格芯 (GFS) 2025 Q3 法說會逐字稿

Thank you for standing by and welcome to the GlobalFoundries' Inc third quarter of fiscal 2025 financial results conference call. (Operator Instructions) As a reminder, today's program is being recorded.

感謝您的耐心等待，歡迎參加 GlobalFoundries 公司 2025 財年第三季財務業績電話會議。（操作說明）提醒各位，今天的節目正在錄製中。

And now I'd like to introduce your host for today's program, Eric Chow, Investor Relations. Please go ahead, sir.

現在，我謹向大家介紹今天節目的主持人，投資者關係部的周先生。請繼續，先生。

Thank you, operator. Good morning, everyone, and welcome to GlobalFoundries third quarter 2025 earnings call. On the call with me today are Tim Breen, CEO; Neil Anderskouv, President and Chief Operating Officer; and Sam Franklin, Interim CFO.

謝謝接線生。各位早安，歡迎參加 GlobalFoundries 2025 年第三季財報電話會議。今天和我一起通話的有：執行長 Tim Breen；總裁兼營運長 Neil Anderskouv；以及臨時財務長 Sam Franklin。

A short while ago, we released GF's third quarter financial results, which are available on our website at investors.gf.com, along with today's accompanying slide presentation. This call is being recorded, and a replay will be made available on our Investor Relations web page. During this call, we will present both IFRS and non-IFRS financial measures. The most directly comparable IFRS measures and reconciliations for non-IFRS measures are available in today's press release and accompanying slides.

不久前，我們發布了GF第三季財務業績，您可以在我們的網站 investors.gf.com 上查看，以及今天隨附的幻燈片簡報。本次電話會議正在錄音，錄音回放將在我們的投資者關係網頁上提供。在本次電話會議中，我們將介紹 IFRS 和非 IFRS 財務指標。最直接可比較的 IFRS 指標和非 IFRS 指標的調節表可在今天的新聞稿和隨附的幻燈片中找到。

Please note that these financial results are unaudited and subject to change. Certain statements on today's call may be deemed to be forward-looking statements. Such statements can be identified by terms such as believe, expect, intend, anticipate and may or by the use of future tense. You should not place undue reliance on forward-looking statements. Actual results may differ materially from these forward-looking statements, and we do not undertake any obligation to update any forward-looking statements we make today.

請注意，這些財務數據未經審計，可能會有所變動。今天電話會議上的某些陳述可能被視為前瞻性陳述。此類陳述可以透過諸如相信、期望、打算、預料和可能之類的詞語或使用將來時態來識別。您不應過度依賴前瞻性陳述。實際結果可能與這些前瞻性陳述有重大差異，我們不承擔更新今天所作任何前瞻性陳述的義務。

For more information about factors that may cause actual results to differ materially from forward-looking statements, please refer to the press release we issued today as well as risks and uncertainties described in our SEC filings, including in sections under the caption Risk Factors in our annual report on Form 20-F and in any current reports on Form 6-K furnished with the SEC.

有關可能導致實際結果與前瞻性聲明存在重大差異的因素的更多信息，請參閱我們今天發布的新聞稿以及我們在提交給美國證券交易委員會的文件中描述的風險和不確定性，包括我們在 20-F 表格年度報告中“風險因素”標題下的章節以及我們向美國證券交易委員會提供的任何 6-K 表格當前報告中的章節。

In terms of upcoming events, we will be participating in a fireside chat at the UBS Global Technology and AI Conference in Scottsdale on December 2. In addition, we are looking forward to hosting a public webcast investor webinar at 10:00 AM Eastern Time on December 3. At this event, we will provide a business, technical and strategy update on the opportunities for GF across the rapidly evolving physical AI market.

在即將舉行的活動中，我們將於 12 月 2 日在斯科茨代爾舉行的瑞銀全球科技與人工智慧大會上參加爐邊談話。此外，我們期待於12月3日美國東部時間上午10:00舉辦一場公開網路直播投資者網路研討會。本次活動我們將介紹GF在快速發展的實體人工智慧市場中的機遇，並提供相關的業務、技術和策略方面的最新資訊。

We will begin today's call with Tim providing a summary update on the current business environment and technologies. Neils will then discuss our recent design wins, highlights and traction across the end markets, after which Sam will provide details on our third quarter results and fourth quarter 2025 guidance. We will then open the call for questions with Tim, Neils and Sam.

今天，我們將首先由 Tim 為大家簡要介紹當前的商業環境和技術情況。接下來，Neils 將討論我們最近在終端市場上的設計成果、亮點和進展，之後 Sam 將詳細介紹我們第三季的業績和 2025 年第四季的業績指引。接下來，我們將開放提問環節，屆時 Tim、Neils 和 Sam 將接受提問。

We request that you please limit your questions to one with one follow-up. I'll now turn the call over to Tim.

請您將問題限制在一個問題內，並可提出後續問題。現在我將把電話交給提姆。

Thank you, Eric, and welcome, everyone, to our third quarter 2025 earnings call. Before I begin, I wanted to express my sincere gratitude to John for his service and contributions to GF. We wish him the best. GF delivered a strong third quarter with revenue, gross margin, operating margin and earnings per share at the high end of the guidance ranges. For the fourth consecutive quarter, we saw strong double-digit percentage year-over-year revenue growth, both in our automotive and communications, infrastructure and data center end markets, which together represented 28% of our total third quarter revenue.

謝謝艾瑞克，也歡迎各位參加我們2025年第三季財報電話會議。在開始之前，我想對約翰為GF所做的貢獻表示衷心的感謝。我們祝他一切順利。GF第三季業績強勁，營收、毛利率、營業利潤率和每股盈餘均達到預期範圍的高端。連續第四個季度，我們在汽車、通訊、基礎設施和資料中心終端市場均實現了強勁的兩位數同比增長，這三個市場合計占我們第三季度總收入的 28%。

We expanded third quarter gross margin both sequentially and year-over-year, which is representative of our relentless drive to growing profitability. With the strength of our differentiated product portfolio, which is highly suited to secular growth markets, the richer mix of high-growth businesses and the clear value proposition of our global footprint, GF is laying a strong foundation for a future of robust profitable growth.

第三季毛利率環比和同比均有所增長，這體現了我們為提高盈利能力而不懈努力的精神。憑藉我們差異化的產品組合（非常適合長期成長市場）、更豐富的高成長業務組合以及我們全球佈局的明確價值主張，GF 正在為未來強勁的獲利成長奠定堅實的基礎。

GF is truly a global company. I recently had the privilege of visiting customers and employees across the U.S., Asia and Europe, including at our marquee Global Technology Summits in all three continents. Having met with over 100 current and prospective customers from across the end markets we serve, the feedback has been consistent and unequivocal. GF brings a unique combination of differentiated technologies that meet the needs of today's secular trends, including the scaling of AI in the data center and the proliferation of AI into the physical world as well as the need to deliver those technologies from a resilient global footprint.

GF 真是一家全球性公司。我最近有幸走訪了美國、亞洲和歐洲各地的客戶和員工，包括參加了我們在三大洲舉辦的全球科技高峰會。我們與來自我們所服務的終端市場的 100 多位現有客戶和潛在客戶進行了會面，得到的回饋一致且明確。GF 帶來了一系列獨特的差異化技術，滿足了當今長期發展趨勢的需求，包括資料中心人工智慧的擴展、人工智慧在物理世界的普及，以及從具有彈性的全球佈局交付這些技術的需求。

Let me address each of these exciting areas. Firstly, scaling AI in the data center with optical networking. After years of R&D, capacity investments and deep innovation with customers, GF is carving out a strong position in the optical market at exactly the right time. Recent commentary by hyperscalers, GPU makers and other players in the data center ecosystem have emphasized the need for silicon photonics in scale up, scale out and scale across networking.

讓我來逐一介紹這些令人興奮的領域。首先，利用光網路在資料中心擴展人工智慧。經過多年的研發、產能投資以及與客戶的深度創新，GF 正值最佳時機在光學市場佔據強勢地位。最近，超大規模資料中心營運商、GPU 製造商和資料中心生態系統中的其他參與者都強調了矽光子技術在網路擴展、橫向擴展和縱向擴展方面的必要性。

The OCP Global Summit last month highlighted a growing shift towards pluggable silicon photonics and co-packaged optics as alternatives to traditional copper interconnects over the next several years, where legacy technology is simply unable to meet the increasing demands in data transmission speed, bandwidth density and power efficiency.

上個月的 OCP 全球高峰會強調，在未來幾年內，可插拔矽光子學和共封裝光學元件將逐漸取代傳統的銅互連技術，因為傳統技術根本無法滿足日益增長的數據傳輸速度、頻寬密度和功率效率需求。

Propelled by this expected transition, we estimate our serviceable addressable market for optical networking will grow by a CAGR of approximately 40% through 2030. We expect GF to be a key participant in this substantial growth and are highly encouraged by our early track record of success in many applications that support optical networking including our silicon photonics platform as well as our high-performance silicon germanium and FDX technologies.

受此預期轉變的推動，我們估計到 2030 年，我們可服務的光網路市場將以約 40% 的複合年增長率成長。我們期望GF能夠成為這一巨大成長的關鍵參與者，並且我們對公司在支援光網路的眾多應用領域（包括我們的矽光子平台以及高性能矽鍺和FDX技術）早期取得的成功記錄感到非常鼓舞。

In Q3 alone, we won three optical networking designs with new customers worth over $150 million of projected lifetime revenue with the first tape-out for one of these designs already completed in the quarter. Silicon photonics alone is on track to reach over $200 million of revenue in 2025, close to doubling year-over-year. As the market continues to require higher and higher performing pluggable optical transceivers and as co-packaged optics adoption meaningfully ramps from 2027, we envision silicon photonics to become a $1 billion-plus run rate business for GF before the end of the decade.

僅在第三季度，我們就贏得了三個光網路設計項目，新客戶預計終身收入超過 1.5 億美元，其中一個設計的首個流片項目已在本季度完成。僅矽光子技術一項，預計到 2025 年收入將超過 2 億美元，接近每年翻倍。隨著市場對性能越來越高的可插拔光收發器的需求不斷增長，以及從 2027 年開始共封裝光器件的採用率顯著提高，我們預計到本十年末，矽光子學將成為 GF 的一項年收入超過 10 億美元的業務。

To support this growth, we will continue to partner with our customers and make the necessary investments to grow our scale as well as adding, organically or inorganically, new complementary capabilities. With gross margins significantly above our target model, we expect long-term growth in silicon photonics to provide a tailwind to GF for years to come.

為了支持這一成長，我們將繼續與客戶合作，並進行必要的投資以擴大規模，同時透過有機成長或無機成長的方式增加新的互補能力。由於毛利率遠高於我們的目標模型，我們預計矽光子技術的長期成長將在未來幾年為GF帶來利多。

The second significant and rapidly evolving secular trend is the advent of AI capabilities being deployed across a broad range of applications in the physical world. Based on discussions with our customers, we believe the ongoing data center AI build-out is merely a prelude to the next step of the AI revolution, real-world applications in the physical space.

第二個重要且快速發展的長期趨勢是人工智慧能力在物理世界的各種應用中廣泛部署。根據與客戶的討論，我們認為正在進行的資料中心人工智慧建構只是人工智慧革命下一步——在實體空間中進行實際應用——的前奏。

From autonomous vehicles and drones to next-generation medical devices and ultimately humanoid robots, we expect the marriage of artificial intelligence with real-time sensing, control and compute capabilities to unlock new previously unthinkable applications and accelerate demand for GF's essential technologies.

從自動駕駛汽車和無人機到下一代醫療設備，最終到人​​形機器人，我們期待人工智慧與即時感知、控制和運算能力的結合，能夠解鎖以前無法想像的新應用，並加速對GF關鍵技術的需求。

The technical demands of this next phase of AI align with GF's deep technical strength in developing feature-rich technologies that play a critical role across multiple applications, which is further complemented with our recent investment in MIPS, which will accelerate the development of real-time processor IP. In the world of physical AI, the market will need vast amounts of feature-rich, low-power connected chips that are secure and cost effective.

下一階段人工智慧的技術要求與GF在開發功能豐富的技術方面的深厚技術實力相契合，這些技術在多個應用中發揮著關鍵作用，而我們最近對MIPS的投資進一步補充了這一點，這將加速實時處理器IP的開發。在實體人工智慧領域，市場將需要大量功能豐富、低功耗、安全且經濟高效的連網晶片。

We believe everything that moves will become autonomous, everything that sensors will be intelligent and many devices that think will also actuate in the real world. GF's product portfolio enables us to play a critical role in this coming revolution. For efficient power management, our FTX and FinFET platforms are specifically designed to support always-on ultra-low leakage so edge devices can run longer and more reliably.

我們相信，所有會動的東西都將實現自主，所有帶有感測器的東西都將變得智能，許多會思考的設備也將在現實世界中發揮作用。GF的產品組合使我們能夠在即將到來的這場變革中發揮關鍵作用。為了實現高效的電源管理，我們的 FTX 和 FinFET 平台經過專門設計，支援始終開啟的超低漏電，因此邊緣裝置可以運行更長時間，更加可靠。

For robotics and real-world object manipulation, our BCD and BCD HV platforms offer a power-efficient architecture that is ideal for motor and joint control as well as battery management. Lastly, for intelligence, sensing and detection, our recently launched UX platform as well as our established FDX and FinFET capabilities enable accurate multimode sensors with capabilities across radar, ultra-wideband, imaging and audio. By coupling all of these technologies with a range of embedded nonvolatile memory solutions, including ESF, MRAM and RRAM, we can go further to enable smart, secure processing in a range of physical applications.

對於機器人和現實世界的物件操作，我們的 BCD 和 BCD HV 平台提供了節能的架構，非常適合馬達和關節控制以及電池管理。最後，在情報、感測和偵測方面，我們最近推出的 UX 平台以及我們成熟的 FDX 和 FinFET 技術能夠實現精確的多模式感測器，這些感測器具備雷達、超寬頻、成像和音訊等功能。透過將所有這些技術與一系列嵌入式非揮發性記憶體解決方案（包括 ESF、MRAM 和 RRAM）結合，我們可以進一步在各種實體應用中實現智慧、安全的處理。

Across all of these GF-served applications, we believe the emerging physical AI opportunity will become more than an $18 billion SAM for GF by 2030. Our momentum with customers is accelerating in edge and physical AI applications. The proof points are already in motion. And in the third quarter, we secured several additional design wins across applications such as AI-enabled glasses, AI-enabled hearables, AI-enabled home appliances and AI-enabled software-defined vehicles.

我們相信，到 2030 年，在所有這些 GF 服務的應用領域中，新興的實體人工智慧機會將為 GF 帶來超過 180 億美元的市場規模。我們在邊緣和實體人工智慧應用領域與客戶的合作勢頭正在加速成長。證據已經顯現。第三季度，我們在人工智慧眼鏡、人工智慧耳機、人工智慧家用電器和人工智慧軟體定義汽車等應用領域獲得了多個額外的設計訂單。

The last theme that remains top of mind for our customers is the critical importance of geographically diversified semiconductor supply. Recent geopolitical conflicts, tariffs and export controls are a consequence of an increasingly fractured and deglobalizing world.

最後一個令我們的客戶最關注的主題是地域多元化半導體供應的關鍵重要性。近期的地緣政治衝突、關稅和出口管制是日益分裂和去全球化的世界的後果。

As a remedy, governments have sought to encourage industry players to reshore or onshore their sourcing of essential chips. It is now common for customers to require, not request non-China, non-Taiwan supply chains and is now also becoming increasingly common to specifically require U.S.-based manufacturing. As many of our customers have now publicly stated, partnering with GF in reshoring technologies to the U.S. has become core to their supply strategy.

作為補救措施，各國政府都試圖鼓勵產業參與者將關鍵晶片的採購遷回國內。現在，客戶通常會要求（而不是要求）非中國、非台灣的供應鏈，現在也越來越普遍地要求在美國進行生產。正如我們許多客戶現在公開表示的那樣，與GF合作將技術轉移回美國已成為其供應鏈策略的核心。

By aligning our investments to our customers' requirements, we are positioning GF to gain share from this secular trend, given our unique and advantaged global footprint across the U.S., Europe and Asia. In June, with support from half a dozen leading customers, including Apple, AMD, SpaceX, Qualcomm, NXP and several other leading technology companies, we announced that we broadened the envelope of our investments to $16 billion in order to expand U.S. manufacturing and advanced packaging capabilities in our facilities in New York and Vermont.

透過調整投資方向以滿足客戶需求，憑藉我們在美國、歐洲和亞洲獨特且具有優勢的全球佈局，GF 正努力從這一長期趨勢中獲取市場份額。6 月，在包括蘋果、AMD、SpaceX、高通、恩智浦等六家領先客戶的支持下，我們宣布將投資規模擴大至 160 億美元，以擴大我們在紐約和佛蒙特州工廠的美國製造和先進封裝能力。

With support from federal, state and local governments, we have established a world-class semiconductor ecosystem in the U.S., rich with employee talent as well as diverse suppliers, customers and OEMs. Notwithstanding the ongoing Section 232 assessment in the U.S., the structural reshaping of global supply chains is well underway, and we believe that GF is at the forefront of supporting this transformation.

在聯邦、州和地方政府的支持下，我們在美國建立了一個世界級的半導體生態系統，擁有豐富的人才資源以及多元化的供應商、客戶和原始設備製造商。儘管美國仍在進行第 232 條款評估，但全球供應鏈的結構性重塑正在順利進行，我們相信 GF 正處於支持這項變革的最前線。

As our customers increasingly seek to mitigate geopolitical risks and enhance their supply chain resilience, GF is helping them navigate trade complexities and optimize their sourcing decisions. An excellent recent example of the progress we are making is our announcement with Silicon Labs to manufacture its wireless SoCs on GF's new ultra-low power platform out of our Malta, New York fab.

隨著客戶越來越尋求降低地緣政治風險並增強供應鏈韌性，GF 正在幫助他們應對貿易複雜性並優化採購決策。最近我們取得的進展的一個很好的例子是我們宣布與 Silicon Labs 合作，將在我們位於紐約州馬耳他的工廠，利用 GF 的新型超低功耗平台生產其無線 SoC。

Beyond the U.S., we have also announced plans to invest an additional EUR 1.1 billion in our Dresden fab, supported by incentives from the German federal government and the state of Saxony under the framework of the European CHIPS Act, the investment will allow us to increase production capacity to more than 1 million wafers a year in Dresden by the end of 2028, making it the largest site of its kind in Europe, approaching gigafab scale.

除了美國之外，我們還宣布計劃在德勒斯登工廠追加投資 11 億歐元。在德國聯邦政府和薩克森州根據《歐洲晶片法案》提供的激勵措施的支持下，這項投資將使我們能夠在 2028 年底前將德累斯頓的年產能提高到 100 萬片以上晶圓，使其成為歐洲同類工廠中規模最大的工廠，接近千兆晶圓廠的規模。

Driven by the needs of key European customers such as NXP, Infineon, Aumovio and Bosch, we are well placed to meet our customers' requirements of EU-based manufacturing from our world-class site. We believe we are only in the early stages of this opportunity and see strong validation of our decade-long strategy to build and scale flexible manufacturing capabilities across our fabs, an area where GF has always been a leader for the industry and intends to continue to do so well into the future.

在NXP、英飛凌、Aumovio和博世等歐洲主要客戶的需求驅動下，我們憑藉著世界一流的生產基地，完全有能力滿足客戶對歐盟製造的要求。我們相信我們目前僅處於這一機會的早期階段，並且看到了我們十年來在晶圓廠構建和擴展柔性製造能力戰略的有力驗證，GF 一直是該行業的領導者，並打算在未來繼續保持這一地位。

In conclusion, at GF, we are committed to being a trusted partner to our customers, utilizing our differentiated chip technologies and global manufacturing capacity. We believe we are well positioned to benefit from the long-term trends driving our industry. Years of work and preparation have established a solid foundation for us to capture these inflection point opportunities, all made possible by the dedication of our global team.

總之，GF致力於成為客戶值得信賴的合作夥伴，充分利用我們差異化的晶片技術和全球製造能力。我們相信，我們已做好充分準備，能夠從推動我們行業發展的長期趨勢中獲益。多年的努力和準備為我們抓住這些轉捩點機會奠定了堅實的基礎，這一切都離不開我們全球團隊的奉獻。

With that, over to you, Niels.

那麼，接下來就交給你了，尼爾斯。

Thank you, Tim, and welcome to everyone on the call. GF's portfolio of diverse and differentiated solutions are enabling us to win more with our customers and serve the defining secular trends of our time. In the third quarter, we secured nearly 150 new design wins across our end markets, more than 50% growth from the same quarter a year ago.

謝謝你，提姆，也歡迎各位參加電話會議。GF 多樣化且差異化的解決方案組合使我們能夠贏得更多客戶的青睞，並順應我們這個時代的發展趨勢。第三季度，我們在終端市場獲得了近 150 個新的設計訂單，比去年同期成長了 50% 以上。

Over the last four quarters, over 90% of our design wins were awarded on a sole-source basis to GF, a consistent proof point of the depth of our customer partnerships and the value of our essential chip technologies. One example of our strong and expanding portfolio of solutions includes our recent technology agreement with TSMC for 650-volt and 80-volt gallium nitride technology. This strategic move will accelerate GF's next generation of GaN products, allowing us to serve an expanded set of customers across a broader range of power applications in markets such as data center, industrial and automotive.

在過去的四個季度中，超過 90% 的設計訂單都是以獨家採購的方式授予 GF 的，這充分證明了我們與客戶建立的深厚合作關係以及我們核心晶片技術的價值。我們強大且不斷擴展的解決方案組合的一個例子是我們最近與台積電達成的 650 伏特和 80 伏特氮化鎵技術技術協議。這項策略措施將加速GF下一代GaN產品的研發，使我們能夠為資料中心、工業和汽車等市場中更廣泛的電力應用領域的更多客戶提供服務。

GF is well suited to capitalize on this opportunity and serve the U.S. market given our existing 200-millimeter GaN capabilities in Burlington, Vermont. We plan to qualify the licensed GaN technology at our fab in Vermont with full production set to begin in the second half of 2026. We made significant strides in our strategy to diversify the business and accelerate the growth of our highest margin product platforms.

鑑於我們在佛蒙特州伯靈頓現有的 200 毫米 GaN 產能，GF 完全有能力抓住這個機會，服務美國市場。我們計劃在佛蒙特州的工廠對授權的氮化鎵技術進行認證，並於 2026 年下半年開始全面投產。我們在業務多角化策略方面取得了重大進展，並加速了利潤率最高的產品平台的成長。

I'm encouraged about the expansion in the number of end applications we serve, including in exciting areas such as optical networking, satellite communications, quantum computing, software-defined vehicles and smart glasses. Given the importance of differentiated technology, enhanced features and the performance requirements from our customers, these fast-growing markets support accelerating growth and improvements to our product mix supporting margin expansion.

令我感到鼓舞的是，我們服務的終端應用數量不斷擴大，其中包括光網路、衛星通訊、量子運算、軟體定義汽車和智慧眼鏡等令人興奮的領域。鑑於差異化技術的重要性、增強的功能以及客戶對性能的要求，這些快速成長的市場支持加速成長和改善我們的產品組合，從而支持利潤率的擴大。

While we have more room to grow and diversify, our progress is already evident in our business results. We have organically grown our automotive end market more than tenfold in the last five years. It now comprises around a quarter of our wafer revenue, and we expect automotive to approach $1.5 billion of annual revenue in 2025. We have line of sight for automotive to become a multibillion-dollar business for us through the end of the decade.

雖然我們還有更多的發展和多元化空間，但我們的進步已經在業務成果中得到了體現。過去五年，我們的汽車終端市場實現了十倍以上的有機成長。目前，汽車產業的收入約占我們晶圓收入的四分之一，我們預計到 2025 年，汽車產業的年收入將接近 15 億美元。我們有理由相信，到本十年末，汽車業務將成為我們價值數十億美元的產業。

We're very encouraged by the strength of our leading silicon photonics products and see strong double-digit growth as it nearly doubles in revenue in 2025 compared to 2024. The application of our silicon photonics portfolio within our communications infrastructure and data center end market is not only margin accretive today, but accretive to our long-term gross margin objectives. As we expand our capacity to meet demand and as the demand for silicon photonics grows, we expect to benefit from additional mix tailwinds.

我們領先的矽光子產品的強勁表現令我們倍感鼓舞，預計 2025 年營收將比 2024 年翻一番，實現兩位數的強勁成長。我們將矽光子產品組合應用於通訊基礎設施和資料中心終端市場，不僅能提高目前的利潤率，還能提高我們的長期毛利率目標。隨著我們擴大產能以滿足需求，以及對矽光子學的需求不斷增長，我們預計將受益於其他利好因素。

Lastly, we've seen strong momentum for fast-growing satellite communications applications, which we expect to contribute approximately $100 million of revenue in 2025 to our communications infrastructure and data center end market, up from de minimis revenue in 2024. The portion of SATCOM served on our NXS platform is a margin-accretive product, thanks to its differentiated features, cost profile and efficient scale despite having an ASP per wafer lower than our corporate average.

最後，我們看到快速成長的衛星通訊應用勢頭強勁，預計到 2025 年，它將為我們的通訊基礎設施和資料中心終端市場貢獻約 1 億美元的收入，而 2024 年的收入微乎其微。由於其差異化的特性、成本結構和高效的規模，儘管每片晶圓的平均售價低於我們公司的平均水平，但我們 NXS 平台上的 SATCOM 產品仍能帶來利潤成長。

With satellite launches expected to grow 150% and SATCOM subscribers set to double in the next five years, we expect the semiconductor SAM for this opportunity to be over $1 billion through the end of the decade with GF as an anchor supplier. Within the end market we serve, GF is well positioned to capitalize on several key secular inflections, and we are making continued progress towards transforming the mix of our business towards the fastest-growing and most profitable platforms.

預計未來五年衛星發射量將成長 150%，衛星通訊用戶數量將翻一番，我們預計到本十年末，這一機遇的半導體市場規模將超過 10 億美元，而 GF 將成為主要供應商。在我們所服務的終端市場中，GF 已做好充分準備，能夠把握幾個關鍵的長期轉折點，並且我們正在不斷努力，將我們的業務組合轉向增長最快、盈利能力最強的平台。

With that, let me walk you through the key highlights for the quarter by end market. Automotive represented approximately 18% of the quarter's total revenue. In the third quarter, we continued our strong momentum in automotive, winning new design wins with 12 unique customers. Highlighting the breadth and depth of our diverse product portfolio, third quarter design wins and new tape-outs included advanced image sensors, body and chassis MCUs, high-performance audio amplifiers, advanced tire monitoring sensors, Ethernet switches and motor controllers on our FinFET, [FDX 4iESF-3] and BCD high-voltage platforms.

接下來，我將按終端市場為您介紹本季的主要亮點。汽車業約佔本季總收入的 18%。第三季度，我們在汽車領域繼續保持強勁勢頭，贏得了 12 家不同客戶的新設計訂單。第三季設計訂單和新流片成果突顯了我們多元化產品組合的廣度和深度，其中包括基於我們的 FinFET、[FDX 4iESF-3] 和 BCD 高壓平台的先進影像感測器、車身和底盤 MCU、高性能音訊放大器、先進輪胎監測感測器、乙太網路交換器和馬達控制器。

Customers across the value chain continue to choose GF for our performance at the highest auto grade standards and strong long-term partnerships. In Q3, we signed an MOU with Hyundai Motor Group that leverages GF's deep semiconductor expertise to equip next-generation vehicles with smart assistance, increased connectivity and enhanced power efficiency. Smart mobile devices represent approximately 45% of the quarter's total revenue.

價值鏈上的客戶持續選擇GF，因為我們以最高汽車級標準提供卓越的性能，並建立了牢固的長期合作夥伴關係。第三季度，我們與現代汽車集團簽署了一份諒解備忘錄，利用GF深厚的半導體專業知識，為下一代汽車配備智慧輔助、增強連接性和提高能源效率。智慧型行動裝置約佔本季總營收的 45%。

In Q3, we secured our first design win for the newly launched CBIC platform with strong engagement with multiple leading fabless RF companies. Developed and manufactured in our Burlington, Vermont fab, CBIC is our highest performing silicon germanium platform to date and is capable of addressing several key markets, including smartphones, wireless infrastructure, optical networking, satellite communications and industrial IoT.

第三季度，我們與多家領先的無晶圓廠射頻公司密切合作，成功拿下新推出的 CBIC 平台的首個設計訂單。CBIC 是我們迄今為止性能最高的矽鍺平台，由我們在佛蒙特州伯靈頓的工廠開發和製造，能夠滿足智慧型手機、無線基礎設施、光網路、衛星通訊和工業物聯網等多個關鍵市場的需求。

For smartphones, the platform enables low-noise amplifiers that reduce power consumption while maintaining ultra-low noise and reducing battery drain. Also in the third quarter, we secured our first NOR Flash memory design win for mobile with a leading Chinese fabless company to enable next-generation mobile and wearables a decision driven specifically by GF's global footprint and the flexibility it provides to our customers.

對於智慧型手機而言，該平台可實現低雜訊放大器，從而在保持超低雜訊的同時降低功耗並減少電池消耗。第三季度，我們還與一家領先的中國無晶圓廠公司獲得了第一個用於行動裝置的 NOR Flash 記憶體設計訂單，以支援下一代行動裝置和穿戴式裝置。這項決定主要得益於 GF 的全球佈局及其為客戶提供的靈活性。

Lastly, we built upon our momentum with a recent design win for microLED display backplane at Sapien for a project with a leading provider of next-generation smart glasses. Home and industrial IoT represented approximately 15% of the quarter's total revenue. Announced at our Global Technology Summit in Asia, GF partnered with Aegis, a leading player in smart sensors to produce the latest generation of smart sensors on GF's BCD platform in Singapore.

最後，我們乘勢而上，最近在 Sapien 贏得了一個微型 LED 顯示器背板的設計項目，該項目是為一家領先的下一代智慧眼鏡供應商設計的。家庭和工業物聯網約佔該季度總收入的 15%。在亞洲舉行的全球技術高峰會上，GF宣布與智慧感測器領域的領導企業Aegis合作，在新加坡利用GF的BCD平台生產最新一代智慧感測器。

This will enable next-generation application-optimized intelligent sensors with best-in-class size, weight, power and cost advantages. These direct time-of-flight sensors are used to gauge depth, a critical feature for next-generation home automation, robotics and other physical AI applications. We also achieved a milestone with our long-time customer and partner, Silicon Labs, shipping more than 10 million Wi-Fi units built on our 40LP platform.

這將使下一代應用優化智慧感測器具備一流的尺寸、重量、功耗和成本優勢。這些直接飛行時間感測器用於測量深度，這是下一代家庭自動化、機器人和其他實體人工智慧應用的關鍵特性。我們還與長期客戶和合作夥伴 Silicon Labs 共同實現了一個里程碑，交付了超過 1000 萬台基於我們 40LP 平台製造的 Wi-Fi 設備。

This platform features low latency standby mode to support power-efficient, always-on intelligent devices and is an integral part of GF's portfolio of advanced technology for sensing applications, delivering exceptional signal-to-noise ratio performance to ensure accurate data capture. Communication infrastructure and data center represent approximately 10% of the quarter's total revenue. I'd like to highlight 3 new optical networking design wins in the third quarter. These include a significant design win with Coherent, a new engagement with a top three U.S. TIA driver supplier and a win with a leading China-based vendor to serve that fast-growing market.

該平台具有低延遲待機模式，可支援節能、始終在線的智慧型設備，是GF先進感測應用技術組合的重要組成部分，可提供卓越的信噪比性能，確保準確的數據採集。通訊基礎設施和資料中心約佔本季總收入的 10%。我想重點介紹第三季獲得的 3 個新的光網路設計專案。其中包括與 Coherent 達成的一項重要設計合作，與美國排名前三的 TIA 驅動器供應商建立新的合作關係，以及與一家領先的中國供應商達成合作，以服務於這個快速增長的市場。

Collectively, these programs deepen our position in next-generation optical interconnects that are critical to AI data center growth. In SATCOM, we continue to build on our success with new wins with global players. During the quarter, we added a digital beam forming win for a Japan-based satellite program as well as an additional ground terminal low noise amplifier win.

這些項目共同鞏固了我們在下一代光互連領域的地位，而下一代光互連對於人工智慧資料中心的發展至關重要。在衛星通訊領域，我們不斷鞏固成功，與全球企業取得了新的合作成果。本季度，我們新增了一項針對日本衛星專案的數位波束成形工程訂單，以及一項地面終端低雜訊放大器訂單。

Overall, the progress we are making across optical networking, satellite communication and quantum computing reflect the strength of our product portfolio and the trust our customers place in us. With these partnerships and our expanding pipeline, I'm confident we are well positioned to capture the long-term growth opportunities ahead.

總體而言，我們在光網路、衛星通訊和量子運算領域的進展反映了我們產品組合的實力以及客戶對我們的信任。憑藉這些合作夥伴關係和不斷擴大的產品線，我相信我們已做好充分準備，抓住未來的長期成長機會。

I'll now pass the call over to Sam for a deeper dive on our financial results and guidance.

現在我將把電話交給 Sam，讓他更深入地分析我們的財務表現和展望。

Thank you, Niels. For the remainder of the call, including guidance, other than revenue, cash flow, net interest income and third quarter CapEx, I will reference non-IFRS metrics, which are included in today's press release and accompanying slides. As Tim noted, our third quarter results came in at the high end of the guidance ranges we provided in our last quarterly update. We delivered third quarter revenue of $1.688 billion, flat over the prior quarter and a 3% decrease year-over-year.

謝謝你，尼爾斯。在本次電話會議的剩餘部分，包括業績指引，除收入、現金流量、淨利息收入和第三季資本支出外，我將參考非國際財務報告準則（IFRS）指標，這些指標已包含在今天的新聞稿和隨附的幻燈片中。正如蒂姆所指出的，我們第三季的業績達到了我們在上一季更新中給出的預期範圍的高端。我們第三季營收為 16.88 億美元，與上一季持平，年減 3%。

We shipped approximately 602,300-millimeter equivalent wafers in the quarter, up 4% sequentially and up 10% from the prior year period. Wafer revenue from our end markets accounted for approximately 88% of total revenue. Non-wafer revenue, which includes revenue from reticles, nonrecurring engineering, expedite fees and other items, accounted for approximately 12% of the total revenue for the third quarter.

本季我們出貨了約 602,300 毫米當量晶圓，較上季成長 4%，比上年同期成長 10%。來自終端市場的晶圓收入約佔總收入的 88%。非晶圓收入（包括光阻收入、非經常性工程收入、加急費和其他項目收入）約佔第三季總收入的 12%。

Let me now provide an update on our revenue by end markets. Smart mobile devices revenue increased approximately 10% sequentially and decreased approximately 13% from the prior year period. The year-over-year change was principally driven by onetime pricing adjustments made in the prior quarter with a limited number of dual source customers.

現在讓我向大家報告我們按終端市場劃分的收入情況。智慧型行動裝置營收季增約 10%，年減約 13%。與去年同期相比的變化主要是由於上一季對少數雙重來源客戶進行的一次性價格調整。

Going forward, we expect to gain a larger share of wallet with these customers. Automotive revenue decreased approximately 17% sequentially and increased 20% from the prior year period. The sequential change was the result of customer shipment timings consistent with the prior year period. Year-over-year revenue gains in our automotive end market were driven by share and content expansion, and we remain on track to grow automotive revenue in the mid-teens percentage range for 2025.

展望未來，我們期望能從這些客戶身上獲得更大的消費份額。汽車產業營收季減約 17%，較去年同期成長 20%。順序變化是由於客戶出貨時間與去年同期一致所致。我們的汽車終端市場收入年增率主要得益於市場份額和產品內容的擴張，我們仍有望在 2025 年實現汽車收入 15% 左右的成長。

Home and Industrial IoT revenue decreased approximately 14% sequentially and 16% from the prior year period. This was principally driven by a year-over-year reduction in wafer revenue associated with aerospace and defense applications as certain products reach end of life with new applications now taping out and expected to move into production in 2026.

家庭和工業物聯網收入較上季下降約 14%，年減約 16%。這主要是由於與航空航天和國防應用相關的晶圓收入同比下降所致，因為某些產品已達到使用壽命終點，而新的應用正在流片，預計將於 2026 年投入生產。

Finally, communications infrastructure and data center revenue increased approximately 2% sequentially and 32% over the prior year period. With improved visibility into our fast ramping optical networking and SATCOM businesses, we now expect full year 2025 revenue in this end market to grow in the low 20s percentage range, up from the high teens outlook indicated on prior earnings calls.

最後，通訊基礎設施和資料中心收入較上季成長約 2%，較去年同期成長 32%。隨著我們對快速成長的光網路和衛星通訊業務的了解不斷加深，我們現在預計 2025 年全年該終端市場的營收將成長 20% 左右，高於先前財報電話會議上預測的 10% 左右。

For the third quarter, we delivered gross profit of $439 million, which was at the high end of our guided range and translates into approximately 26% gross margin. Notwithstanding flat sequential revenue, gross margin expanded sequentially and year-over-year approximately 80 basis points and 130 basis points, respectively.

第三季度，我們實現了 4.39 億美元的毛利，處於我們預期範圍的高端，毛利率約為 26%。儘管營收環比持平，但毛利率環比和同比分別增長了約 80 個基點和 130 個基點。

Gross margin expansion remains a key focal area for GF, and we believe we're beginning to see the benefits associated with a shift towards a more accretive product mix and increased revenue from non-wafer technology services. R&D for the quarter was $111 million, and SG&A was $68 million. Total operating expenses of $179 million were up marginally quarter-over-quarter and represented approximately 11% of total revenue.

毛利率擴張仍然是GF的一個重點領域，我們相信，隨著產品組合向更具增值性的方向轉變以及非晶圓技術服務收入的增加，我們已經開始看到由此帶來的好處。本季研發費用為 1.11 億美元，銷售、管理及行政費用為 6,800 萬美元。總營運支出為 1.79 億美元，比上一季略有成長，約佔總收入的 11%。

We delivered operating profit of $260 million for the quarter and an operating margin of 15.4%, which is at the high end of our guided range and 180 basis points above the prior year period. Third quarter net interest income was $18 million, and we incurred income tax expense of $46 million in the quarter. We reported third quarter net income of $232 million, an increase of approximately 1% from the prior year period.

本季我們實現了 2.6 億美元的營業利潤，營業利潤率為 15.4%，處於我們預期範圍的高端，比去年同期高出 180 個基點。第三季淨利息收入為 1,800 萬美元，本季所得稅支出為 4,600 萬美元。我們公佈第三季淨利為 2.32 億美元，比上年同期成長約 1%。

As a result, based on a fully diluted share count of approximately 559 million shares, we reported diluted earnings of $0.41 per share for the third quarter, which was at the high end of our guided range. Let me now provide some key balance sheet and cash flow metrics. Cash flow from operations for the third quarter was $595 million. CapEx for the quarter was $189 million or roughly 11% of revenue. Adjusted free cash flow for the quarter was $451 million, which represented an adjusted free cash flow margin of approximately 27% in the quarter.

因此，根據約 5.59 億股的完全稀釋股份數，我們報告第三季稀釋後每股收益為 0.41 美元，處於我們預期範圍的高端。現在讓我提供一些關鍵的資產負債表和現金流量指標。第三季經營活動產生的現金流量為 5.95 億美元。本季資本支出為 1.89 億美元，約佔營收的 11%。本季調整後自由現金流為 4.51 億美元，本季調整後自由現金流利潤率約 27%。

At the end of the third quarter, our balance sheet remains strong with our combined total cash, cash equivalents and marketable securities at approximately $4.2 billion. Our total debt was $1.2 billion, and we also have a $1 billion revolving credit facility, which remains undrawn. Next, let me provide you with our outlook for the fourth quarter of 2025.

第三季末，我們的資產負債表依然穩健，現金、現金等價物和有價證券總額約 42 億美元。我們的總債務為 12 億美元，此外我們還有 10 億美元的循環信貸額度，目前尚未動用。接下來，我將為大家介紹我們對 2025 年第四季的展望。

We expect total GF revenue to be $1.8 billion, plus or minus $25 million. Of this, we expect non-wafer revenue to be approximately 13% of total revenue. We expect gross margin to be approximately 28.5%, plus or minus 100 basis points, which reflects the sequential and year-over-year growth in gross margin. Excluding share-based compensation, we expect total operating expenses to be $210 million, plus or minus $10 million. We expect operating margin to be in the range of 16.8%, plus or minus 170 basis points.

我們預計 GF 總收入將達到 18 億美元，上下浮動 2,500 萬美元。其中，我們預計非晶圓收入將佔總收入的約 13%。我們預計毛利率約為 28.5%，上下浮動 100 個基點，這反映了毛利率的環比和同比增長。不計股權激勵費用，我們預計總營運費用為 2.1 億美元，上下浮動 1,000 萬美元。我們預計營業利潤率將在 16.8% 左右，上下浮動 170 個基點。

At the midpoint of our guidance, we expect share-based compensation to be approximately $63 million, of which roughly $16 million is related to cost of goods sold. We expect net interest and other income for the quarter to be between $4 million and $12 million and income tax expense to be between $40 million and $62 million, which translates to an effective tax rate of approximately mid- to high teens percentage for the full year 2025.

根據我們的預期，股權激勵支出中位數約為 6,300 萬美元，其中約 1,600 萬美元與銷售成本相關。我們預計本季淨利息及其他收入將在 400 萬美元至 1,200 萬美元之間，所得稅支出將在 4,000 萬美元至 6,200 萬美元之間，這意味著 2025 年全年的有效稅率約為 15% 至 10% 的中高稅率。

Based on a fully diluted share count of approximately 559 million shares, we expect diluted earnings per share for the fourth quarter to be $0.47, plus or minus $0.05. Finally, a brief update on our capital allocation activities. GF continues to generate strong, consistent adjusted free cash flow while retaining healthy balance sheet fundamentals. In 2025 alone, we have significantly reduced our outstanding debt, continued to optimize our capacity footprint by technology transfers and completed critical acquisitions to enable future growth, such as the recently closed MIPS transaction.

基於約5.59億股的完全稀釋股份數，我們預期第四季稀釋後每股收益為0.47美元，上下浮動0.05美元。最後，簡單介紹一下我們的資本配置活動。GF持續產生強勁、穩定的調整後自由現金流，同時維持健康的資產負債表基本面。僅在 2025 年，我們就大幅減少了未償債務，透過技術轉移繼續優化產能佈局，並完成了關鍵收購以實現未來的成長，例如最近完成的 MIPS 交易。

Looking ahead to 2026, we expect to continue with our objectives to reinvest in the business as well as planning for a systematic approach to returning an appropriate portion of free cash flow to shareholders. In closing, I want to express my appreciation to our employees worldwide for their dedication and execution that helped deliver this quarter's strong financial performance.

展望 2026 年，我們預計將繼續推進業務再投資的目標，並計劃採取系統性的方法，將適當比例的自由現金流返還給股東。最後，我要感謝我們全球員工的奉獻和執行力，正是他們的努力使得本季取得了強勁的財務表現。

Over the last few years, I've had the privilege of leading our business finance and operations functions, working with exceptional team members from around the world, and I remain focused on executing a smooth transition for our finance and operations functions and partnering with Tim and Neils to advance our long-term strategic objectives.

在過去的幾年裡，我有幸領導我們的業務財務和營運職能，與來自世界各地的優秀團隊成員一起工作，我將繼續專注於實現財務和營運職能的平穩過渡，並與 Tim 和 Neils 合作，推進我們的長期戰略目標。

With that, let's open the call for Q&A. Operator?

接下來，我們進入問答環節。操作員？

Ross Seymore, Deutsche Bank.

羅斯·西摩，德意志銀行。

I want to ask one long-term one and then a shorter-term one. On the long-term one, you went into great details about your silicon photonics business. But I just wanted to ask two follow-ups on that. First, what do you believe to be the core differentiation of what GF offers versus any other foundry peers? And second, what sort of capital and CapEx needs to be applied if you're going to quintuple that business over the next five years?

我想問一個長期問題，然後再問一個短期問題。在長期發展方面，您詳細介紹了您的矽光子學業務。但我還有兩個後續問題想問。首先，您認為GF提供的服務與其他代工廠相比，其核心差異為何？其次，如果要在未來五年內將業務規模擴大五倍，需要投入多少資金和資本支出？

Very good. Thank you, Ross. So maybe I'll start with the first one. I mean just to recap, right? This is GF's core play in the data center. We've talked about two sets of data center priorities, one, of course, being power that we've spoken a bit about in our GaN announcements, but then networking, optical networking specifically being the secular trend that we see the industry now fully adopting both for pluggable optical transceivers and the transition to co-packaged optics.

非常好。謝謝你，羅斯。那我或許會從第一個開始。我的意思是，簡單回顧一下，對吧？這是GF在資料中心領域的核心佈局。我們已經討論過資料中心的兩大優先事項，一是電源，我們在 GaN 發布會上已經稍微談到過；二是網絡，特別是光網絡，這是我們現在看到業界全面採用的長期趨勢，無論是可插拔光收發器還是向共封裝光器件的過渡。

In many ways, GF was early in developing silicon photonics. We've been doing this for now more than a decade. As a result, we believe we have best-in-class device performance, really focusing around the electrical to optical -- the optical to electrical, excuse me, signal conversion. We do that through innovation around device structure, around material and increasingly around packaging and especially as we make the transition to co-packaged optics, some of the innovation we've been driving around how those packages get put together will play, I think, a critical role in that rollout and that adoption.

在許多方面，GF公司較早開發了矽光子學技術。我們從事這項工作已經超過十年了。因此，我們相信我們擁有同類最佳的設備性能，真正專注於電信號到光信號——抱歉，是光信號到電信號轉換。我們透過圍繞裝置結構、材料以及越來越多的封裝進行創新來實現這一目標，尤其是在我們向共封裝光學元件過渡的過程中，我認為，我們一直在推動的關於如何組裝這些封裝的一些創新，將在這一推廣和應用過程中發揮關鍵作用。

I think the other aspect of differentiation is the ecosystem we have been building around it to enable design support for our customers and also to enable critical components. For example, our announcement with Corning around the detachable fiber connector, a very important part of how can you make these devices both high performing, but also serviceable, maintainable in those data center context. So I think we're very bullish about the adoption story. We're very bullish also about GF differentiation. I'm going to let Sam comment about the CapEx.

我認為差異化的另一個方面是我們圍繞它構建的生態系統，旨在為客戶提供設計支持，並實現關鍵組件的交付。例如，我們與康寧公司就可拆卸光纖連接器發布的公告，對於如何在資料中心環境中使這些設備既高效能又易於維護和保養而言，這是一個非常重要的部分。所以我認為我們對市場普及前景非常樂觀。我們對GF差異化也非常看好。我打算讓 Sam 來評論一下資本支出方面的問題。

Yeah. Ross, just a quick follow-on there as far as the CapEx is concerned. Look, we've been on a bit of a journey, as you know, from a capacity and a CapEx point of view for really the last five years. We began that journey at roughly 2 million wafers of capacity a year, and we set ourselves a target to get to 3 million wafers of capacity. Now as we've gone through that, obviously, the demand environment has changed slightly.

是的。羅斯，關於資本支出方面，我還有一個補充說明。你看，正如你所知，過去五年裡，我們在產能和資本支出方面經歷了一段漫長的歷程。我們起步時年產能約 200 萬片晶圓，我們為自己設定了達到 300 萬片晶圓產能的目標。現在，隨著我們經歷這一切，很明顯，需求環境已經略有變化。

And so over the course of the last couple of years, you've seen us moderate some of that CapEx in and around the 10% of revenue versus that sort of broader model target of roughly 20%. So looking out to 2026, obviously, it's a little bit too soon to guide CapEx specifically, but you can infer from what Tim is saying around the opportunity that we see within silicon photonics that we'd expect to see a pickup in CapEx going into next year, call it, the midpoint of that range that we've trended in over the course of the last few years as well.

因此，在過去的幾年裡，你們已經看到我們調整了部分資本支出，使其達到收入的 10% 左右，而不是之前設定的 20% 左右的更廣泛目標。展望 2026 年，顯然現在就具體指導資本支出還為時過早，但從 Tim 所說的矽光子學領域的機會來看，我們可以推斷，我們預計明年資本支出將有所回升，可以說是在過去幾年我們一直追踪的區間的中點。

So hopefully, that helps. As we think about it beyond 2026, obviously, the foundational principle of why we invest in our capacity is tied to customer demand. And so if it's to be seen around the ramp in demand for silicon photonics and the continuation of the customer partnerships that we've certainly seen during the course of this year, it would justify incremental CapEx is a highly value-accretive end market for us.

希望這能有所幫助。當我們展望 2026 年以後時，很顯然，我們投資產能的根本原則與客戶需求息息相關。因此，如果矽光子學需求的成長以及我們今年看到的客戶合作關係的延續能夠證明增加資本支出是合理的，因為矽光子學是一個對我們來說價值增值的終端市場。

And maybe, Ross, if I can just add a little bit more color on the nature of that CapEx for photonics wafer production, these are highly valuable wafers. So from a wafer volume point of view, it's relatively small from a wafer value point of view, relatively high. And so very CapEx efficient when it comes to adding wafer capacity. Some of the CapEx that Sam alluded to will also be around packaging capacity because that goes alongside especially the co-packaged optics transition. So that will be -- both of those will be featured in 2026.

羅斯，或許我可以再補充一點關於光子晶圓生產資本支出的性質，這些晶圓價值極高。所以從晶圓體積的角度來看，它相對較小；但從晶圓價值的角度來看，它相對較高。因此，在增加晶圓產能方面，資本支出效率非常高。Sam提到的部分資本支出也將與封裝能力有關，因為這與共封裝光學元件的轉型特別相關。所以，這兩項都將在 2026 年推出。

Great. I guess as my follow-up in the shorter-term question is just in the fourth quarter, I just want to talk about the end markets and what you're assuming sequentially in your revenue guide. You gave the full year guidance for automotive and com data center. So those ones seem to be quite obvious. But I guess what I'm getting at is the smart mobile device side of things.

偉大的。我想就短期問題提出的後續問題是，在第四季度，我只想談談終端市場以及您在收入指南中逐季做出的假設。您提供了汽車和通訊資料中心的全年業績指引。所以這些看起來都很明顯。但我真正想說的應該是智慧型行動裝置方面的事情。

How are you seeing that in the fourth quarter? How did the ASP cuts lead to any unit share gains? And when do you think that segment could return to year-over-year growth?

你如何看待第四季的情況？平均售價下調如何帶來了市場佔有率的成長？您認為該細分市場何時才能恢復年成​​長？

Sure, Russ. So I'll kick off there, and then I'll let Tim and Neils add any other commentary in terms of the long-term opportunities that we're seeing in smart mobile more specifically. But you hit the nail on the head as it relates to some of those dynamics that we saw in the third quarter. And the way we think about our business, Ross, it's really on a year-over-year basis.

當然，拉斯。那麼我就先從這裡開始，然後讓蒂姆和尼爾斯就我們在智慧移動領域看到的長期機會發表其他評論。但你一針見血地指出了我們在第三季度看到的一些動態。羅斯，我們看待業務的方式，實際上是按年來計算的。

And look, we've continued to see very strong year-over-year growth from an automotive point of view in the third quarter, comms infrastructure and data center was up 32%, and we've also had a high contribution associated with wafer from non-wafer revenue services. So all said and done, we're seeing the right momentum in growth as it relates to the end markets where we see most of that accelerated opportunity.

而且，從汽車產業的角度來看，第三季我們繼續看到非常強勁的年成長，通訊基礎設施和資料中心成長了 32%，此外，非晶圓收入服務也為晶圓業務做出了很高的貢獻。總而言之，我們看到了與終端市場相關的成長勢頭良好，而我們認為這些市場蘊藏著最大的加速成長機會。

Now look, the balance on that, and again, I'll talk more specifically around the fourth quarter, but in the context of the full year is that you can infer that from a mid-teens expected full year growth in automotive that sequentially, we'd expect quite a strong ramp going into 2024 -- excuse me, going into Q4, which is quite consistent with the 2024 sequential ramp that we saw last year as well. Similarly, as it relates to comms infrastructure and data center, we provided that updated guidance now in the low 20s range, so you can infer what sequentially that looks like.

現在，關於這一點，我再具體談談第四季度的情況，但就全年而言，你可以推斷，從汽車行業全年預期兩位數中段的增長來看，我們預計到 2024 年（抱歉，是到第四季度）將出現相當強勁的增長，這與我們去年看到的 2024 年環比增長非常一致。同樣，就通訊基礎設施和資料中心而言，我們已提供更新後的指導意見，目前在 20 左右的範圍內，因此您可以推斷其順序變更情況。

Now the -- if you like, the offsets as it relates to smart mobile devices and IoT more specifically, for the full year, IoT, we expect to be down about mid-single digits. That's really a function of some of that aerospace and defense revenue that we saw falling out and we commented in the prepared remarks.

現在，如果你願意的話，可以談談與智慧行動裝置和物聯網相關的抵銷因素，更具體地說，就全年而言，物聯網，我們預計會下降個位數百分比左右。這實際上是由於我們看到的航空航天和國防收入有所下降，我們在準備好的演講中也提到了這一點。

And then as it relates to smart mobile devices, clearly a function of some of those onetime pricing adjustments, which are in the rearview mirror now, but that will contribute to kind of a low double-digit percentage decline on a year-over-year basis. So that's how we think about it full year, and you can infer from what that means the quarter-to-quarter dynamics. And then I'll let Tim and Neils comment on the longer term where we see those opportunities.

至於智慧型行動設備，顯然這與一些一次性的價格調整有關，這些調整現在已經過去了，但這將導致其同比出現兩位數的百分比下降。這就是我們對全年情況的思考方式，你可以從中推斷出季度間的動態變化。然後我會讓提姆和尼爾斯來談談我們認為有哪些長期機會。

Yeah. Thank you, Sam. I think on the longer term, Ross, for smart mobile, we're very focused on where we can be the most differentiated. And so I'd say we see great traction in areas like audio, haptics, advanced display, advanced imaging, areas where GF technologies play a key role, both by the way, in the handsets of today, but also engagements like we mentioned in areas like smart glasses, that form factor becoming increasingly viable, I think, from a high-volume perspective.

是的。謝謝你，山姆。羅斯，我認為從長遠來看，對於智慧型手機領域，我們非常關注的是如何做到最大程度的差異化。因此，我認為我們在音頻、觸覺、先進顯示、先進成像等領域看到了巨大的發展勢頭，這些領域中，GF 技術發揮著關鍵作用，不僅在當今的手機中如此，而且在我們提到的智能眼鏡等領域也是如此，我認為，從大批量生產的角度來看，這種外形尺寸正變得越來越可行。

And so we see longer-term good traction in smart mobile in those differentiated areas. And I think that's true also in the IoT space. Obviously, you've got a broad set of end markets contained with IoT, but you see good traction in medical. You see good traction in industrial. And even in consumer, some of the announcements we've made, including companies like Silicon Labs, again, indicating good long-term growth in those markets as well.

因此，我們看到智慧移動技術在這些差異化領域具有長期良好的發展動能。我認為這一點在物聯網領域也同樣適用。顯然，物聯網涵蓋了廣泛的終端市場，但在醫療領域取得了良好的發展勢頭。在工業領域，我們看到了良好的發展動能。即使在消費領域，我們發布的一些公告，包括對 Silicon Labs 等公司的公告，也再次表明這些市場將有良好的長期成長。

David O'Connor, BNP Paribas.

David O'Connor，法國巴黎銀行。

Maybe a question on the onshoring side of things. So firstly, congrats on the expanded partnership there with Silicon Labs. After the Apple deal last quarter, it seems to be increasing demand and traction for the U.S. onshore manufacturing that's starting to come through now. Maybe could you just talk about what that pipeline actually looks like? And then related to that, just your ability to support additional really high-volume wins out of the multifab.

或許可以問一個關於本土化的問題。首先，恭喜你們與 Silicon Labs 擴大了合作關係。繼上季蘋果收購案之後，美國本土製造業的需求和發展動能似乎正在增強，而這種勢頭現在也開始顯現。能否簡單介紹一下這條管道的實際結構？與此相關的是，你能否支援多晶圓廠實現更多真正的大量訂單。

No, thank you for that question. Obviously, it's a trend that we have been quite public about the engagement from customers over the last couple of quarters now. Just for those keeping score, we've had 8 specific customer announcements regarding U.S. onshoring. If you just do a rough cut of how much that customer set spends in terms of silicon in our addressable market, you're talking about between $15 billion and $20 billion of total spend.

不，謝謝你的提問。顯然，在過去的幾個季度裡，我們一直非常公開地談論客戶參與度，這已成為一種趨勢。僅供參考，我們已經發布了 8 個關於美國本土化的具體客戶公告。如果粗略估算一下我們目標市場中該客戶群在晶片方面的支出，總支出大約在 150 億美元到 200 億美元之間。

And so these are large representative customers that have significant opportunity to reshore capacity to the U.S. And from that point of view, we see a very strong share gain opportunity for GF. And they're coming for the footprint, but they're also coming for the differentiated technology. So we see that as very strong. There's a significant pipeline on top of that, to your question.

因此，這些都是具有代表性的大客戶，他們有很大的機會將產能遷回美國。從這個角度來看，我們認為GF有很大的市佔率成長機會。他們覬覦的是市場份額，但他們也覬覦的是差異化技術。所以我們認為這非常有力。至於你的問題，除此之外還有一條重要的管道。

A lot of other customers saying, look, what can we do, when can we do it? And that's again, that match of capacity and footprint being very important. I think from a timing point of view, we're talking about ramps in 2027, largely and beyond. And this is a secular shift that's durable. And so obviously, we're going through those product design wins, product qualification cycles that are necessary. U.S. is a large part of this and obviously very visible. But actually, the story is also replicating in areas outside the U.S.

很多其他客戶都在問，我們能做什麼？什麼時候可以做？再次強調，產能與佔地面積的匹配非常重要。我認為從時間安排的角度來看，我們主要討論的是 2027 年及以後的匝道建設。而且這是一種持久的世俗轉變。因此，很顯然，我們正在經歷那些必要的產品設計成功案例和產品認證週期。美國在其中扮演著重要角色，而且顯然非常引人注目。但實際上，這種情況也在美國以外的地區重演。

I think our announcements in Dresden a couple of weeks ago for our relatively smaller expansion investments we're making there are still backed by significant Europe for Europe, let's say, customer demand, key players like Infineon, like NXP, like Aumovio, like Bosch, all kind of publicly supporting the investments we're making there to build that fab to even further scale.

我認為，幾週前我們在德累斯頓宣布的相對較小的擴張投資計劃，仍然得到了歐洲的大力支持，例如客戶需求，以及像英飛凌、恩智浦、Aumovio、博世等主要參與者的公開支持，他們都支持我們在那裡進行的投資，以進一步擴大晶圓廠的規模。

And obviously, with that comes very accretive economics for that fab. And we're even seeing examples outside that in Singapore, and I think one that we mentioned in the prepared remarks, even Chinese fabless companies looking to have their own version of a diversified supply chain. The NOR Flash win that we had, we mentioned for Q3 is a good example of that moving to Singapore. So I think the story of supply diversification is just extraordinarily clear globally and only picking up in pace.

顯然，這將為這家工廠帶來非常可觀的經濟效益。我們甚至在新加坡以外的地方也看到了這樣的例子，而且我認為我們在準備好的發言稿中也提到過，甚至中國的一些無晶圓廠公司也在尋求建立自己的多元化供應鏈。我們在第三季贏得的 NOR Flash 就是一個很好的例子，說明這種模式正在向新加坡轉移。所以我認為，全球供應鏈多元化的趨勢非常明顯，而且這個趨勢還在加速發展。

Did you have a follow-up, David?

大衛，你還有後續問題嗎？

Yeah, I do. Maybe one on the technology side on the gallium nitride -- on the GaN side of things. So TSMC recently exited that GaN business and at the time, citing kind of low profitability and just the competition there was quite intense. Can you maybe talk about your GaN strategy, how that is kind of different? And how are you addressing these concerns?

是的，我有。或許可以從技術層面，例如氮化鎵（GaN）方面著手。所以台積電最近退出了氮化鎵業務，當時給出的理由是獲利能力較低，而且競爭非常激烈。您能否談談您的氮化鎵策略，以及它與其他策略有何不同？你們是如何解決這些問題的？

No, thank you for the question. Maybe I'll start and then Niels can add a little bit of color as well. Look, we're very excited about GaN. From a simple technology point of view, this is a way of achieving a significant improvement in power density, significantly reduced losses in switching and power conversion. If you think about where that matters, of course, one of the areas that matters most is in the data center, right?

不，謝謝你的提問。也許我可以先開始，然後尼爾斯也可以添加一些顏色。我們對氮化鎵（GaN）非常興奮。從簡單的技術角度來看，這是一種顯著提高功率密度、顯著降低開關和功率轉換損耗的方法。如果你仔細想想這在哪些方面很重要，當然，最重要的領域之一就是資料中心，對吧？

When you're talking about enormous amounts of power consumption based on the build-out, GaN plays a critical role in that market. Of course, it also plays a broader role in critical infrastructure. It plays a role in automotive. And so it actually has plenty of uses and even longer term plays a role in radio frequency and high-performance communication. So from a secular trend point of view, it's a great technology fit.

當談到基於構建的巨大功耗時，GaN 在該市場中扮演著至關重要的角色。當然，它在關鍵基礎設施中也發揮更廣泛的作用。它在汽車行業中發揮作用。因此，它實際上有很多用途，甚至從長遠來看，它在射頻和高效能通訊中也發揮作用。因此，從長期趨勢的角度來看，這是一項非常契合的技術。

From a customer traction point of view, we also see customers very much focused on sourcing that technology in the U.S. again, key differentiation from GF. We're building that in Burlington, Vermont, a fab that is well -- good track record in various complex technologies and one that customers trust to deliver in the future. So I think our strategy is quite different than TSMC, and it's a case of us focusing on where we're a natural athlete and they're focusing on where they are. And I think this is a good win-win for both of us.

從客戶吸引力的角度來看，我們也看到客戶非常注重在美國採購該技術，這與GF的關鍵差異在於此。我們正在佛蒙特州伯靈頓建造這座工廠，該工廠在各種複雜技術領域擁有良好的業績記錄，並且客戶相信它能夠在未來交付產品。所以我認為我們的策略與台積電截然不同，我們專注於我們擅長的領域，而他們則專注於他們不擅長的領域。我認為這對我們雙方來說都是雙贏的局面。

Neils?

尼爾斯？

Yeah. Maybe just to add to that, and you may recall from one of the previous earnings calls, our strategy on GaN is very focused around highly reliable, safe, high-quality devices. And obviously, in data centers, that is crucial to ensure there's no downtime.

是的。或許還要補充一點，您可能還記得我們在先前的財報電話會議上提到過，我們在氮化鎵領域的策略非常注重高可靠性、安全性和高品質的裝置。顯然，在資料中心，這一點至關重要，可以確保沒有停機時間。

In addition to that, we are actually focused on the technology in a very similar fashion to the BCD technologies, meaning we are not just going for the discrete device implementation, but we're adding technologies around the discrete devices that enable us to get more differentiated and higher performing and more reliable solutions to the market. So very, very, very focused strategy from our side, lots of cost of interest like Tim said, and the U.S. footprint is really just the cherry on the top.

除此之外，我們實際上以與 BCD 技術非常相似的方式專注於這項技術，這意味著我們不僅致力於分立裝置的實現，而且還在分立元件周圍添加技術，使我們能夠為市場提供更具差異化、性能更高、更可靠的解決方案。因此，我們採取了非常、非常、非常專注的策略，正如提姆所說，利息成本很高，而美國市場只是錦上添花而已。

Chris Caso, Wolfe Research.

克里斯‧卡索，沃爾夫研究公司。

The first question would be on gross margins and utilization. And obviously, you have the step-up here in the fourth quarter. But how should we think about that as we go into the new year that typically, you see some seasonality as you go into the March quarter. And ultimately, I think what drives the gross margin is going to be getting utilization rate up. Could you give some commentary on where you see that going as you go into next year?

第一個問題是關於毛利率和利用率的。顯然，第四節比賽會更加精彩。但是，隨著新年的到來，我們該如何看待這個問題呢？通常情況下，到了三月份，你會看到一些季節性現象。最終，我認為提高毛利率的關鍵在於提高產能利用率。您能否就明年的發展趨勢談談您的看法？

Yeah. Chris, Sam here. Happy to give that. I'll probably start with the third quarter dynamics, and then I think that's a good layup into how we're thinking about the fourth quarter as well. So look, taking a step back, third quarter gross margin up 80 basis points quarter-over-quarter, up about 130 basis points year-over-year. Now that's on a declining revenue profile on a year-over-year basis, flat revenue on a quarter-over-quarter basis.

是的。克里斯，我是山姆。樂意效勞。我可能會先從第三節的比賽動態說起，然後我認為這可以很好地引出我們對第四節的思考。所以，退一步來看，第三季毛利率較上季成長 80 個基點，年增約 130 個基點。現在，按年計算，收入呈下降趨勢；按季度計算，收入則持平。

So we set out with a very clear mission at the start of this year, which was notwithstanding some of the consumer-driven demand environment, focusing on improving profitability, consistent free cash flow generation. And that's really what you're seeing come through in the third quarter. Actually, all the more notable as well, Chris, given the fact in the third quarter of 2024, we still had about $40 million to $50 million of underutilization payments falling through at that point.

因此，今年年初，我們制定了一個非常明確的目標，那就是儘管面臨一些消費者驅動的需求環境，仍專注於提高獲利能力，持續產生自由現金流。而這正是你在第三節比賽中真正看到的。事實上，克里斯，更值得注意的是，在 2024 年第三季度，我們仍有大約 4,000 萬至 5,000 萬美元的未充分利用款項未能到位。

So call that roughly 2 points to 3 points of margin benefit in the third quarter of last year that we didn't get in the third quarter of this year. So it's very much a case of where we've been focusing on opportunities to improve the profitability structure within the business and also continuing to mix into accretive end markets. And what you're seeing is really a reflection of that starting to come through.

所以，我們可以說，去年第三季我們獲得了大約 2 到 3 個百分點的利潤率優勢，而今年第三季我們卻沒有獲得這些優勢。因此，我們一直專注於尋找機會來改善公司內部的獲利結構，並持續擴大增值終端市場。而你現在看到的，正是這種趨勢開始顯現的體現。

Obviously, we've increased and had some incremental benefit come through from our non-wafer technology services. That's also a strong leading indicator in terms of where we see future production ramp as well as we kind of develop those projects from a mask, a reticle nonrecurring engineering perspective as well and really kind of embarking on those new projects with customers.

顯然，我們的非晶圓技術服務有所增加，並帶來了一些增量效益。這也強烈預示著我們未來生產規模的提升方向，因為我們從掩模、光罩等非重複性工程的角度開發這些項目，並真正與客戶一起開展這些新項目。

A little bit of benefit came through, obviously, from D&A, which we talked about at the start of this year. And utilization has been probably the lowest of the contributors towards that margin dynamic. We started out this year in roughly the low 80s. We've been trending around the kind of mid-80s for the last couple of quarters, possibly a minor pickup in the fourth quarter.

顯然，D&A 帶來了一些好處，我們在今年年初就討論過這個問題。而利用率可能是造成這種利潤率動態變化的因素中最低的。今年年初，我們的氣溫大約在華氏 80 度左右。過去幾個季度，我們的匯率一直徘徊在 80 多美元左右，第四季可能會略有回升。

But again, just switching to the fourth quarter, what you're seeing is roughly 3 points of incremental benefit on a year-over-year basis at the midpoint of that guide. And actually, from a guide-to-guide perspective, about 3 points as well. And again, that's really a confluence of those initiatives that we focused on from continuing to improve the mix dynamics, focusing on productivity, improving the cost structure of the business and obviously taking a modest benefit from a utilization and D&A perspective.

但再來看看第四季度，你會發現，與去年同期相比，該季度中點的收益大約增加了 3 個百分點。實際上，從導遊的角度來看，也有大約 3 點。再次強調，這實際上是我們持續改善產品組合動態、提高生產力、改善業務成本結構以及從利用率和折舊角度獲得適度收益等一系列舉措的匯合。

Did you have a follow-up, Chris?

克里斯，你還有後續問題嗎？

I did. And there's a question with regards to the mobile business. And obviously, we've seen at least the potential for some consolidation in that business on the RF side. Some of that consolidation would affect some of your customers. What's your thoughts on that going forward of the potential effect of consolidation among your customers in the mobile business?

我做了。關於行動業務，還有一個問題。顯然，我們已經看到射頻領域存在一些整合的潛力。部分整合可能會影響到您的部分客戶。您認為行動業務客戶整合可能對您的業務產生哪些影響？

Yeah. Thank you, Chris. And I presume you're largely referring to the announcements by Skyworks and Qorvo. Obviously, we're not to comment on that merger itself. But look, I'd say for both companies, we have a very long track record of serving both of them, and those partnerships go back even before GF with GF in some parts of our business. And it's partly because of technology leadership.

是的。謝謝你，克里斯。我猜你主要指的是 Skyworks 和 Qorvo 的公告。顯然，我們不便對此次合併本身發表評論。但是，我想說的是，對於這兩家公司而言，我們都有著非常悠久的為他們服務的記錄，而且在某些業務領域，我們與GF的合作關係甚至可以追溯到GF成立之前。部分原因在於其技術領先地位。

They're obviously leaders in the RF field and have been key partners for us in building and deploying our road maps. And so that's been a very tight collaboration in the case of both those companies. I think both of them are also increasingly focused on supply security, U.S. manufacturing and so on. So I think all the ingredients for strong relationships, strong future business are there with both companies. I don't think that will change whether they're one company or two companies going forward.

他們顯然是射頻領域的領導者，也是我們制定和部署路線圖的關鍵合作夥伴。因此，這兩家公司一直保持著非常緊密的合作關係。我認為他們兩人也越來越關注供應鏈安全、美國製造業等等。所以我認為兩家公司都具備建立牢固關係和實現穩固未來業務發展的所有要素。我認為無論未來他們是一家公司還是兩家公司，情況都不會改變。

Harlan Sur, JPMorgan.

哈蘭‧蘇爾，摩根大通。

Many of your customers are coming off the bottom of the nearly two-year long down cycle, right? But not seeing that sort of early cycle kind of strong recovery trajectory profile, but instead seeing a return to a more normal kind of seasonal profile in their businesses. You guys are already starting wafers for the March quarter given your manufacturing lead times.

您的許多客戶都正從近兩年的低迷週期中走出來，對嗎？但他們並沒有看到早期週期中那種強勁的復甦軌跡，而是看到他們的業務恢復到更正常的季節性模式。鑑於你們的生產週期，你們已經開始為三月的季度生產晶圓了。

I think normal seasonality is for the team is for revenues to be down about 10%, 12% sequentially. Is that how you are seeing the shipment profile early next year? Or maybe could it be down slightly more sequentially just given non-wafer revenues potentially kind of normalizing back to that sort of 10%, 11% of the mix?

我認為正常的季節性波動會導致團隊收入較上季下降約 10% 至 12%。這是您對明年年初出貨量的預測嗎？或者，考慮到非晶圓收入可能逐漸恢復到佔總收入的 10% 或 11% 左右，或許季減幅度會略大？

Yeah. Harlan, Sam here. Just happy to take the first part of that question. And look, I think one of the dynamics that you need to keep in mind, particularly when you look at our business is the diversification that we have across the portfolio today and actually increasing diversification. You take automotive, content for and data center, that continues to contribute a larger piece of the overall revenue stack.

是的。哈蘭，我是山姆。樂意回答這個問題的第一部分。我認為，尤其是在審視我們的業務時，需要牢記的一個動態是，我們目前在投資組合方面已經實現了多元化，並且實際上還在不斷提高多元化程度。汽車、內容和資料中心等產業在整體收入結構中持續佔據越來越大的份額。

And so the point there being that there's no single cyclical trend that actually is the determining factor in terms of where we see the revenue profile in the business. Now it's a little bit too soon to go into guiding the first quarter of 2026 at this point. I think you've heard from our customers that they're expecting, as you say, that kind of typical seasonal range, which you outlined on the call.

因此，關鍵在於，並沒有任何單一的周期性趨勢能夠真正決定企業的收入狀況。現在預測 2026 年第一季的走勢還為時過早。我想您已經從我們的客戶那裡了解到，他們期望的是像您所說的那樣，具有典型的季節性波動範圍，正如您在電話會議上概述的那樣。

And maybe just to cover off a little bit of your second part of your question, which is around the non-wafer revenues. Look, this has been a healthy tailwind as we've gone through this year. We signaled it on prior calls. Some of that really is a function of where we see customer dialogue and the timing of new products for our customers, the timing of new engagements on engineering services as well.

也許可以稍微回答一下你問題的第二部分，也就是關於非晶圓收入的問題。你看，今年以來，這一直是個有利的順風。我們在之前的電話會議中已經表達過這個意思。這其中一部分確實取決於我們與客戶的對話情況、新產品上市的時機以及工程服務新合作的時機。

And so that's what you're seeing starting to come through really in the third quarter. And obviously, we guided 13% expectation of revenue in the fourth quarter as well. So really a function of those activities. But more broadly, this plays very much to the increased suite of services that as GF, we're able to offer our customers. And clearly, as we think about '26 and beyond and continuing to integrate MIPS into the business and the offering that they have in terms of expanding suite of services for customers, those non-wafer technology services become an increasingly important component of the business.

所以，這就是你在第三節開始真正看到的情況。顯然，我們對第四季營收的預期也為 13%。所以這其實是這些活動的結果。但更廣泛地說，這與我們作為GF能夠為客戶提供的日益豐富的服務密切相關。顯然，當我們展望 2026 年及以後，並繼續將 MIPS 整合到業務中，以及他們為客戶提供的擴展服務套件中時，這些非晶圓技術服務將成為業務中越來越重要的組成部分。

Did you have a follow-up?

您有後續跟進嗎？

Yeah, I did. So thanks for giving us an update on the diversification efforts. Obviously, geographical diversification, supply chain diversification is extremely important for your customers. With that in mind, can you guys just give us an update on your China for China strategy? I think you guys announced the partnership with Zen Semiconductor in China last quarter.

是的，我做了。感謝您向我們報告多元化發展的最新進展。顯然，地域多元化和供應鏈多元化對您的客戶來說極為重要。有鑑於此，你們能否為我們介紹一下你們的「中國對中國」戰略的最新進展？我認為你們上個季度宣布了與中國Zen Semiconductor的合作關係。

The GF team, I think, has had very strong success within the domestic China automotive markets, for example, with your differentiated technology. And for your non-China customers, obviously, they want local supply to ship to their China customers, right? So what's the time line for transferring, qualifying, ramping production of their manufacturing processes at Zensemi? And is the business model royalty based? Or are you splitting profits? Any insights here would be very helpful.

我認為，GF團隊憑藉其差異化技術，在中國國內汽車市場取得了非常大的成功。而對於你的非中國客戶來說，顯然他們希望從本地採購，然後發貨給他們的中國客戶，對吧？那麼，Zensemi 的製造流程轉移、認證和量產的時間表是什麼？該商業模式是基於版稅的嗎？還是你們在平分利潤？任何見解都將非常有幫助。

Yeah. Thank you, Harlan. So look, we've spoken on a few different calls about China for China and how we've been addressing that. Again, as a recap, our strategy has been for specific technologies where there is strong local manufacturing design from our customers to make those available locally in Guangzhou, as you mentioned. That technology is typically in the microcontroller space, automotive imaging, increasingly also technologies in the power space, all relevant for that local automotive build-out and beyond.

是的。謝謝你，哈蘭。所以，我們已經在幾次不同的電話會議上談到了中國問題以及我們是如何應對的。再次總結一下，正如您所提到的，我們的策略是針對客戶在本地製造設計方面具有強大實力的特定技術，在廣州本地提供這些技術。這項技術通常應用於微控制器領域、汽車成像領域，並且越來越多地應用於電源領域，所有這些都與本地汽車建造及其他領域相關。

I'd say customer traction has been very, very strong. We spoke briefly in the prepared remarks about our Global Technology Summit. We do three, the third of which was in Shanghai with very, very strong customer traction. Interestingly, not just from those multinational companies serving the China market, which perhaps was where we started this engagement, but more and more from also local Chinese players who are looking for both manufacturing in China, but also that diversification for their global exports outside China. So if anything, incrementally more bullish on the China story for us in terms of demand.

我認為客戶反應非常非常強烈。我們在事先準備好的演講稿中簡要地談到了我們的全球技術高峰會。我們做了三場，第三場在上海，獲得了非常非常強大的客戶反應。有趣的是，不僅是那些服務於中國市場的跨國公司（這或許是我們最初接觸這些公司的地方），而且越來越多的中國本土企業也希望在中國進行生產，同時也希望實現其全球出口在中國以外的多元化。因此，就需求而言，我們對中國市場前景更加樂觀。

And just as a level set, remember, our direct China business today is probably -- the nearly lowest amongst peers for larger semiconductor companies. And so net-net, we see China as actually quite a good upside for us over time. Obviously, led by saying these are automotive technologies, so they go through a product development cycle, a qualification cycle, but customers are excited about that. And obviously, we're supplying them already out of our global footprint in the meantime, while that ramp is still taking place.

需要說明的是，我們目前的中國直接業務規模在大型半導體公司中可能是最低的。因此總而言之，我們認為從長遠來看，中國實際上對我們來說是一個相當不錯的成長點。顯然，首先要說明的是，這些都是汽車技術，所以它們要經歷產品開發週期、認證週期，但客戶對此感到興奮。顯然，與此同時，我們已經從我們的全球業務範圍內向他們供貨，而產能爬坡仍在進行中。

C.J. Muse, Cantor Fitzgerald.

C.J. Muse，坎托·菲茨傑拉德。

Christopher MuseAnalyst

克里斯多福·繆斯分析師

I guess first question on non-wafer revenues. Based on your guidance, that business is going to grow 20% in 2025. Curious if you can give a little more color on what's driving that incremental growth? And if you could kind of help us understand whether we should assume similar type of growth into calendar '26?

我想先問一個關於非晶圓收入的問題。根據您的指導，該業務在 2025 年將成長 20%。能否詳細解釋一下推動這種漸進式成長的因素？如果您能幫我們了解一下，我們是否應該假設 2026 年也會有類似的成長？

Yeah. Thank you, CJ. No, it's a great question, and Sam sort of touched briefly on this. And let's look what is in non-wafer revenue, so we're clear on what goes in there. That consists of reticles for masks, for tape-out and nonrecurring engineering, increasingly other technology services, licensing is starting to also be where you're seeing the IP revenue starting to layer in. Sam mentioned MIPS as a driver of that.

是的。謝謝你，CJ。不，這是一個很好的問題，Sam 也曾簡要地談到這一點。讓我們來看看非晶圓收入包含哪些內容，這樣我們就能清楚了解其中的組成。這包括掩模版、流片和非重複性工程的掩模版，以及越來越多的其他技術服務，授權也開始成為智慧財產權收入的來源。Sam提到MIPS是造成這種情況的原因之一。

So look, I think there are some good tailwinds leading to that generally growing. Part of that is a higher number of design wins. We spoke about that leads to a high number of tape-outs, which tends to positively improve our non-wafer revenue. And then, of course, the acquisition of MIPS now starting, I'd say, starting to impact that as well. So I think that is a good trend, and we'll broaden that category going forward in years to come.

所以你看，我認為有一些有利因素會推動它整體成長。部分原因是設計得標數量較多。我們討論過，這會導致大量的流片，而這往往會對我們的非晶圓收入產生積極影響。當然，現在收購 MIPS 也開始對這一點產生影響了。所以我認為這是一個很好的趨勢，我們將在未來幾年擴大這一類別。

Great. And then I guess maybe to follow up on Ross' question around smartphones or smart mobility, sorry. As you think about calendar '26 and you reflect kind of the reset to pricing, but the hopeful gains in unit volumes, is that a business that can turn and grow now in calendar '26? Or are there still kind of headwinds that we should be thinking about?

偉大的。然後，我想就羅斯提出的關於智慧型手機或智慧旅行的問題做個後續討論，抱歉。展望 2026 年，考慮到價格的調整以及銷售量有望成長，這樣的企業能否在 2026 年實現轉虧為盈和發展？或者，是否還有其他我們應該考慮的不利因素？

Yeah. No, no, it's a great question. I think let's break it down into those two pieces. As we said, on the pricing side, this is dual source business that we took proactive steps to reset pricing with customers in order to gain more share. And the calculus there is that's more profitability for GF, and it's a win-win for us and our customers. So that reset is done, and that reset is now in place for the duration of some of those contracts and those contracts now still extend out several years.

是的。不，不，這是一個很好的問題。我認為我們可以把它拆分成這兩個部分來分析。正如我們所說，在定價方面，這是一個雙重來源的業務，我們採取了積極主動的措施與客戶重新調整價格，以獲得更大的市場份額。這樣看來，GF就能獲得更高的利潤，對我們和我們的客戶來說也是雙贏的。因此，重置已經完成，並且該重置措施將在某些合約的期限內生效，而這些合約現在仍將持續數年。

So we don't expect another kind of step down on the pricing side. And as you said, we do expect increased volumes relative to baseline for that. So I think on that dual source component, which is a limited part of that business, that dynamic is there. But then I think what you're seeing on the rest of mobile is two factors, right? One is the ramp of more differentiated solutions, what we're doing in the RF front end. Neils talked about CBIC, right? This is an incredibly interesting and exciting silicon germanium technology, strong customer traction, right?

因此，我們預期價格方面不會再出現其他形式的下調。正如你所說，我們預計銷量會比基線水平有所增長。所以我認為，在雙源元件（這是該業務的有限部分）中，這種動態是存在的。但我認為你在行動裝置上看到的其他情況是由兩個因素造成的，對嗎？一是逐步推出更多差異化解決方案，這是我們在射頻前端所做的。尼爾斯談到了CBIC，對吧？這是一種非常有趣且令人興奮的矽鍺技術，客戶反應強烈，對吧？

How do you improve performance of low-noise amplifiers, power amplifiers in the future. These are the technologies that are difficult to do, but where GF has a strong track record, and that's just one of the several technologies we're bringing to the mobile market going forward. And of course, a lot of things have cellular connectivity. And so the technologies have broader applications as well.

如何提高低雜訊放大器和未來功率放大器的性能？這些技術雖然難以實現，但GF在這方面有著良好的業績記錄，而這只是我們未來將引進行動市場的眾多技術之一。當然，很多東西都具備蜂窩網路連線功能。因此，這些技術也具有更廣泛的應用前景。

So we're very much focused on the differentiation. That will be a mix tailwind over time in mobile and those new form factors that I spoke about as well, things like smart glasses, increasingly good traction. Too early to call 2026, but definitely, we think this market, we have plenty of room to grow and plenty of areas to play within the differentiated technologies we have.

所以我們非常注重差異化。隨著時間的推移，這將對行動裝置以及我之前提到的那些新型設備形態（例如智慧眼鏡）產生積極影響，這些設備正越來越受到歡迎。現在預測 2026 年還為時過早，但可以肯定的是，我們認為這個市場還有很大的成長空間，在我們擁有的差異化技術範圍內也有很多發展領域。

Maybe the last comment, some of the customers we've talked about in the onshoring story are also significant players in the handset. And although those ramps are largely kind of '27 and beyond as we've talked about, they're obviously based on diversifying their supply and building a more global sourcing strategy for them. So that will also be a longer term, I think, tailwind for our mobile business.

最後補充一點，我們在本土化過程中提到的一些客戶也是手機產業的重要參與者。儘管正如我們之前討論過的，這些坡道大多是針對 2027 年及以後的車型，但它們顯然是基於供應多元化和建立更全球化的採購策略而製定的。所以我認為這也將是我們行動業務長期發展的順風。

Krish Sankar, TD Cowen.

Krish Sankar，TD Cowen。

Congrats on the silicon photonics win. I had two questions. First one, wafer shipments were up 4% Q-o-Q despite flat revenues. And you said for the full year, you're still tracking around 13% shipment on flattish revenues. So I'm just trying to figure out what does that imply for ASPs, both blended ASPs and ASP ex mobile? And then I had a follow-up.

恭喜你們在矽光子學領域取得勝利。我有兩個問題。首先，儘管營收持平，但晶圓出貨量較上季成長了 4%。您說，全年來看，在收入基本持平的情況下，出貨量仍維持在 13% 左右。所以我想弄清楚這對於ASP（包括混合型ASP和行動端ASP）意味著什麼？然後我還有後續跟進。

Yeah. So that's maybe, Krish, a good build on what I just talked about in terms of the pricing dynamics. And so it is very much that story of very specific customers where we proactively chose to make price changes from a share of wallet perspective and increase overall profit dollars to GF and obviously in a way our customers are supportive. That is the vast majority of the dynamic on pricing affecting both the quarter and the full year trajectory.

是的。所以，Krish，這或許可以很好地補充我剛才談到的價格動態方面的內容。因此，這很大程度上就是一個關於特定客戶的故事，我們主動從錢包份額的角度選擇改變價格，增加 GF 的整體利潤，顯然，我們的客戶在某種程度上也給予了支持。這就是影響季度和全年價格走勢的大部分動態因素。

Even within mobile outside those customers, actually, we see mix being a tailwind as we ramp additional high margin or higher-margin differentiated technologies. And across all of the other end markets, that's where we have very much sole-source business, and so pricing has been largely stable.

實際上，即使在行動領域，除了這些客戶之外，我們也看到產品組合成為一種順風，因為我們正在逐步增加高利潤或更高利潤的差異化技術。而在其他所有終端市場，我們基本上都是單一來源業務，因此價格基本上保持穩定。

And maybe also worth adding that this is the in-year pricing, but even the pricing that we're winning new designs on, and as Neils mentioned, year-on-year, we're winning significantly more designs, pricing is very stable. Customers are happy to pay for the value of what they're really looking for. They're looking for differentiated technology. They're looking for time to market. They're looking for supply security. They're looking for capacity, and they're willing to put the right price on that. And so we feel the overall price environment remains actually very constructive.

或許值得一提的是，這是年內定價，但即使是我們贏得新設計訂單的價格，正如尼爾斯所提到的，我們每年贏得的設計訂單數量都顯著增加，價格也非常穩定。顧客樂於為他們真正想要的東西支付相應的價值。他們正在尋找差異化技術。他們正在尋找上市時間。他們尋求的是供應安全。他們正在尋找產能，並且願意為此支付合適的價格。因此，我們認為整體價格環境實際上仍然非常有利。

And Krish, maybe just to add one point to that. I think a critical dynamic you need to continue to focus on is the margin structure within the business. Actually, the correlation, I think, between where some of the pricing movements versus where the margin is, we sort of dispelled some of that focus around pure ASP. And so from our point of view, the fact that we've incurred some of those ASP trends associated with a limited number of customers in smart mobile and still grown margin year-over-year, quarter-over-quarter, I think, is a good proof point there to focus on.

克里什，或許我還要補充一點。我認為你需要繼續關注的一個關鍵動態是企業內部的利潤結構。實際上，我認為，價格變動與利潤率之間的相關性，讓我們不再那麼關注純粹的平均售價。因此，從我們的角度來看，儘管我們在智慧移動領域擁有有限數量的客戶，但仍然實現了利潤率的逐年、逐季增長，我認為這是一個很好的證明點，值得我們關注。

Got it. Very helpful. Then a quick follow-up on the Silicon Labs expanded partnership -- is this a share gain thing where slab is moving more wafers to GlobalFoundries from another foundry? Or is it more new chip designs? How to think about that?

知道了。很有幫助。接下來快速跟進一下 Silicon Labs 擴大合作關係——這是否是為了增加市場份額，即 Slab 將更多晶圓從其他代工廠轉移到 GlobalFoundries？還是說會有更多新的晶片設計？如何看待這個問題？

Yeah. It's absolutely a share gain. Silicon Labs, you should ask them about their sourcing strategy, but what they've been clear about with us is that they are very keen to have a strong U.S. sourcing footprint for their business. Today, they do source from other foundries.

是的。這絕對是股價上漲。對於 Silicon Labs，你應該問問他們的採購策略，但他們已經明確告訴我們，他們非常渴望在美國建立強大的採購基礎。如今，他們也從其他鑄造廠採購原料。

I think over time, you would expect that to diminish. And given what we're offering them. And again, it's not just the U.S. sourcing, it's also a very strong focus on their technology platforms. Literally, everything they make is our technologies that we support and invest in. And I think it's not just a capacity partnership, it's also a technology partnership.

我認為隨著時間的推移，這種情況會逐漸減少。考慮到我們向他們提供的條件。而且，不只是美國本土採購，他們也非常重視自身的技術平台。實際上，他們生產的所有產品都是我們支持和投資的技術。我認為這不僅是能力合作，也是技術合作。

Joseph Moore, Morgan Stanley.

約瑟夫‧摩爾，摩根士丹利。

You've addressed a lot of the opportunities geopolitically. And I know you have a lot of capacity headroom overall. Can you give us a sense for that by region and to the extent that you get Silicon Labs type deals in the U.S. or in Europe, do you have capacity to continue to grow those businesses? Do you have space if you need to spend more money to grow? I think Malta was space constrained at one point. Can you just give us an update on how that utilization is by region?

你已經談到了很多地緣政治方面的機會。我知道你們整體上還有很多產能提升空間。您能否按地區為我們介紹一下情況？如果您在美國或歐洲達成類似 Silicon Labs 的交易，您是否有能力繼續發展這些業務？如果需要投入更多資金發展，你有足夠的空間嗎？我認為馬耳他一度面臨空間限制。能否提供一下各地區利用率的最新情況？

Yeah. Maybe I'll talk about how we think about capacity and then maybe Sam can talk a little bit more tactically about utilization. Joe, our footprint utilization, meaning our floor space utilization, we still have significant upside or room to grow within our current four walls.

是的。也許我會談談我們如何看待產能，然後也許 Sam 可以從戰術層面更詳細地談談利用率。Joe，就我們的佔地面積利用率而言，也就是我們的樓層空間利用率，我們在目前的四面牆內仍然有很大的成長空間。

In some of our sites, obviously, we're running up against the headroom there like in Dresden, where we start to make small investments to expand the footprint, converting in that case, our former test facility using that space and so on. Malta has significant floor space to grow. And so I think we have very, I'd say, short time to market for that growth because we -- again, we're not building new fabs to get all of that growth started.

顯然，在我們的一些場地，例如德累斯頓，我們已經遇到了空間不足的問題，在那裡我們開始進行小額投資來擴大規模，例如將我們以前的測試設施改造成利用該空間的場地等等。馬耳他擁有龐大的發展空間。所以我認為，我們實現這一成長的市場化時間非常短，因為我們——再說一遍，我們並沒有建造新的晶圓廠來啟動所有這些成長。

And what you also have to bear in mind is that we have significant -- probably the highest we've ever had in terms of level of government incentive programs to support that new CapEx for sure in the U.S., but also that's what we expect in Germany, and we will continue to have similar positive support levels in Singapore as we have had in the past there as well.

您還必須記住，我們擁有大量的政府激勵計劃來支持新的資本支出，這在美國肯定是前所未有的，我們在德國也期望如此，而且我們將繼續在新加坡提供與過去類似的積極支持。

So I think it's very, very capital efficient, very short time to market and with that government support alongside. That said, we're tough on ourselves. We scrutinize every dollar of incremental CapEx heavily. Is it based on real demand that we have line of sight of and is it in those areas that are highly kind of differentiated from a technology point of view. So silicon photonics will definitely prioritize those kind of areas when it comes to the investments. But overall, we're very disciplined in how we think about adding capacity. Do you have a follow-up, Joe?

所以我認為它的資本效率非常高，上市時間非常短，而且還有政府的支持。話雖如此，我們對自己要求很嚴格。我們對每一筆新增資本支出都進行嚴格審查。它是否基於我們所能看到的實際需求，以及它是否在技術角度上具有高度差異化的領域。因此，矽光子學在投資方面肯定會優先考慮這些領域。但總的來說，我們在考慮如何增加產能方面非常嚴謹。喬，你還有後續問題嗎？

Yeah, I do. I think you mentioned the sort of categories of non-wafer revenue. You talked about expedite fees. I'm curious, are you seeing a lot of that at this point? Are there any of the data center markets giving you expedites or just anything new that you see on that front?

是的，我有。我想你提到了非晶圓收入的類別。你提到了加急費。我很好奇，你現在常常看到這種情況嗎？資料中心市場是否有任何加速交付的措施，或者您在這方面看到了什麼新情況？

Yeah. So expedite is a portion of non-wafer revenue. I'd say, if anything, we do see a little bit more desire for expedites across different markets. I think we also see specific capacity corridors closer to full utilization, which is a great sign of demand for those differentiated areas. Are we in an extraordinary kind of scarcity situation across every part of the business? Not yet. But I think you are seeing increasing tightening across those very differentiated corridors.

是的。因此，加急服務是晶圓以外收入的一部分。我認為，如果有什麼變化的話，那就是我們看到不同市場對加急服務的需求略有增加。我認為我們還可以看到一些特定產能走廊接近滿載運轉，這很好地表明了這些差異化領域的需求。我們是否正處於企業各環節都面臨的異常短缺局面？還沒有。但我認為你會看到這些差異很大的通道之間的收緊程度正在加劇。

Just on the design win side, we talked about in this quarter and the previous quarter, up 100% in the previous quarter, up 50% of the number of design wins. That directly translates into tape-outs. So you continue to see the number of tape-outs coming up, meaning the reticle revenue going into non-wafer revenue as well. We expect to continue to grow, which is obviously a good sign for future growth of revenue overall.

單就設計訂單數量而言，我們在本季度和上一季度都提到了這一點，上一季度增長了 100%，設計訂單數量增長了 50%。這直接轉化為磁帶輸出。因此，你會看到越來越多的晶片流片，這意味著光阻收入也會轉化為非晶圓收入。我們預計將繼續成長，這顯然是未來整體營收成長的一個好兆頭。

This does conclude the question-and-answer session of today's program. I'd like to hand the program back to Eric Chow for any further remarks.

今天的問答環節到此結束。我謹將發言權交還給周永康，請他再作發言。

Thank you, Jonathan. Thank you for joining. We look forward to seeing you at the UBS Conference on December 2, and please do tune in to our investor webinar on December 3, focusing on physical AI. Thank you.

謝謝你，喬納森。感謝您的參與。我們期待在 12 月 2 日的瑞銀大會上見到您，也請您收看 12 月 3 日的投資者網路研討會，本次研討會將重點放在實體人工智慧。謝謝。

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude the program. You may now disconnect. Good day.

感謝各位女士、先生參加今天的會議。節目到此結束。您現在可以斷開連線了。再會。