，所以我們對此感覺良好，並確保我們能服務客戶、取得我們應得的市佔份額。

Joshua Buchalter, TD Cowen.

Joshua Buchalter，TD Cowen。

Maybe following up on Francois' question from earlier. You've had a couple of your peers very clearly call out restocking in the Auto market. You guys, during this past cycle, I think for good reason, were conservative with inventory on your books and in the channel. I guess, is there any reason, anything about your portfolio that makes it like you would see restocking later, or is there any conservatism on your part that's driving the comment about not seeing restocking?

也許延續先前 Francois 的問題。你們有幾家同業非常明確地提到汽車市場正在補庫存（restocking）。你們在這一輪循環中，我認為是有充分理由的，在帳上與通路端的庫存都相對保守。我想問的是，你們的產品組合是否有任何原因會讓你們較晚才看到補庫存？或者你們是否因為自身的保守態度，才會說沒有看到補庫存？

Yeah, Joshua. Hey, this is Bill. I think one of the unique things about NXP is our company-specific growth drivers where Rafael shared in his prepared remarks that 47% is coming from this content that is typically, I would say, 3 times larger than our core. And so, we are in a different area. We play in a different area of Auto, and we expect that 47% to grow towards 50% next year.

是的，Joshua。嗨，我是 Bill。我認為 NXP 的一個獨特之處在於我們公司特有的成長驅動因素；拉斐爾在事先準備的發言中分享，47% 來自這些內容（content），而這些內容通常我會說是我們核心的 3 倍大。因此，我們所處的領域不同。我們在汽車領域切入的位置不同，而且我們預期這 47% 明年會成長到接近 50%。

Related to restocking, again, we have a very good handle on distribution. We know exactly what's going in and what's going out that's serving more broader customers. And specifically in Auto, majority of our Asia customers go through the channel, but in the Western Tier 1s, we track all that and we triangulate it. And as you all know, the Tier 1, the working capital needs are quite tight.

關於補庫存，再次強調，我們對經銷通路（distribution）掌握得非常清楚。我們確切知道哪些貨進來、哪些貨出去，以服務更廣泛的客戶群。特別是在汽車方面，我們在亞洲的大多數客戶是透過通路，但在西方的 Tier 1，我們會追蹤所有這些數據並進行交叉驗證。而且各位都知道，Tier 1 的營運資金需求相當吃緊。

We still see late orders coming in and hand-to-mouth, and margins for them are not that healthy. And so, they know we have some inventory. You see it on our balance sheet. And they are still providing late orders to us. So we monitor this very carefully, but we have not seen the restocking effect specifically with our Western Tier 1s.

我們仍然看到臨時下單（late orders）以及「手到口」（hand-to-mouth）的採購方式，而且他們的利潤率並不那麼健康。因此，他們知道我們有一些庫存。你們也能在我們的資產負債表上看到。而他們仍然對我們提出臨時訂單。所以我們非常謹慎地監控這件事，但我們在西方 Tier 1 客戶身上，尚未看到明顯的補庫存效應。

Okay, thank you. Appreciate that. And then I was hoping to ask about gross margins in the second half. I think on a previous call you talked about utilization rates going from the low 80s to the mid-80s. Is that still the right metric to think about as we continue in this upcycle and as we think about the third-quarter gross margin guide?

好的，謝謝。感謝你的說明。接著我想問下半年毛利率。我記得在之前的電話會議上，你們談到產能利用率會從 80% 出頭提升到 80% 中段。在我們持續處於這一輪上行週期、以及思考第三季毛利率指引時，這仍然是合適的衡量指標嗎？

Last quarter, you called out some higher wafer access fees that could potentially impact you. Did those play any role in the quarter or the guide?

上一季你們提到一些較高的晶圓取得費（wafer access fees）可能會對你們造成影響。這些在本季或指引中有扮演任何角色嗎？

Sure. Let me first address that. Obviously, our gross margins are doing quite well. From Q1 year over year, they're up 150 basis points. Last year, they're going up another 150 basis points. And we plan to bring them into the mid-80s, so I'd like to confirm that. And that does help the second half gross margins continue to improve.

當然。我先回應這一點。很明顯，我們的毛利率表現相當不錯。以 Q1 來看，年對年提升了 150 個基點。相較去年，又再提升 150 個基點。而我們計畫把（產能利用率）提升到 80% 中段，所以我想確認這一點。這也確實有助於下半年毛利率持續改善。

And related to increased costs from foundries and access fees, we haven't seen that play out yet. That's something more I would think comes to us maybe in Q4, but more in 2027 when we enter into a new agreement with our foundry partners.

至於來自晶圓代工廠的成本上升與取得費，我們目前還沒有看到它實際發生。我認為那可能會在 Q4 才開始影響我們，但更可能是在 2027 年、當我們與代工夥伴進入新的協議時。

And, Joshua, what we said last quarter was we did see inflationary input costs on back-end type of things, so piece parts, substrates, precious metals, things like that. But in terms of wafers on the front-end, we operate within kind of a boundary condition, an envelope. As long as we operate within that envelope agreed to with our partners, we don't see tactical price increases. If we go outside of that envelope, yes, we would see price increases and we would then have to pass that along to our customers.

另外，Joshua，我們上一季所說的是，我們確實看到後段（back-end）相關項目的投入成本有通膨壓力，例如零組件、基板、貴金屬等。但就前段（front-end）的晶圓而言，我們是在某種邊界條件、也就是一個範圍（envelope）內運作。只要我們在與夥伴約定的那個範圍內運作，就不會看到戰術性的漲價。如果我們超出那個範圍，是的，我們就會看到漲價，然後我們必須把它轉嫁給客戶。

Tom O'Malley, Barclays.

Tom O'Malley，Barclays。

Rafael, I wanted to dive back into the Kinara commentary. So the funnel is expanding, going to $1.5 billion. You've seen in the industry a lot of acquisitions taking place, both Synaptics and Halo. So do you think that these acquisitions are going after that same area that you guys have already targeted? And then when you look at the TAM and how big that can be, maybe could you try to size what that market looks like in a couple of years just because the funnel would indicate it's a pretty large opportunity.

拉斐爾，我想再深入回到 Kinara 的評論。所以漏斗正在擴大，達到 15 億美元。你也看到產業內有很多併購在發生，包括 Synaptics 和 Halo。你認為這些併購是在追逐你們已經鎖定的同一個領域嗎？另外，當你看 TAM 以及它可能有多大時，能否嘗試估算一下這個市場在幾年後大概會是什麼規模？因為漏斗顯示這是一個相當大的機會。

Indeed. I think you pointed to basically the acquisitions that I think you mentioned. It's just a confirmation of our strategy. The cognification of the edge is happening, and I think you are not able to play without having a strong AI platform and a strong AI roadmap. And so, yes, I think that's exactly the case. I think we do believe that the platform that we have and the asset that we acquired, now incorporated into the NXP roadmap, is best-in-class.

確實如此。我想你指出的基本上就是你提到的那些併購案。這只是對我們策略的一種確認。邊緣端的「認知化」（cognification）正在發生，而我認為如果沒有強大的 AI 平台與強大的 AI 路線圖，你就無法參與競爭。所以，是的，我認為情況正是如此。我們確實相信，我們所擁有的平台以及我們收購的資產，現在已經納入 NXP 的路線圖，是同級最佳（best-in-class）。

We have incorporated now that idea of AI into monolithic integration into our i.MX processors and our S32N platform in Automotive. We are discussing how to actually engage with customers on a discrete NPU which attaches to other platforms, even non-NXP. And not only that, we're developing a very complex AI software framework, which includes now Agentic AI, which is going to be the way that the edge becomes completely autonomous.

我們現在已把 AI 的概念納入我們 i.MX 處理器與汽車用 S32N 平台的單晶整合（monolithic integration）。我們也在討論如何以一顆可分離的 NPU 與客戶互動，該 NPU 可附加到其他平台上，甚至是非 NXP 的平台。不僅如此，我們正在開發一個非常複雜的 AI 軟體框架，現在也包含 Agentic AI；這將會是讓邊緣端變得完全自主的方式。

So now, while other companies are acquiring assets and trying to integrate them, we are trying to evolve into what is going to be the next phase of AI, which we strongly believe is going to be the deployment of AI in Agentic AI.

因此，當其他公司正在收購資產並嘗試整合時，我們則是在演進到 AI 的下一個階段；我們堅信那將是以 Agentic AI 形式部署 AI。

And then on the Auto side, I know that you're saying that you're not seeing Tier 1s bringing back up inventory, but I know that around the last quarter, there was a great variation among your customers where some were well below the standard channel number of weeks and then some were well above. Have you at least seen some normalization there where there's been some standardization around that 11-week mark on both the end customer and maybe at Tier 1? You can comment on the metrics as you will, or do you still see this big disparity where some people really aren't getting it? I just want to see if there's still some normalization yet to come. That's the genesis of the question.

接著在汽車端，我知道你說你沒有看到 Tier 1 把庫存補回來，但我也知道大約在上一季，你的客戶之間差異很大：有些遠低於通路標準的週數，有些則遠高於標準。你是否至少看到那邊有一些正常化，也就是在終端客戶端、以及可能在 Tier 1 端，都更趨近於那個 11 週的水位、並有一些標準化？你可以依你想要的方式評論這些指標，或者你仍然看到這種很大的落差，有些人真的還是沒有跟上？我只是想確認是否還有一些正常化尚未到來。這就是這個問題的起源。

Yeah, hey, Thomas. It's similar to what we've seen in the previous quarters. There is a dislocation between low and high, there's a mix for whatever reason, how they want to control their own working capital and so forth. So no change there. But we track this very carefully.

是的，嗨，Thomas。這跟我們前幾季看到的情況類似。低端與高端之間確實存在錯位；基於各種原因，他們想如何控管自己的營運資金等等，造成了這種組合。所以那邊沒有變化。但我們非常仔細地追蹤這些。

We're just basically -- finally, I think in Q4, Q1, we were finally shipping to real end demand, inventory digestion is behind us, but we have not seen any pull forwards or restocking efforts specifically with their Tier 1s in the Western world.

我們基本上——終於，我想在 Q4、Q1，我們終於是在依照真實的終端需求出貨，庫存去化已經在我們身後了，但我們在西方世界並沒有看到任何提前拉貨或補庫存的動作，特別是在他們的 Tier 1 端。

Tore Svanberg, Stifel. Please go ahead.

Tore Svanberg，Stifel。請繼續。

Yes, thank you, and congratulations on the record revenue. Rafael, I wanted to go back to the Physical AI pipeline, the $1.5 billion. How broad-based is that? Is this a few verticals? Is it many different applications? And which sort of applications should we assume you're going to ramp the earliest?

是的，謝謝，並恭喜創下營收新高。Rafael，我想回到 Physical AI 的管線，那 15 億美元。這有多廣泛？是少數幾個垂直領域嗎？還是很多不同的應用？另外，哪些類型的應用我們應該假設會最早開始放量？

Well, the excitement is that there is actually broad-based interest in deployment of Physical AI or Edge AI. And it's happening whether it's, think about HMI applications, basically where a human controls the device via voice commands and replaces whether mechanical buttons or display buttons. And the support is there's all sorts of use cases around that. There's plenty of use cases around visual and vision. There's plenty of cases of predictive maintenance.

嗯，令人振奮的是，市場對於部署 Physical AI 或 Edge AI 的興趣其實是相當廣泛的。而且它正在發生——例如想想 HMI 應用，基本上是人透過語音指令來控制裝置，取代機械按鍵或螢幕按鍵。而且圍繞這點有各式各樣的使用情境。在視覺與影像辨識方面也有很多使用情境。在預測性維護方面也有很多案例。

The deployment of Physical AI is really broad-based. I think we see a very strong interest in Industrial. We see a strong interest also in Automotive for in-cabin type of applications that we're getting pulled into. And so, I think it's really going to be such a massive opportunity, massive, I would say, transformation that is happening at the Edge. And it's going to be Physical AI, we've determined, is the driver for content growth in Industrial, and it's also an additional content driver for Automotive moving forward.

Physical AI 的部署確實非常廣泛。我想我們看到工業領域的興趣非常強。我們也看到汽車領域的興趣很強，特別是車艙內（in-cabin）類型的應用，我們也被拉進去參與。所以我認為這將會是一個非常巨大的機會——非常巨大——我會說，邊緣端正在發生一場轉型。而我們判斷 Physical AI 將是推動工業內容成長的驅動力，同時也會成為未來汽車領域額外的內容驅動因素。

And then, as my follow-up, if we think about the current environment, it's a bit strange because you're seeing obviously great bookings momentum, a lot of demand, and on the other hand, we have these obviously capacity constraints with other components and so on and so forth. So I'm just wondering, are you seeing any of that potentially impacting some of your customers' demand, meaning they want to do more, but they can't because there are shortages of other components?

接著作為追問，如果我們看目前的環境，感覺有點奇怪，因為你顯然看到很強的訂單動能、很多需求；但另一方面，我們又有其他元件等等的產能限制。所以我想問的是，你是否看到這些因素可能正在影響你某些客戶的需求——也就是他們想做更多，但因為其他元件短缺而做不到？

Well, I think that, let me answer the question with a specific -- I mean, you can see our Q3 guide has Mobile already down year on year, even though it grew sequentially, and our position with respect to market share has not changed. It's already seeing an impact, right? Year on year, Mobile is down.

嗯，我想——讓我用一個具體的例子來回答——你可以看到我們 Q3 的指引，Mobile 已經是年對年下滑，儘管它是季對季成長，而我們在市占率方面的定位並沒有改變。它已經開始受到影響了，對吧？年對年來看，Mobile 是下滑的。

And so, I would say that you're starting to see it. You saw it in Mobile. Now, the concern about memory, especially memory and the constraints that right now the market is -- literally everybody is talking about it. Everybody is trying to design around it, whether it's different products, different packages, different types of DDR. And I think we are literally helping our customers to actually go and make sure that we help them manage the constraints they have. So in pockets, we see it.

所以我會說，你開始看到這個現象了。你在 Mobile 看到了。現在，對於記憶體的擔憂——尤其是記憶體，以及目前市場上的限制——基本上每個人都在談。每個人都在嘗試繞開設計限制，不管是不同產品、不同封裝、不同類型的 DDR。而我想我們確實正在協助客戶，確保我們幫他們管理他們所面臨的限制。所以在某些區塊，我們看得到。

Will Stein, Truist Securities.

Will Stein，Truist Securities。

I wanted to ask about the interplay between the backlog that keeps growing and your lead times and your customers' intentions. If your lead times were shorter, would you have been shipping more? In other words, are customers wanting more than what you can deliver now and you're sort of facing some constraints of your own, and that's why the backlog is growing, or is it that customers feel emboldened given the demand signals they see in the market, and so they're just lining up to place orders with longer duration? If you could linger on that for a moment, I think it would help us understand what's going on there.

我想問一下持續成長的積壓訂單（backlog）與你們交期（lead times）以及客戶意圖之間的互動關係。如果你們的交期更短，你們是否就會出貨更多？換句話說，是不是客戶想要的量超過你們目前能交付的量，而你們自身也面臨一些限制，所以 backlog 才會增加；還是說，客戶因為看到市場上的需求訊號而更有信心，因此只是排隊下更長週期的訂單？如果你能在這點上多談一下，我想會幫助我們理解那邊到底發生了什麼。

Hey, Will, this is Bill. Yeah, I mean, I think what you're saying is what we see, right? We do see escalations. You see our inventory coming down. Late orders have been coming in over the last couple quarters, and now they are realizing they have to place orders, specifically in areas where lead times are longer.

嗨，Will，我是 Bill。是的，我的意思是，我想你說的就是我們看到的情況，對吧？我們確實看到加急（escalations）。你也看到我們的庫存正在下降。過去幾季有延遲下單的訂單進來，而現在他們意識到必須下單，特別是在交期較長的領域。

I mean, it's not our whole entire portfolio, but I would say that greater than 16 weeks, a big chunk of our portfolio has extended versus last quarter. And so you've got to place your orders in line. And so we're seeing that play out, but we're still seeing these late orders come in. So it's a combination of both, I would say.

我的意思是，這不是我們整個產品組合都如此，但我會說，交期超過 16 週的部分，我們產品組合中有很大一塊相較上一季已經拉長了。所以你必須把訂單排進隊列。因此我們看到這種情況正在發生，但我們仍然看到這些延遲下單的訂單持續進來。所以我會說兩者兼而有之。

But it's going back to more of a normal type of way of how you should place orders with us in their appropriate lead times. But there are folks that are still trying to place late orders and struggle with it.

但這也正在回到比較正常的狀態——也就是客戶應該依照對應的交期來向我們下單。不過仍然有人試圖延遲下單，並因此遇到困難。

Right. And then one follow-up, if I can. There are some things going on in the world that are normally disruptive to the electronics supply chain, the wars in particular. Have you seen any change in the impact on your supply chain from the renewed activity in Iran or anything else, any other geopolitical developments? Are you seeing any effect of that that's different from what we've seen over the last few months?

了解。然後如果可以的話再追問一個。世界上有一些事情通常會對電子供應鏈造成干擾，特別是戰爭。你們是否看到伊朗近期活動升高或其他任何地緣政治發展，對你們供應鏈的影響有任何改變？你們是否看到任何不同於過去幾個月的影響？

No, I would say it's similar. Obviously, we are getting a direct impact on higher input costs. And like Rafael said, we try to offset those operationally first. And if we can't, we want to protect our gross margins, and unfortunately, we have to pass those on to our customers.

沒有，我會說差不多。顯然，我們直接受到較高投入成本的影響。而如 Rafael 所說，我們會先嘗試用營運方式來抵銷。如果做不到，我們希望保護毛利率，很不幸地，我們就必須把這些成本轉嫁給客戶。

So that's been the mode we start to work in. We started to see this in Q2. We're in the mode of Q3 related to it. Indirectly, I mean, of course, there are probably things that will impact us indirectly, which is more macro. But if you look at the macro indications, both PMI is doing quite well.

所以這就是我們開始運作的模式。我們在 Q2 就開始看到這個情況。我們在 Q3 也處於因應這件事的模式中。間接來說，當然，可能會有一些事情以更宏觀的方式間接影響我們。但如果你看宏觀指標，PMI 的表現都相當不錯。

GDP has ticked up slightly. Next year's GDP, as you know, is sitting at 3.2 versus today's, I believe, 2.5 or 2.6. So, I mean, it's a balance, I would say. And so we see the macro signals, we see our own internal signals, and we just want to make sure that we're there to support our customers and provide value for them.

GDP 已略微上升。如你所知，明年的 GDP 目前在 3.2，相較於今天我認為是 2.5 或 2.6。 所以，我的意思是，我會說這是一種平衡。因此我們看到總體經濟訊號，也看到我們自身的內部訊號，我們只是想確保我們能在那裡支持客戶並為他們提供價值。

And, Lisa, we'll take our last question here today.

Lisa，我們今天在這裡將進行最後一個問題。

The last question will be coming from the line of Chris Caso of Wolfe Research. Please go ahead.

最後一個問題將由 Wolfe Research 的 Chris Caso 提問。請開始。

Yes, thank you. It's just a follow-up question with regard to what you had said on pricing. When pricing was starting to move higher during the last cycle, you guys were very specific about it being neutral to gross margins. Is that the case today?

是的，謝謝。這只是針對你們先前談到定價的追問。在上一個週期中，當定價開始走高時，你們非常明確地表示這對毛利率是中性的。今天也是如此嗎？

And in terms of what you're seeing with regard to pricing, could you level set us with regard to the magnitude? I know you said you're going to have to wait a while to see the magnitude, but you've obviously made some assumption in terms of your third-quarter guidance.

另外，就你們看到的定價情況而言，你能否幫我們校準一下幅度？我知道你說需要等一段時間才能看到幅度，但你們顯然在第三季指引中已做了一些假設。

Yeah, Chris, let me tackle the -- I'll let Bill tackle the gross margin piece, but let me tackle it the way we account for pricing because I want to be clear, right? Pricing is not necessarily what drives our model. The driver is content growth. It is architecture-led content growth per system, and so that's the main driver for our revenue.

好的，Chris，我先來處理——毛利率的部分我會讓 Bill 來回答，但我先從我們如何認列定價的角度來說，因為我想講清楚，對吧？定價不一定是驅動我們模型的因素。驅動因素是內容成長。也就是由架構帶動、每套系統的內容成長，因此這才是我們營收的主要驅動力。

In pricing, with respect to the way right now we're framing it, pricing seems to be a little bit monolithic in the way you're framing the question with respect to just a tactical move to overcome higher input costs. But in reality, pricing is very dynamic, and it's a strategic lever for us. We use pricing sometimes to capture value, sometimes we use it to increase market share, and in some cases, like the one we discussed right now, to offset input costs.

就定價而言，以我們目前的表述方式，你在問題中把定價框得有點單一，好像只是為了因應更高的投入成本而做的戰術性調整。但實際上，定價非常動態，對我們而言是一個策略槓桿。我們有時用定價來捕捉價值，有時用來提升市占率，而在某些情況下——例如我們現在討論的這個——則用來抵消投入成本。

But we do price adjustments every quarter. And every quarter we provide a next-quarter guide that has an estimate of that. And so I don't think this time is any different. So Bill if you want --

但我們每一季都會做價格調整。而且每一季我們都會提供下一季指引，其中包含對此的估計。所以我不認為這次有任何不同。所以 Bill，如果你想——--

Yeah. I would just add to what Rafael said is on pricing, where typically, we give an update once a year. In the beginning of this year, we said we'd be down in the low-single digits. At the end of this year, we'll update that. We'll probably be a bit better, I would say, because again, with the selective pricing we're doing because of the higher input costs.

是的。我只想補充 Rafael 關於定價的說法：通常我們一年更新一次。在今年年初，我們說會在低個位數的下滑。到今年年底，我們會更新這個數字。我會說可能會好一些，因為同樣地，鑑於投入成本上升，我們正在做選擇性的定價。

And the related gross margins, obviously we want to make sure the value we capture, we pass that on to the owners. And obviously if there's one that we can't offset and they're higher input costs like inflation, we have to make sure that, unfortunately, we have to pass that to our customers. And you see that play out throughout the entire supply chain.

至於相關的毛利率，顯然我們希望確保我們所捕捉到的價值能回饋給股東。而且很明顯，如果有些成本我們無法抵消，而投入成本又上升（例如通膨），我們就必須確保——很遺憾地——必須把這些轉嫁給客戶。你會看到這在整個供應鏈中發生。

But I think we've been very disciplined here. It is a bit different from COVID, which was more broad based and was a supply issue. This is more of an inflationary issue, I would say, at this time of where we are.

但我認為我們在這方面一直非常自律。這與 COVID 時期有些不同，當時影響更廣泛，且是供給面的問題。我會說，就我們目前所處的階段而言，這更像是一個通膨問題。

Got it. And as a follow-up, if you could go into a little more detail. You talked a bit about the AI-enabled processors. How does that value come to NXP? Is it a form of higher content, higher ASPs for the products? Is it driving unit growth? Is it market share? Perhaps a combination of all those?

了解。再追問一下，如果你能再多談一些細節。你提到了一些 AI 賦能的處理器。這個價值如何回到 NXP？是以更高的內容量、產品更高的 ASP（平均售價）呈現嗎？是在推動出貨量成長嗎？還是市占率？或許是以上的組合？

I think you answered the question, but I'll just rephrase what you said. It starts with the products themselves, they have more content, right? So Physical AI will drive content from a product perspective. They tend to be higher-performance processors, more AI inference content, more software content, and more enablement.

我想你已經回答了這個問題，但我只是把你說的再換個說法。首先從產品本身開始，它們有更多的內容，對吧？所以 Physical AI 會從產品角度帶動內容成長。它們往往是更高效能的處理器，具備更多 AI 推論內容、更多軟體內容，以及更多使能（enablement）。

Then the system itself becomes also more complex. You have more connectivity, because you have AI, you have more security, and in the case of Physical AI and robotics, you must have functional safety. And then you go in the evolution of what's going to happen right now with respect to Agentic AI. Now you need a little bit more of a software framework associated with that.

接著，系統本身也會變得更複雜。因為有 AI，你需要更多連線能力，也需要更多安全性；而在 Physical AI 與機器人領域，你必須具備功能安全。然後你再看接下來在 Agentic AI 方面將會發生的演進。此時你還需要一個與之相關的、更多的軟體框架。

So in reality, Physical AI for us is a very important driver of content growth, and we intend to position our roadmap to lead in this market.

所以實際上，Physical AI 對我們而言是內容成長非常重要的驅動力，而我們打算把產品藍圖（roadmap）定位為在這個市場中領先。

Thank you. And that concludes the Q&A session for today. I would like to turn the call back over to Rafael, CEO, for closing remarks. Please go ahead.

謝謝。今天的問答環節到此結束。我想把電話交回給執行長 Rafael 作結語。請開始。

Thank you, everyone, for joining us and for your thoughtful questions. I want to leave you with three thoughts. First, our growth is structural driven by Software-Defined Vehicles, Physical AI at the industrial Edge, and a nascent Data Center franchise. Second, we're entering a decade-long adoption of Physical AI, which is transforming industries. Through relentless innovation and customer intimacy, NXP is best positioned to lead this transformation.

謝謝各位參與，也謝謝你們提出深思熟慮的問題。我想留給大家三點想法。第一，我們的成長是結構性的，受軟體定義車輛（Software-Defined Vehicles）、工業邊緣端（industrial Edge）的 Physical AI，以及一個新興的資料中心事業所驅動。第二，我們正進入 Physical AI 長達十年的採用期，它正在改變各個產業。透過不懈的創新與貼近客戶，NXP 最有條件引領這場轉型。

Third, our financial model is scaling exactly as designed. Margin expansion is structural, capital allocation is disciplined, and we are positioned to deliver expanding profitability and growing returns for years to come. The long-term opportunity for NXP has never been clearer. Thank you.

第三，我們的財務模型正如設計般精準地擴張。毛利率擴張是結構性的，資本配置是有紀律的，而我們已就位，將在未來多年帶來持續擴大的獲利能力與不斷成長的報酬。NXP 的長期機會從未如此清晰。謝謝。

Thank you so much for joining. You may now disconnect.

非常感謝各位參與。你們現在可以掛線。