格芯 (GFS) 2026 Q2 法說會逐字稿

Thank you for standing by, and welcome to the GlobalFoundries, Inc. second-quarter fiscal year 2026 financial results. (Operator Instructions) As a reminder, today's program is being recorded. And now I'd like to introduce your host for today's program, Eric Chow, Head of Investor Relations. Please go ahead, sir.

感謝各位稍候，歡迎參加格羅方德（GlobalFoundries, Inc.）2026 會計年度第二季財務業績電話會議。(接線員指示) 提醒各位，今天的會議將被錄音。現在我想介紹今天會議的主持人：投資人關係主管 Eric Chow。先生，請開始。

Thank you, operator. Good morning, everyone, and welcome to GlobalFoundries second-quarter 2026 earnings call. On the call with me today are Tim Breen, CEO; and Sam Franklin, CFO. A short while ago, we released GF's second-quarter 2026 financial results, which are available on our website at investors.gf.com, along with today's accompanying slide presentation.

謝謝，接線員。各位早安，歡迎參加格羅方德 2026 年第二季財報電話會議。今天與我一同出席的有執行長 Tim Breen，以及財務長 Sam Franklin。稍早我們已發布 GF 2026 年第二季財務結果，可於我們網站 investors.gf.com 查閱，並附有今天的簡報投影片。

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 made available in today's press release and accompanying slides. Please note that these financial results are unaudited and subject to change.

本次電話會議將被錄音，並將於我們投資人關係網頁提供重播。在本次會議中，我們將同時呈報 IFRS 與非 IFRS 財務衡量指標。與非 IFRS 指標最直接可比的 IFRS 指標，以及非 IFRS 指標的調節表，已於今日新聞稿與隨附投影片中提供。請注意，這些財務結果尚未經審計，且可能變更。

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 the future tenants. 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.

今天電話會議中的某些陳述可能被視為前瞻性陳述。此類陳述可透過 believe（相信）、expect（預期）、intend（打算）、anticipate（預估）、may（可能）等用語，或使用未來時態來辨識。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，我們亦不承擔更新今日所作任何前瞻性陳述之義務。

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 caption Risk Factors in our annual report on Form 20-F and in any current reports on Form 6-K furnished with the SEC.

關於可能導致實際結果與前瞻性陳述有重大差異之因素的更多資訊，請參閱我們今日發布的新聞稿，以及我們向美國證券交易委員會（SEC）提交文件中所述之風險與不確定性，包括我們 Form 20-F 年報中「風險因素」章節，以及向 SEC 提交的任何 Form 6-K 當期報告。

In terms of upcoming events, we will be participating in a fireside chat at the Goldman Sachs Communacopia and Technology Conference in San Francisco on September 8. We will begin today's call with Tim providing a summary update on the business environment and technologies, followed by Sam, who will provide details on our second-quarter results and third-quarter guidance. We will then open the call for questions with Tim and Sam. We request that you please limit your questions to one with one follow-up.

就近期活動而言，我們將於 9 月 8 日在舊金山參加高盛 Communacopia 與科技大會的爐邊談話（fireside chat）。今天的會議將先由 Tim 概述商業環境與技術的最新進展，接著由 Sam 說明第二季業績與第三季展望指引。之後我們將由 Tim 與 Sam 開放問答。也請各位將提問限制為一題，並可加問一個追問。

I'll now turn the call over to Tim.

現在把電話交給 Tim。

Thank you, Eric, and welcome, everyone, to our second-quarter 2026 earnings call. GF delivered strong results in the second quarter with revenue and non-IFRS profitability metrics at or above the high end of our guidance ranges. The team continued its rigorous execution ramping critical technology corridors where we see accelerating customer demand and the opportunity to create and capture value.

謝謝你，Eric，也歡迎各位參加我們 2026 年第二季財報電話會議。GF 在第二季交出強勁成果，營收與非 IFRS 獲利能力指標達到或高於我們指引區間的上緣。團隊持續以嚴謹的執行力推進關鍵技術走廊（technology corridors）的擴產與爬坡，在這些領域我們看到客戶需求加速成長，並有機會創造與掌握價值。

In particular, our comms infrastructure and data center end market delivered over 60% year-over-year growth in Q2, driven by continued demand for optical networking applications across our Silicon Photonics and Silicon Germanium platforms. This marked one of the fastest quarters of year-over-year revenue growth for an end market in our company history.

特別是，我們的通訊基礎設施與資料中心終端市場在第二季年增超過 60%，主要受惠於矽光子（Silicon Photonics）與矽鍺（Silicon Germanium）平台在光網路應用上的持續需求。這是我們公司歷史上某一終端市場最快的單季年對年營收成長之一。

We believe our value proposition has never been more in demand. Our differentiated technology portfolio and resilient global manufacturing footprint continue to strengthen our position with customers. We are seeing meaningful momentum as we execute our strategy and drive towards the long-term targets we shared at this year's Investor Day.

我們相信，我們的價值主張從未像現在這樣受到市場需求。我們具差異化的技術組合與具韌性的全球製造佈局，持續強化我們在客戶端的地位。隨著我們落實策略並朝今年投資人日所分享的長期目標邁進，我們正看到顯著的動能。

Let me now update you on three key developments in the quarter that are accelerating our strategic path. One, Quantum Technology Solutions; two, optical networking and power opportunities in the AI data center; and three, IP, software, and custom silicon.

接下來我將更新本季三項加速我們策略路徑的關鍵進展。第一，量子技術解決方案（Quantum Technology Solutions）；第二，AI 資料中心中的光網路與電源機會；第三，IP、軟體與客製化晶片（custom silicon）。

Starting with Quantum. A paradigm shift that will define the next chapter of high-performance computing over the coming decade and beyond. In May, we launched Quantum Technology Solutions, a new dedicated team and set of capabilities within GF that will enable the quantum industry to move from prototypes to high-volume production.

先從量子談起。這是一場典範轉移，將在未來十年乃至更長時間，定義高效能運算的下一個篇章。5 月，我們成立了量子技術解決方案（Quantum Technology Solutions），這是 GF 內部一支新的專責團隊與一套能力組合，將協助量子產業從原型走向高量產。

Just the CPUs, GPUs, and AI ASICs define today's compute paradigm, we believe quantum processor units or QPUs will be an essential part of tomorrow's. Advanced semiconductor manufacturing built securely here in the US will be essential for the scaling of this technology. Establishing the right production capability is now the critical enabler. The ability to manufacture and integrate complex quantum devices with consistency, yield, and scale. This is precisely where GF wins.

正如 CPU、GPU 與 AI ASIC 定義了今日的運算典範，我們相信量子處理單元（QPU）將成為明日運算典範不可或缺的一部分。在美國本土以安全方式建置的先進半導體製造能力，將是這項技術擴展的關鍵。建立正確的量產能力，如今已成為最關鍵的推動因素。也就是能以一致性、良率與規模來製造並整合複雜量子元件的能力。這正是 GF 的優勢所在。

Our Quantum strategy is cubit agnostic, meaning our manufacturing platforms are expected to support a broad range of leading modalities, including superconducting, trapped ion, photonic, topological, and spin. Our proven platforms like FDX provides a cryogenic CMOS foundation, and we are extending our advanced packaging capabilities into the cryogenic regime to enable the 3D heterogeneous integration that these systems require.

我們的量子策略採「量子位元（qubit）不可知」（qubit agnostic）方式，也就是我們的製造平台預期可支援多種領先的技術路徑（modality），包括超導、離子阱、光子、拓撲與自旋。我們已驗證的平台（如 FDX）可提供低溫 CMOS 基礎；同時我們也正將先進封裝能力延伸至低溫環境，以支援這些系統所需的 3D 異質整合。

As highlighted in our recent announcements, we are working closely with eight of the world's leading quantum-computing players, including partnerships with PsiQuantum, Quantinuum, and Quantum Motion as well as new endorsements from the quantum arms of large hyperscalers. Since launching just three months ago, we have already embarked on four new customer-specific quantum engagements with accelerating commercial momentum ahead.

如我們近期公告所強調，我們正與全球八家領先的量子運算業者密切合作，其中包括與 PsiQuantum、Quantinuum 與 Quantum Motion 的合作夥伴關係，以及來自大型超大規模業者（hyperscalers）量子部門的新背書。自三個月前成立以來，我們已啟動四項新的客戶專屬量子合作案，且商業動能正加速向前。

Advancing our quantum capabilities is anchored by an expected $375 million grant from the US Department of Commerce to accelerate the research development and build-out of quantum manufacturing capacity in the US. This critical partnership with the US government underscores why Quantum is not only a business opportunity, but also a national priority.

推進我們量子能力的基礎之一，是預期將獲得美國商務部 3.75 億美元的補助，用以加速研發並在美國擴建量子製造產能。與美國政府的這項關鍵合作，凸顯量子不僅是商業機會，同時也是國家優先事項。

We are only in the early stages of the nascent quantum opportunity. Over the next one to three years, we expect to generate quantum-related revenue largely through engineering engagements with customers reported within our Technology Services revenue. As customer platforms qualify and move into volume production, we expect quantum-related revenue from manufacturing services to ramp towards the end of the decade.

我們仍處於量子新興機會的早期階段。在未來一到三年內，我們預期量子相關營收將主要來自與客戶的工程合作案，並計入我們的技術服務（Technology Services）營收。隨著客戶平台完成認證並進入量產，我們預期來自製造服務的量子相關營收將在本十年末期開始爬坡。

Ultimately, our early momentum and customer-proof points in this emerging area perfectly encapsulate the outsized value GF provides, a strongly differentiated technology deep customer partnerships and a global secure manufacturing footprint.

最終而言，我們在此新興領域的早期動能與客戶驗證成果，完美體現 GF 所提供的超額價值：高度差異化的技術、深厚的客戶夥伴關係，以及全球且安全的製造佈局。

Let me turn to AI data center, where we continue to build momentum through new customer design wins and increasing engagement across the ecosystem. In the second quarter alone, we secured seven new optical networking design wins with customers across both pluggable transceiver suppliers as well as major hyperscaler and networking players.

接著談 AI 資料中心。我們透過新的客戶設計導入（design wins）以及在整體生態系中更深的合作參與，持續累積動能。僅在第二季，我們就取得 7 項新的光網路設計導入，客戶涵蓋可插拔收發器（pluggable transceiver）供應商，以及主要的超大規模業者與網路設備業者。

Silicon Photonics and Silicon Germanium each play critical roles in optical networking systems and combined with data center power represents three high-quality, long-term secular growth drivers which underscore our conviction in the ability to grow in the data center for years to come.

矽光子與矽鍺在光網路系統中各自扮演關鍵角色；再加上資料中心電源，合計構成三項高品質、長期的結構性成長驅動力，強化我們對未來多年在資料中心持續成長能力的信心。

I will walk through an update on each of these. For Silicon Photonics, let's start with pluggables, which contributes the vast majority of our Silicon Photonics revenue today. Thanks to our differentiated technology and advanced 300 millimeter photonics manufacturing footprint, we are actively engaged with four of the top five optical transceiver players.

我將逐一說明這些領域的最新進展。就矽光子而言，先從可插拔產品談起，這部分目前貢獻了我們矽光子營收的絕大多數。憑藉我們具差異化的技術與先進的 300 毫米光子製造佈局，我們正積極與前五大光收發器業者中的四家合作。

Given our strong capabilities and robust capacity ramp, we now expect our Silicon Photonics revenue as reported within the comms infrastructure and data center end market to more than double in 2026 compared with the year prior. Beyond just this year, we are progressing well on our multiyear road map to advance the enablement of modules delivering [1.6t, 3.2t], and beyond. High-volume manufacturing of our 200 gig per lane technology is underway. We have already demonstrated 400-gig capability and solutions for even greater bandwidth are in development.

鑑於我們強大的能力與穩健的產能爬坡，我們目前預期：以通訊基礎設施與資料中心終端市場口徑揭露的矽光子營收，2026 年將較前一年成長逾一倍。不僅如此，我們在多年期路線圖上也進展良好，持續推進可支援提供 [1.6t, 3.2t] 及更高規格之模組的技術使能。我們的每通道 200G 技術已進入高量產。我們也已展示 400G 能力，並正開發可支援更高頻寬的解決方案。

In addition to our robust pluggable offerings today, we see significant customer interest in our scale platform, the industry's first OCI MSA compatible solution for near and co-packaged optics. We currently have seven active engagements with leading companies on our scale platform and customer feedback on the merits of our technology and manufacturing capabilities has been very positive. We are already delivering tangible results for our customers today, having taped out a scale-related design win in Q2, and we expect to take out another in Q3.

除了我們目前強勁的可插拔產品組合之外，我們也看到客戶對我們的規模化平台展現出顯著興趣；該平台是業界首個與 OCI MSA 相容、用於近封裝與共封裝光學的解決方案。目前我們在規模化平台上已與多家領先公司展開七項進行中的合作案，客戶對我們技術與製造能力優勢的回饋非常正面。我們已在今年第二季完成一項與規模化相關的設計案 tape-out，為客戶交付了具體成果，並預期在第三季再完成另一項。

Specifically for near package optics, we see NPL as an important application and exciting opportunity ahead of the broader adoption of co-package solutions. Because Neo and co-packaged optics are built on a common photonic IC and because many components of GF scale solution support both near and co-packaged optics. Our customers benefit from the same underlying platform. As a result, we expect GF to benefit from the Silicon Photonics opportunity regardless of the rate and pace of various form factor adoptions by our customers.

就近封裝光學而言，我們將 NPL 視為在共封裝解決方案更廣泛採用之前，一項重要的應用與令人振奮的機會。由於 Neo 與共封裝光學建立在共同的光子 IC 之上，且 GF 規模化解決方案的許多元件同時支援近封裝與共封裝光學。我們的客戶可受益於相同的底層平台。因此，無論客戶採用各種外形規格的速度與節奏如何，我們預期 GF 都將受惠於矽光子（Silicon Photonics）的機會。

As we increase investments into our Silicon Photonics capabilities, the importance of government partnerships continues to grow. Last week, GF entered into a letter of intent with the US Department of Commerce for a $300 million award to accelerate the development of next-generation silicon photonics technologies in the US.

隨著我們加大對矽光子能力的投資，政府合作夥伴關係的重要性持續提升。上週，GF 與美國商務部簽署意向書（LOI），擬獲得 3 億美元獎助，以加速在美國開發下一代矽光子技術。

The funding will support advanced optical materials, modulated technologies, and packaging innovations that will enable next-generation near and co-packaged optics architectures, building directly on GF scale platform. The endorsement from our partners across the industry has reinforced our strategic conviction, including the world's top XPU providers, hyperscalers, AI connectivity leaders, and ecosystem partners.

該資金將支持先進光學材料、調變技術與封裝創新，進而實現下一代近封裝與共封裝光學架構，並直接建立在 GF 規模化平台之上。來自產業各方合作夥伴的背書強化了我們的策略信念，其中包括全球頂尖 XPU 供應商、超大規模雲端業者、AI 連結領導者與生態系夥伴。

We are pleased to take a central role in advancing optical innovation and development in the US and believe this recognition validates the strategic importance of Silicon Photonics, the excellent relationship we enjoy with our partners and GF leadership in these technologies.

我們很高興能在推動美國光學創新與發展中扮演核心角色，並相信這項肯定驗證了矽光子的重要策略地位、我們與合作夥伴之間良好的關係，以及 GF 在這些技術上的領導地位。

Another driver of data center momentum is high-performance silicon germanium, which powers the analog and mixed signal electronics at the heart of optical interconnects for AI and cloud infrastructure. Our differentiated SiGe platform delivers the bandwidth, signal integrity and power efficiency required for increasingly demanding optical networking applications, making it strongly complementary or Silicon Photonics portfolio.

推動資料中心動能的另一項因素是高效能矽鍺（SiGe），其為 AI 與雲端基礎設施之光互連核心的類比與混合訊號電子提供動力。我們具差異化的 SiGe 平台可提供日益嚴苛的光網路應用所需的頻寬、訊號完整性與功耗效率，使其與我們的矽光子產品組合形成高度互補。

During the quarter, we secured multiple new SiGe TIA and driver design wins across networking customers. Demand for SiGe remains strong, and we are oversubscribed throughout 2027. We are actively expanding capacity in our Vivant facility to support this demand. We believe SiGe represents another key growth opportunity for GF. Combined with our leadership in Silicon Photonics, GF offers a uniquely differentiated set of technologies that help address the power efficiency and signal integrity requirements of next-generation AI systems. The momentum we are seeing today reinforces our belief that we will be a key leader in optical networking for years to come.

本季期間，我們在網路客戶中取得多項新的 SiGe TIA 與驅動器設計案。SiGe 需求依然強勁，我們在 2027 年之前的產能已超額預訂。我們正積極擴充 Vivant 廠的產能以支援此需求。我們相信 SiGe 代表 GF 另一項關鍵成長機會。結合我們在矽光子領域的領導地位，GF 提供一套獨特且具差異化的技術組合，有助於滿足下一代 AI 系統對功耗效率與訊號完整性的要求。我們目前看到的動能強化了我們的信念：未來多年我們將成為光網路領域的重要領導者。

The third strong opportunity we see in the AI data center relates to power. In July, we closed the strategic acquisition of the custom power team from Potion Technologies in Europe, bringing an experienced design team focused on integrated voltage regulators or IVR. Together with our BCD, GaN, and integrated inductor capabilities, IVR further strengthens our road map depth in Power Technologies and expands our serviceable market in one of the fastest-growing opportunities within AI data centers.

我們在 AI 資料中心看到的第三個重大機會與電源相關。7 月，我們完成在歐洲對 Potion Technologies 客製化電源團隊的策略性收購，納入一支專注於整合式電壓調節器（IVR）的資深設計團隊。結合我們的 BCD、GaN 與整合電感能力，IVR 進一步強化我們在電源技術路線圖的深度，並擴大我們在 AI 資料中心中成長最快的機會之一的可服務市場。

Our goal is to help enable a new power architecture for AI infrastructure, one that brings power conversion closer to the processor and addresses the increasing efficiency, power density bandwidth, high current, and transient response requirements of next-generation XPUs.

我們的目標是協助推動 AI 基礎設施的新電源架構，將電源轉換更靠近處理器，並滿足下一代 XPU 對效率提升、功率密度、頻寬、高電流與瞬態響應日益提高的需求。

As AI workloads continue to scale, XPUs are consuming more power than ever before, increasing the need for solutions that can reduce power losses and deliver higher performance with an increasingly constrained thermal and physical footprint. Closed in Q2, this transaction brings new differentiated IVR technology, specialized engineering talent, and additional R&D capabilities that strongly complement of our portfolio, allowing us to capture a larger share of the growing power opportunity in AI data centers.

隨著 AI 工作負載持續擴大，XPU 的耗電量較以往更高，因而更需要能降低功率損耗、在日益受限的散熱與實體空間下提供更高效能的解決方案。此交易於第二季完成，帶來具差異化的 IVR 新技術、專業工程人才與額外的研發能力，與我們的產品組合高度互補，使我們能在 AI 資料中心不斷成長的電源機會中取得更大份額。

Finally, moving to another key element of our long-term strategy, our IP, software, and custom silicon capabilities. In June, we completed our previously announced acquisition of Synopsys ARC Processor IP Solutions business, an important milestone in advancing our strategy in physical AI and a notable step change in expanding GF serviceable addressable market.

最後，談到我們長期策略的另一個關鍵要素：我們的 IP、軟體與客製化晶片能力。6 月，我們完成先前宣布對 Synopsys ARC Processor IP Solutions 業務的收購，這是推進我們實體 AI（physical AI）策略的重要里程碑，也是在擴大 GF 可服務市場規模方面的一項顯著躍升。

As a recap, the strategic rationale is multifold. As AI increasingly moves beyond the data center into the physical world around us, it is transforming automotive, industrial automation, robotics, and intelligent edge devices. In that context, customers are looking for partners that can help them navigate the growing complexity of software, compute architectures, and semiconductor design.

回顧而言，其策略性理由包含多個面向。隨著 AI 日益從資料中心走向我們周遭的實體世界，它正在改變汽車、工業自動化、機器人與智慧邊緣裝置。在此背景下，客戶正在尋找能協助他們因應軟體、運算架構與半導體設計日益複雜性的合作夥伴。

Together with MIPS, this acquisition bolsters GS capabilities across risk five-processor IP, software development tools, and custom silicon design, enabling us to support customers from architecture and software through high-volume silicon production. We acquired a broad set of CPU, DSP, NPU, and broader risk five technologies as well as a proven software development toolkit and application-specific processor design capabilities.

結合 MIPS，此次收購強化了 GS 在 RISC-V 處理器 IP、軟體開發工具與客製化晶片設計方面的能力，使我們能從架構與軟體一路支援客戶到高量產的晶片生產。我們取得了廣泛的 CPU、DSP、NPU 以及更廣泛的 RISC-V 技術，同時也獲得經驗證的軟體開發工具套件與特定應用處理器設計能力。

With over 150 patents, 300 existing customers, and 400 R&D engineers around the world, this acquisition meaningfully expands our ecosystem reach and depth. Importantly, we are already seeing significant strategic benefits from our acquisition by combining MIPS and Synopsis ARC under one roof we are engaging with more customers earlier in the design cycle, shaping application-specific compute architectures and creating deeper, longer-lasting customer partnerships.

憑藉超過 150 項專利、300 家既有客戶，以及遍布全球的 400 名研發工程師，此次收購在實質上擴大了我們生態系的觸及範圍與深度。更重要的是，透過將 MIPS 與 Synopsys ARC 納入同一體系，我們已開始看到顯著的策略效益：能在設計週期更早期與更多客戶互動，塑造特定應用的運算架構，並建立更深、更長久的客戶夥伴關係。

To accelerate customer enablement, we are increasing investment in a number of R&D initiatives. These are focused, high-return programs that position us to capitalize on expanding opportunities while helping our customers innovate faster. Over time, we believe this creates a pathway to create custom silicon opportunities and enabling physical AI customers to run their AI inference workloads on GF and MIPS-based processing platforms.

為加速客戶導入，我們正增加對多項研發計畫的投資。這些是聚焦且高報酬的專案，使我們能把握不斷擴大的機會，同時協助客戶更快創新。長期而言，我們相信這將開闢一條創造客製化晶片機會的路徑，並使實體 AI 客戶能在以 GF 與 MIPS 為基礎的處理平台上執行其 AI 推論工作負載。

In summary, we made meaningful progress across several strategic growth areas this quarter. We delivered a record quarter for design wins across both communications infrastructure and data center and smart mobile devices. In differentiated areas such as display back planes for AI glasses, PMIC for premium smartphones and smart power stage gate drivers for data center power.

總結來說，本季我們在多個策略性成長領域取得了實質進展。我們在通訊基礎設施、資料中心以及智慧行動裝置等領域的設計案數量創下單季新高。在差異化領域方面，例如 AI 眼鏡的顯示背板、旗艦智慧型手機的 PMIC，以及資料中心電源用的智慧功率級閘極驅動器。

Our differentiated capabilities are helping customers solve increasingly complex challenges while positioning GF as a trusted technology partner. We are making critical investments and integrating strategic acquisitions that strengthen our competitive position, diversify our growth drivers, and provide a durable foundation for long-term profitable growth. I am proud of the team's diligent execution this quarter and excited about the opportunities ahead. I'll now pass the call over to Sam for a deeper dive on second-quarter 2026 financials.

我們具差異化的能力正協助客戶解決日益複雜的挑戰，同時使 GF 成為值得信賴的技術夥伴。我們正在進行關鍵投資並整合策略性收購，以強化競爭地位、分散成長動能來源，並為長期獲利成長提供穩固基礎。我為團隊本季的嚴謹執行感到自豪，也對未來的機會充滿期待。接下來我將把電話會議交給 Sam，請他更深入說明 2026 年第二季財務表現。

Thank you, Tim. For the remainder of the call, including guidance other than revenue, cash flow, and net interest income, I will reference non-IFRS metrics. GF delivered strong results in the second quarter, with revenue and non-IFRS gross margin exceeding the high end of our guidance ranges. Thanks to the efforts from our teams around the world to improve structural costs, raise manufacturing productivity, and accelerate growth in value-accretive secular end markets, we grew our gross margin by nearly 500 basis points year over year.

謝謝你，Tim。在本次電話會議剩餘部分，包括除營收、現金流與淨利息收入以外的指引，我將引用非 IFRS 指標。GF 在第二季交出強勁成果，營收與非 IFRS 毛利率均高於我們指引區間的上緣。得益於全球團隊在改善結構性成本、提升製造生產力，以及加速在具價值增益的長期趨勢終端市場成長方面的努力，我們的毛利率年增近 500 個基點。

Not only did this represent a second-quarter record, we delivered on our expectation to reach approximately 30% gross margin, well before the end of 2026. Driven by a richer mix of revenue, this quarter's results demonstrated a meaningful step forward towards our long-term objectives to achieve structurally higher margins and profitability.

這不僅創下第二季紀錄，我們也兌現了將毛利率提升至約 30% 的預期，且遠早於 2026 年底。在更佳的營收組合帶動下，本季結果顯示我們朝向長期目標——實現結構性更高的毛利率與獲利能力——邁出了重要一步。

Now on to the results. We delivered second-quarter revenue of $1.786 billion, up 9% sequentially and 6% year over year. We shipped approximately [625,000], a 300-millimeter equivalent wafers in the quarter, up 8% sequentially and 8% from the prior year period. Revenue from manufacturing services accounted for approximately 89% of total revenue. Revenue from Technology Services, which includes revenue from IP, licensing software, reticles, non-recurring engineering, expedite fees, and other items accounted for approximately 11% of total revenue for the second quarter.

現在來看業績結果。我們第二季營收為17.86億美元，較前一季成長9%，較去年同期成長6%。本季出貨約[625,000]片（以300毫米晶圓當量計），較前一季成長8%，較去年同期成長8%。製造服務營收約占總營收的89%。技術服務營收（包含IP、授權軟體、光罩、一次性工程費（NRE）、加急費及其他項目）約占第二季總營收的11%。

Following the acquisitions of MIPS and the Synopsys ARC IP business, we expect revenue contribution of approximately $100 million to $120 million towards our full-year 2026 Technology Services revenue, up from our prior expectation of $60 million to $100 million as these acquisitions continue to drive new opportunities with our customers.

在完成收購MIPS與Synopsys ARC IP業務後，我們預期這些收購將為2026年全年技術服務營收貢獻約1億至1.2億美元，高於先前預期的6,000萬至1億美元，因為這些收購持續為我們與客戶帶來新的機會。

In addition, driven by strong conversion of our design win pipeline, and an expanding scope of partnerships with customers, we expect sustained momentum in our revenue contribution from Technology Services. As a result, we expect Technology Services revenue towards the high end of the 10% to 12% range of total revenue in 2026, with a gross margin profile significantly higher than our corporate targets.

此外，在我們設計導入（design win）管線強勁轉換，以及與客戶合作夥伴關係範疇擴大的帶動下，我們預期技術服務的營收貢獻將維持動能。因此，我們預期2026年技術服務營收將接近總營收10%至12%區間的高端，且其毛利率結構將顯著高於公司的整體目標。

Let me now provide an update on our revenue and outlook by end market. Communications Infrastructure and Data Center represented approximately 16% of second-quarter total revenue. Revenue increased 20% sequentially and 62% year over year. This marked the seventh consecutive quarter of double-digit percentage year-over-year growth for Communications Infrastructure and Data Center, and the fastest quarterly year-on-year growth since 2022.

接下來我將按終端市場提供我們的營收更新與展望。通訊基礎設施與資料中心約占第二季總營收的16%。營收較前一季成長20%，較去年同期成長62%。這是通訊基礎設施與資料中心連續第七季達成兩位數的年增率，也是自2022年以來最快的單季年增幅。

Within this end market, we saw strong customer demand for our silicon photonics and silicon germanium offerings. In both of these high-margin technologies we're ramping capacity and making the necessary investments to unlock increases in demand indicated by our customers.

在此終端市場中，我們看到客戶對矽光子（Silicon Photonics）與矽鍺（SiGe）產品的需求強勁。在這兩項高毛利技術上，我們正提升產能並進行必要投資，以釋放客戶所指示的需求增量。

Beyond optical networking, we saw strong double-digit year-over-year growth in applications across both wireless infrastructure and storage. Given the accelerating demand outlook from our customers, we now expect to achieve full-year 2026 revenue growth in the range of 50% to 60% for our Communications Infrastructure and Data Center end market, up from our prior expectations of high [30s%] year-over-year growth, which we believe is an early indication of the long-term growth opportunities ahead for GF in this end market.

除光網路之外，我們也看到無線基礎設施與儲存等應用皆呈現強勁的兩位數年增。鑑於客戶需求展望加速，我們目前預期2026年全年通訊基礎設施與資料中心終端市場的營收成長可達50%至60%，高於先前預期的年增高[30%]區間；我們認為這是GF在該終端市場長期成長機會的早期跡象。

Beyond growth opportunities across Silicon Photonics and SiGe outlined by Tim, we also closed a first-of-a-kind design win of smart power stage gate drivers on our BCD platform. We see this as just one notable step forward in the rapidly evolving market for data center power applications.

除了Tim所提及在矽光子與SiGe的成長機會外，我們也在BCD平台上拿下首創的智慧功率級閘極驅動器（smart power stage gate drivers）設計導入。我們將此視為資料中心電源應用快速演進市場中的一項重要進展。

Automotive represented approximately 19% of second-quarter total revenue. Automotive revenue decreased 13% sequentially and 10% year over year, principally driven by customer-led shipment timings. However, for the full year, we continue to expect low double-digit percentage revenue growth for our automotive end market with a higher weighting towards the fourth quarter.

車用約占第二季總營收的19%。車用營收較前一季下滑13%，較去年同期下滑10%，主要受客戶主導的出貨時點影響。然而，就全年而言，我們仍預期車用終端市場營收將呈現低兩位數百分比成長，且權重將更偏向第四季。

As automotive semiconductor content continues to grow, we're encouraged by our design-win momentum with customers and the long-term growth opportunities these present. In the second quarter, we secured a significant automotive power design win for 5-volt and 10-volt power management integrated circuits built on our BCD platform. In addition, we also taped out an ADAS radar built on our FDX platform for Bosch, a notable milestone, and the combination of years of close partnership. These highlights reflect the strong momentum we continue to see across automotive power, processing, sensing, and safety applications.

隨著車用半導體含量持續提升，我們對與客戶的設計導入動能以及其帶來的長期成長機會感到鼓舞。第二季，我們在BCD平台上取得一項重要的車用電源設計導入，產品為5伏與10伏電源管理IC。此外，我們也為Bosch在FDX平台上完成一款ADAS雷達的tape-out，這是一項重要里程碑，也是多年緊密合作的成果。這些亮點反映出我們在車用電源、運算、感測與安全等應用上持續看到的強勁動能。

Smart mobile devices represented approximately 36% of second-quarter total revenue. Revenue increased 15% sequentially and decreased 6% from the prior year period. As noted by peers and customers across the industry, 2026 smart mobile handset forecasts have reduced meaningfully over the last quarter, principally due to the continued impact from memory pricing and associated shortages.

智慧行動裝置約占第二季總營收的36%。營收較前一季成長15%，較去年同期下滑6%。如同產業同業與客戶所指出，2026年智慧型手機預測在過去一季明顯下修，主要原因是記憶體價格持續影響及相關短缺。

As a result, we currently expect smart mobile devices to decline by a low-teens percentage year-over-year in 2026. Customer design win momentum for new generations of smart mobile devices continues to be positive. In the second quarter, we secured a notable design win on GF BCD platform with MediaTek, further validating our expanding power platform. This marked GF first-ever power management integrated circuit design win with our long-standing customer.

因此，我們目前預期2026年智慧行動裝置將呈現年減低十位數百分比。客戶在新一代智慧行動裝置的設計導入動能仍維持正向。第二季，我們在GF BCD平台上與聯發科（MediaTek）取得一項重要設計導入，進一步驗證我們持續擴展的電源平台。這也是GF與這家長期客戶首次取得電源管理IC的設計導入。

In addition, we continue to strengthen our position with next-generation augmented reality wearables at a leading hyperscaler, winning a new design for micro LED display back plans. Finally, Home Industrial IoT represented approximately 19% of second-quarter total revenue. Revenue increased 30% sequentially and 10% year over year.

此外，我們也持續強化在某家領先超大規模業者（hyperscaler）的下一代擴增實境穿戴裝置布局，贏得一項micro LED顯示背板（back planes）的新設計。最後，家用與工業物聯網約占第二季總營收的19%。營收較前一季成長30%，較去年同期成長10%。

In the second quarter, IoT revenue growth marked the fastest year-over-year growth since 2022, driven by a breadth of demand for applications across AI-enabled image processing, healthcare wearables, and next-generation MCUs for edge AI compute. As inventory normalizes, customer demand signals improve and the next generation of production ramps commenced in the second half of the year, we expect our revenue for the Home and Industrial IoT end market to grow in the range of 10% to 15% in 2026, up notably from our prior expectations for mid-single-digit percentage growth.

第二季物聯網營收年增率為自2022年以來最快，動能來自AI賦能影像處理、醫療穿戴，以及用於邊緣AI運算的下一代MCU等多元應用需求。隨著庫存正常化、客戶需求訊號改善，以及下半年啟動下一代量產爬坡，我們預期2026年家用與工業物聯網終端市場營收將成長10%至15%，顯著高於先前預期的中個位數百分比成長。

In the second quarter, we secured three strategic chiplet design wins with Lockheed Martin on our FinFET and FDX platforms, creating a foundational aerospace and defense chiplet ecosystem that further extends GF leadership as a trusted US foundry. We also expanded our relationship with Microchip with a meaningful design win on our FinFET platform, another notable proof point for the growth of our embedded compute and Edge AI offerings.

第二季，我們在FinFET與FDX平台上與洛克希德馬丁（Lockheed Martin）取得三項策略性chiplet設計導入，建立基礎性的航太與國防chiplet生態系，進一步延伸GF作為值得信賴的美國晶圓代工廠之領導地位。我們也擴大與Microchip的合作關係，在FinFET平台上取得具規模的設計導入，這也是我們嵌入式運算與邊緣AI產品成長的另一項重要佐證。

Moving now to other key financial performance metrics in the quarter. In the second quarter, we delivered gross profit of $534 million, which translates into approximately 29.9% gross margin, above the high end of the guidance range and up 470 basis points year over year. A richer mix of manufacturing and Technology Services revenue, structural improvements in manufacturing costs, and improved utilization all contributed to favorable year-over-year margin expansion. R&D for the quarter was $144 million and SG&A was $92 million.

接著談本季其他關鍵財務績效指標。第二季毛利為5.34億美元，對應毛利率約29.9%，高於財測區間上緣，且較去年同期提升470個基點。製造與技術服務營收組合更佳、製造成本的結構性改善，以及產能利用率提升，皆促成年增毛利率擴張。本季研發費用為1.44億美元，銷售、一般及行政費用（SG&A）為9,200萬美元。

Total operating expenses of $236 million were up 16% quarter over quarter and represented approximately 13% of total revenue. We delivered operating profit of $298 million for the quarter at an operating margin of 16.7%, above the midpoint of our guided range and up 140 basis points from the prior year period.

總營業費用為2.36億美元，較前一季增加16%，約占總營收的13%。本季營業利益為2.98億美元，營業利益率16.7%，高於財測區間中點，且較去年同期提升140個基點。

Second-quarter net interest income was $9 million. Other expense was $12 million, and we incurred tax expense of $39 million in the quarter. We delivered second-quarter net income of approximately $256 million, an increase of approximately $22 million from the prior year period. Diluted earnings of $0.46 per share was at the high end of the guidance range based on a fully diluted share count of approximately 556 million shares.

第二季淨利息收入為900萬美元。其他費用為1,200萬美元，本季所得稅費用為3,900萬美元。我們第二季淨利約2.56億美元，較去年同期增加約2,200萬美元。以完全稀釋後股數約5.56億股計算，稀釋後每股盈餘為0.46美元，位於財測區間上緣。

Let me now provide some key cash flow and balance sheet metrics. Cash from operations in the second quarter was $405 million. Second-quarter CapEx net of proceeds from government grants was $408 million or roughly 23% of revenue. Adjusted free cash flow for the quarter was negative $3 million, as indicated in our prior quarter's guidance.

接下來提供幾項關鍵現金流與資產負債表指標。第二季營運活動現金流為4.05億美元。第二季資本支出（CapEx）扣除政府補助款收入後為4.08億美元，約占營收的23%。本季調整後自由現金流為負300萬美元，符合我們在上一季財測中的指引。

At the end of the second quarter, our combined total of cash, cash equivalents, and marketable securities stood at approximately $3.3 billion. Our total debt was $1.1 billion, and we also have a $1 billion revolving credit facility which remains undrawn.

在第二季末，我們合併後的現金、約當現金及有價證券總額約為33億美元。我們的總負債為11億美元，此外我們還有一筆10億美元的循環信用額度，目前仍未動用。

On July 14, we paid GF first-ever quarterly cash dividend of $0.12 per share, an important milestone that reflects both the progress we have made in strengthening the business and our confidence in its future cash-generating capacity. Supported by a strong balance sheet and disciplined capital allocation framework, we remain committed to investing in profitable growth while returning excess cash to shareholders.

7月14日，我們支付了GF史上首次季度現金股利，每股0.12美元；這是一個重要里程碑，反映我們在強化業務方面所取得的進展，以及我們對未來現金創造能力的信心。在強健的資產負債表與嚴謹的資本配置框架支持下，我們仍致力於投資於具獲利性的成長，同時將多餘現金回饋給股東。

As outlined at our Investor Day, our objective is to return up to 50% of trailing 12-month non-IFRS adjusted free cash flow after investments through a combination of dividends and share repurchases over time. Pursuant to this strategic objective, I'm pleased to announce that our Board of Directors approved a quarterly cash dividend of $0.12 per share payable on October 9, 2026, and to shareholders of record as of September 23, 2026.

如同我們在投資人日所說明的，我們的目標是：在完成投資後，透過股利與庫藏股回購的組合，隨時間推進返還最高達過去12個月非IFRS調整後自由現金流的50%。依據此一策略目標，我很高興宣布董事會已核准每股0.12美元的季度現金股利，將於2026年10月9日支付，並發放給截至2026年9月23日登記在冊的股東。

In addition, approximately $100 million remains under the share repurchase authorization approved by our Board of Directors and we expect to be flexible with the deployment of the remaining authorized amount.

此外，董事會核准的股票回購授權項下仍約有1億美元額度，我們預期在運用剩餘核准額度時將保持彈性。

Next, let me provide you with our outlook for the third quarter of 2026. We expect total GF revenue to be $1.885 billion, plus or minus $25 million. We expect gross margin to be approximately 30.5%, plus or minus 100 basis points. which at the midpoint reflects approximately 450 basis points of year-over-year expansion.

接下來，我提供我們對2026年第三季的展望。我們預期GF總營收為18.85億美元，上下浮動2,500萬美元。我們預期毛利率約為30.5%，上下浮動100個基點；以中位數計，代表年增約450個基點的擴張。

Excluding share-based compensation, we expect total operating expenses to be $260 million, plus or minus $10 million. We expect operating margin in the range of 16.7%, plus or minus 170 basis points. At the midpoint of our guidance, we expect share-based compensation to be approximately $76 million of which roughly $18 million is related to cost of goods sold. We expect net interest and other income for the quarter to be between $3 million and $11 million, and income tax expense to be between $28 million and $52 million.

排除股份基礎給付後，我們預期總營業費用為2.60億美元，上下浮動1,000萬美元。我們預期營業利益率介於16.7%，上下浮動170個基點。以我們指引的中位數計，我們預期股份基礎給付約為7,600萬美元，其中約1,800萬美元與銷貨成本相關。我們預期本季淨利息及其他收益介於300萬至1,100萬美元之間，所得稅費用介於2,800萬至5,200萬美元之間。

Based on a fully diluted share count of approximately 556 million shares, we expect diluted earnings per share for the third quarter to be $0.51, plus or minus $0.05. Now let me provide an update on some broader financial drivers as we evolve the mix of our business and aim to deliver the growth model set out at our recent Investor Day.

以完全稀釋後股數約5.56億股計算，我們預期第三季稀釋每股盈餘為0.51美元，上下浮動0.05美元。接著我將就一些更廣泛的財務驅動因素提供更新，因為我們正調整業務組合，並力求實現近期投資人日所提出的成長模型。

With respect to pricing, we're encouraged by the improving industry dynamics as well as the evolving mix of our business towards highly accretive technologies. In addition to these positive mix shifts, in the second quarter, we implemented pricing increases in partnership with our customers across several technology corridors.

就定價而言，我們對產業動態的改善以及我們業務組合朝向高度增益（highly accretive）技術演進感到鼓舞。除了這些正向的組合轉變外，在第二季，我們也與客戶合作，於多個技術走廊（technology corridors）實施了漲價。

Following the satisfactory conclusion of these customer conversations, we expect the pricing adjustments to be reflected in revenue commencing in 2027. The magnitude of these pricing increases varies by end market and technology and contemplates the differentiated value we provide, the ongoing supply and demand dynamics, and the inflationary absorption across our industry in recent years.

在與客戶的討論圓滿結束後，我們預期這些價格調整將自2027年起反映在營收中。這些漲價幅度會因終端市場與技術而異，並考量我們所提供的差異化價值、持續的供需動態，以及近年來整個產業對通膨的吸收情況。

Conversations with our customers have been very constructive, and we'll continue to assess pricing for 2027 and through the second half of 2026. With respect to operating expenses, consistent with the strategic updates we set out at our Investor Day in May, we believe that R&D will rise as a percentage of revenue as we integrate recent acquisitions and accelerate our R&D capabilities to support key growth opportunities.

與客戶的對話非常具建設性，我們將持續評估2027年的定價，以及2026年下半年期間的定價。就營業費用而言，與我們在5月投資人日所提出的策略更新一致，我們認為隨著整合近期收購並加速研發能力以支援關鍵成長機會，研發費用占營收比重將上升。

We've strengthened our portfolio of capabilities through the acquisitions of the Synopsys ARC IP business in June as well as the IVR business from Potion Technologies in early July, adding critical R&D, IP, and engineering resources.

我們透過6月收購Synopsys ARC IP業務，以及7月初自Potion Technologies收購IVR業務，強化了我們的能力組合，新增關鍵的研發、IP與工程資源。

Following these acquisitions, we now expect quarterly operating expenses in the second half of 2026 to be consistent with our third-quarter guidance as we accelerate critical R&D investments while ramping talent and capabilities intended to support key growth opportunities across the AI data center, physical AI, quantum computing, and advanced packaging.

完成這些收購後，我們目前預期2026年下半年每季營業費用將與第三季指引一致，因為我們將加速關鍵研發投資，同時擴增人才與能力，以支援AI資料中心、實體AI（physical AI）、量子運算與先進封裝等領域的關鍵成長機會。

These timely and necessary investments are targeted to accelerate our technology road map, deepen our customer engagements, and expand future growth opportunities in the years ahead. Moving now to tax where we expect an effective tax rate in the mid-teens percentage range for the full year of 2026. And principally due to the expected geographical mix of wafers shipped in the second half of this year.

這些及時且必要的投資，旨在加速我們的技術路線圖、深化客戶合作，並在未來數年擴大成長機會。接著談到稅務，我們預期2026年全年有效稅率將落在百分比中十位數（mid-teens）的區間。主要原因在於今年下半年預期出貨晶圓的地理組合。

Finally, for the full-year 2026, we continue to expect an adjusted free cash flow margin of approximately 10%. In conclusion, I'd like to thank our global teams for their continued commitment and diligent execution towards our strategic goals. GF drove another quarter of meaningful year-over-year margin expansion and achieved new second quarter records across a range of growth and profitability metrics.

最後，就2026年全年而言，我們仍預期調整後自由現金流利潤率約為10%。總結來說，我要感謝我們全球團隊持續投入並勤勉執行，以達成我們的策略目標。GF再度實現一季具意義的年增毛利率擴張，並在多項成長與獲利指標上創下第二季新紀錄。

Our strategic initiatives and investments executed over the last year are demonstrating good momentum across the end markets that we serve. And the continued mix shift in our business is driving improved diversification across our end market portfolio. Looking ahead, we intend to continue executing towards a richer mix of business targeting continued structural cost improvements and improved manufacturing productivity, all of which we believe are forming a strong foundation for increasing shareholder value in the years ahead.

過去一年我們所執行的策略性舉措與投資，正在我們所服務的終端市場展現良好動能。而我們業務持續的組合轉變，也正推動終端市場組合的多元化提升。展望未來，我們打算持續推進更優質的業務組合，目標是持續的結構性成本改善與製造生產力提升；我們相信這些正為未來數年提升股東價值奠定堅實基礎。

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

接下來，我們開放問答。接線員。

(Operator Instructions) Chris Caso, Wolfe Research.

(接線員指示) Chris Caso，Wolfe Research。

I guess the first question is about the common data center growth and some of the capacity expansion that's occurring in that segment. And I know while it's growing strongly, you're capacity constrained. Can you help us with to the extent you can, timing and magnitude of that capacity expansion? When does additional capacity come online? And I did also notice that you received, I think it was a 300 million chips at grant for the Silicon Photonics expansion. Can you speak to how that helps to defray some of the net CapEx for that?

我想第一個問題是關於一般資料中心的成長，以及該區隔正在進行的一些產能擴張。我知道雖然成長很強勁，但你們受到產能限制。在你們可分享的範圍內，能否協助說明該產能擴張的時程與規模？新增產能何時上線？另外我也注意到你們似乎獲得了一筆用於矽光子擴產的補助，我想是3億顆晶片的補助。你們能談談這如何有助於抵減部分淨資本支出（CapEx）嗎？

Yeah. Thank you, Chris. So I'll kick us off on that. So in terms of capacity, as you say, demand has been strengthening across basically all data center applications. We're feeling that very strongly in the optical networking space that particularly pulls on silicon photonics and silicon germanium. But we're also seeing it in other parts of the business starting to pick up, including areas like power. So our strategy will be to add capacity in those areas.

好的。謝謝你，Chris。我先來回答這題。就產能而言，如你所說，需求在幾乎所有資料中心應用上都在增強。我們在光網路領域感受特別明顯，尤其帶動對矽光子與矽鍺（SiGe）的需求。但我們也看到業務的其他部分開始回升，包括電源等領域。因此，我們的策略將是在這些領域增加產能。

One advantage for us is we're building that capacity out within our existing fab footprint, and we have, let's say, ample fab footprint today to ramp capacity relatively quickly. By day one of the contributors to us upping our full-year view about our CID end market is actually our confidence about bringing that capacity on and driving factory-level productivity improvements to be able to get basically wafers out through the back half of this year and even further into 2027. So we feel good about the ability to meet that growth with additional capacity expansion.

對我們而言的一項優勢是，我們是在既有晶圓廠（fab）版圖內擴建這些產能，而且以目前來說，我們的fab版圖相當充足，能夠相對快速地提升產能。順帶一提，我們之所以提高對全年CID終端市場的展望，其中一個因素其實是我們對導入這些產能、並推動工廠層級生產力改善的信心，從而在今年下半年、甚至延伸到2027年，基本上都能把晶圓順利出貨。因此，我們對透過額外的產能擴張來滿足該成長需求感到樂觀。

Maybe I'll turn to the $300 million partnership with the US government. We couldn't be more excited about this. I think it's really important to bear in mind that the shift to optical networking is very much a secular shift, and we see this only at the very early innings of penetrating the data center. We've spoken in the past about 70% of data center links being optical by 2030, I think every piece of evidence today points to that being perhaps even conservative relative to what's happening, including the penetration, not just to scale out but also scale up networking.

也許我會轉到與美國政府的 3 億美元合作夥伴關係。我們對此感到無比振奮。我認為非常重要的一點是要記住，向光學網路的轉移是一個非常典型的長期結構性趨勢，而我們看到這在資料中心的滲透仍然處於非常早期的階段。我們過去談到到 2030 年資料中心連結有 70% 會是光學的；我認為今天的每一項證據都顯示，考量到正在發生的情況（包括滲透率，不僅是 scale out，也包括 scale up 的網路），這個數字甚至可能偏保守。

And so look, we're very excited about the prospects of optical networking and within that Silicon Photonics. But a lot of what that will require is higher performance technologies in the future. And so what we announced really has three components continuous innovation at the PIC level, right? So improved modulated technology so we can go to 400 gig per lane and beyond. New materials.

所以你看，我們對光學網路的前景非常興奮，而其中也包括矽光子。但其中很大一部分將需要未來更高效能的技術。因此，我們所宣布的內容其實有三個組成部分：在 PIC 層級的持續創新，對吧？也就是改良調變技術，讓我們能夠達到每通道 400G 甚至更高。以及新材料。

At some point, we will introduce new materials into the system, think about barium titan a thin film, lithium niobate, think about indium phosphide, all areas of technological innovation to produce higher performing systems and more integrated systems going forward.

在某個時間點，我們會把新材料導入系統，例如思考鈦酸鋇薄膜、鈮酸鋰，也思考磷化銦；這些都是技術創新的領域，用以打造未來更高效能、且更高度整合的系統。

And then the last piece was extraordinarily important, particularly for near and co-packaged optics is packaging. And so being able to build those integrated optic engines for both those applications using our scale solution also requires continued capability and capacity. So that partnership allows us to accelerate that. And we're very grateful to have the US government as a strong, let's say, partner in our corner supporting that innovation happen right here in the US.

然後最後一塊也極其重要，特別是對於近端（near）與共封裝（co-packaged）光學而言，就是封裝。因此，能夠運用我們的規模化解決方案，為這兩類應用打造整合式光學引擎，也需要持續的能力與產能。所以這項合作夥伴關係讓我們能夠加速推進。我們也非常感謝美國政府作為強而有力的、可以說是在我們這一邊的合作夥伴，支持這些創新就在美國本土發生。

Do you have a follow-up question?

你有追問嗎？

I do. A follow-up. I'll ask on gross margins. And can you speak to what's the driver of the gross margin expansion as you go into the third quarter in terms of utilization mix and pricing and perhaps give us some color on the trajectory of gross margins into next year, particularly in light of some of your comments with regard to pricing?

有的。我有個追問。我想問毛利率。你能否談談進入第三季時毛利率擴張的驅動因素是什麼，從產能利用率、產品組合與定價的角度；另外也請你提供一些對明年毛利率走勢的看法，特別是考量到你們對定價的部分評論？

Yes. Very happy to you, Chris. And look, I will start by saying that we're very encouraged by the continued progression and expansion in our gross margins. And I think it's a continued reflection of the progress that we've seen during the course of this year, we had almost 500 basis points of more expansion in the second quarter.

可以。很樂意回答你，Chris。我先說，我們對毛利率持續改善與擴張的進展感到非常鼓舞。我認為這持續反映了我們今年以來看到的進展；第二季我們的毛利率擴張接近 500 個基點。

We had over 500 basis points of margin expansion in the first quarter. And as I said in my prepared remarks, if you take the midpoint of our inferred guide, that implies about another 450 basis points of margin expansion. So this is really playing to the thesis and the levers that we discussed at our Investor Day just a couple of months ago now. And frankly, it's falling through on a relatively healthy basis when you look at the revenue take revenue a year ago and compare it to the same period this year, about $100 million of revenue growth. Look at that adjusted gross profit and you see about $100 million of gross falling through as well.

第一季我們的毛利率擴張超過 500 個基點。而且如我在事先準備的發言中所說，如果以我們推導出的財測指引中點來看，這意味著大約還有另外 450 個基點的毛利率擴張。所以這確實呼應了我們在幾個月前投資人日所討論的論點與槓桿。坦白說，當你看營收的落地情況——把一年前的營收與今年同期間相比，大約增加了 1 億美元的營收成長。再看調整後毛利，你也會看到大約同樣有 1 億美元的毛利落地。

So we're very encouraged by that relative fall through to the underlying gross margin as it relates to our revenue growth. And you sort of touched on it a little bit in your question, Chris. Mix has been a big and continues to be a big driver of that. And the way to think about mix is twofold. It's mix from a manufacturing services point of view, and its mix from a Technology Services point of view as well. Both of those have been encouraging tailwinds for us, particularly when you look at the relative strength and growth within some of those end markets, which I touched on around it comes in for in data center being highly accretive to those targets. Utica Technology Services revenue up a little under $40 million year over year. That's about one point of benefit that comes through there.

因此，隨著營收成長帶動的底層毛利率落地（fall-through）表現，我們感到非常鼓舞。而你在問題中也稍微提到，Chris。產品組合一直是、而且仍然是主要驅動因素。而理解組合可以從兩個面向來看。一是從製造服務（manufacturing services）的組合角度，二是從技術服務（Technology Services）的組合角度。這兩者對我們而言都是令人鼓舞的順風，特別是當你看到某些終端市場的相對強勁與成長——我剛才也提到，資料中心的成長對這些目標具有高度增益（highly accretive）。Utica 的技術服務營收年增略低於 4,000 萬美元。這大約帶來 1 個百分點的貢獻。

So the combination of the mix across manufacturing and Technology Services has been encouraging. We expect that to continue. As I said previously, productivity within our manufacturing sites and driving structural cost improvements has been a big driver as well.

因此，製造與技術服務兩端的組合效應相加，表現令人鼓舞。我們預期這會持續。如我先前所說，我們製造據點的生產力提升，以及推動結構性成本改善，也是一個重要驅動因素。

Utilization, we were in the second-quarter high 80s% from a utilization point of view. So we still feel we've got a good amount of our existing capacity to be able to grow into and see positive margin movements over time. And frankly, all of that is against some of the benefits we had in the year-ago period. We had things like liquidated damages in early 2025, which have fallen out. So again, it sort of reflects the strong growth we've seen from a margin point of view.

產能利用率方面，第二季我們的利用率在 80% 高段。所以我們仍然認為，現有產能還有相當空間可以承接成長，並隨時間帶來正向的毛利率變動。而且坦白說，這一切還是在對比去年同期的一些有利因素之下。例如我們在 2025 年初有一些違約金（liquidated damages）收入，現在已經不再有了。所以再一次，這也反映了我們從毛利率角度所看到的強勁成長。

We bumped up on that 30% target margin that we said we were looking to solve for at the exit of 2026. In our second quarter, we're above that in the third-quarter guidance. So expectation now for the full year is that we should be at about 30 points of gross margin for the full year rather than just that exit target that we had at the beginning of the year. I hope that helps, Chris.

我們已經觸及我們所說、希望在 2026 年底達成的 30% 目標毛利率。第二季我們已經高於該水準，而第三季指引也在其上。因此，現在對全年而言，我們的預期是全年毛利率應該約在 30 個點左右，而不只是年末退出時點（exit）的目標——這是我們年初的設定。希望這能幫到你，Chris。

Krish Sankar, TD Cowen.

TD Cowen 的 Krish Sankar。

I just wanted to first follow up on the Silicon Photonic CID year-over-year growth, almost doubling from your prior outlook. I'm just kind of curious what changed in the last three months that the outlook has been revised almost materially higher? And any color you can give on your PIC solutions compared to your two competing foundries. Then I had a quick follow-up on Quantum too.

我想先追問一下矽光子 CID 的年增成長，幾乎比你們先前的展望翻倍。我有點好奇，過去三個月到底發生了什麼，讓展望幾乎大幅上修？另外，你能否談談你們的 PIC 解決方案相較於另外兩家競爭晶圓代工廠的情況？然後我也有一個關於量子（Quantum）的快速追問。

Yeah. Great. Thank you very much, Krish. Look, Photonics remains a very strong driver for us. I think every customer meeting is all about what more can we do, how much faster can we go? There is clearly strong demand today. And by the way, in a market like this, we don't just validate that demand with our direct customers. We spend time throughout the ecosystem, including with the big hyperscalers, and you've seen many of them are supporting a lot of what we're doing here in the US and around the world.

好的。很好。非常感謝你，Krish。你看，光子技術仍然是我們非常強勁的成長驅動力。我想每一次客戶會議都在談：我們還能做更多什麼？我們能多快？目前顯然有強勁需求。順帶一提，在這樣的市場裡，我們不只是透過直接客戶來驗證需求。我們也會花時間走訪整個生態系，包括大型超大規模業者（hyperscalers）；你也看到其中許多正在支持我們在美國以及全球各地所做的很多事情。

So we're validating the demand, and we believe it's very real today and durable going forward. That's giving us confidence to continue to invest. We'll increase our investments in Photonics capacity. And as I mentioned earlier, there's nothing our factories love more than being challenged to get more output literally every single week. We're calling in from (inaudible) New York right now, and the factory is hard at work producing more wafers every day for these oversold corridor.

因此我們正在驗證需求，而且我們相信目前的需求非常真實，且未來也具持續性。這讓我們有信心持續投資。我們會增加在光子產能上的投資。而且如我先前提到，沒有什麼比每週都被要求把產出再提高更能讓我們的工廠感到興奮的了。我們現在是從（聽不清）紐約連線，而工廠正全力運作，每天為這些供不應求的走廊（corridor）生產更多晶圓。

So I think very conviction about continuing to grow Silicon Photonics. But yet, we're still at the very beginning of this, and that those growth targets we set for kind of end of 2018 and through 2030, I'd say today, we are very much on track and potentially ahead of those targets in terms of our Silicon Photonics growth.

所以我認為，我們對持續擴大矽光子成長非常有信念。但同時，我們仍然處於非常早期的階段；而我們為 2018 年底到 2030 年所設定的那些成長目標，我會說以今天來看，我們在矽光子成長方面非常符合進度，甚至可能領先於那些目標。

Got it. Very helpful, Tim. And then a quick follow-up on Quantum. You recently got the $375 million grant. Can you talk a little bit about the opportunity set there and whether being the Quantum commercialization curve. And have you seen more interest or share gains given Ion just recently closed acquisition of Sky Water?

了解。非常有幫助，Tim。接著我想快速追問一下 Quantum。你們最近拿到 3.75 億美元的補助金。能否談談那邊的機會版圖，以及你們是否正處於 Quantum 商業化曲線上。另外，鑑於 Ion 最近剛完成收購 Sky Water，你們是否看到更多興趣或市占提升？

Yes. So thank you for that question. I mean, quantum is extremely exciting, and there's a few reasons behind that. I think one is talking to now basically all the players in the sector, everyone is facing the same kind of transition. But this is not a -- can I prove it in a lab discussion. This is can I scale to high-volume manufacturing.

是的。謝謝你的提問。我的意思是，量子非常令人振奮，背後有幾個原因。我認為其中之一是，現在基本上和這個領域所有玩家都在交流，每個人都面臨同樣類型的轉型。但這已經不是——我能不能在實驗室證明它——的討論。而是我能不能擴展到高量產製造。

And so the conversations we have, and in our announcement, we had both dedicated quantum players, but also hyperscalers and others comments and support that. initiative, the conversations are similar because it's all about transitioning to that high-volume scale and cracking different problems that they've proven at lab scale that need to prove now as they transition to high volume.

因此，我們進行的對話，以及在我們的公告中，我們同時獲得了專注量子的玩家、也有超大規模業者（hyperscalers）與其他夥伴的評論與支持該項倡議；這些對話很相似，因為核心都是要轉向高量產規模，並解決他們已在實驗室規模驗證、但在邁向量產時必須再度證明與攻克的不同問題。

Since that announcement, we've launched four significant new engagements with some -- a subsidy of the players that are supporting us in that announcement. And we see that ramp continuing. Those engagements have some common features. For example, some of the stuff we're doing around cryogenic CMOS for readout, ICs for different modalities. That's very exciting because it builds on existing platforms that we have. But also there are those who have very specific requirements.

自那次公告以來，我們已啟動四項重要的新合作案，合作對象包括——公告中支持我們的部分玩家。而且我們看到這樣的爬坡仍在持續。這些合作案有一些共同特徵。例如，我們在用於讀出（readout）的低溫 CMOS、以及適用不同模態的 IC 等方面所做的一些工作。這非常令人振奮，因為它建立在我們既有的平台之上。但同時也有一些客戶有非常特定的需求。

And that's where also I'm quite excited about the technology benefits of us investing in Quantum. And so I'll give you an example. A couple of -- actually more than two, three, or four players are doing things linked to the Photonic side in that Quantum Solution that has excellent read-across for us in our long-term Silicon Photonics road map. So think of that as a very synergetic with what we're doing in that space.

而這也正是我對我們投資 Quantum 所帶來的技術效益感到相當興奮的地方。我舉個例子。有幾家——其實不只兩家，可能三家或四家——玩家正在做與 Quantum 解決方案中「光子（Photonic）」面向相關的事情，這對我們長期的矽光子（Silicon Photonics）路線圖有非常好的延伸與對照效益。所以你可以把它視為與我們在該領域所做的事情高度協同。

And so Quantum is actually reinvigorating a number of our long-term technology road maps even further and faster than otherwise it would be happening. So that's very positive. As Sam mentioned in the prepared remarks, we'll see the financial profile of Quantum in our Technology Services revenue this year and definitely into next year. Think of it more medium term as a kind of call option on the scale to high-volume module manufacturing. Too early to call exactly when those ramps will happen. But clearly, that is the objective of these players that are engaging with us is develop and crack the solutions and then scale them to high volume together.

因此，Quantum 其實正在讓我們多項長期技術路線圖更進一步、也更快地被重新注入動能，否則不會這麼快發生。這是非常正面的。如同 Sam 在事先準備的發言中提到的，我們今年會在 Technology Services 的營收中看到 Quantum 的財務表現，並且肯定會延續到明年。從中期來看，可以把它想成是一種「買權（call option）」，押注未來擴展到高量產模組製造的規模化機會。現在要精準判斷那些爬坡何時發生還太早。但很明顯，這些與我們合作的玩家目標就是共同開發並攻克解決方案，然後一起把它們擴展到高量產。

Karl Ackerman, BNP Paribas.

Karl Ackerman，法國巴黎銀行（BNP Paribas）。

Two, if I may. Tim, you spoke about the three-fourths of growth, including Photonics, Quantum, IP, and custom silicon. But could you speak to the revenue and OpEx contribution of the ARC and Foton Technologies IVRT, the September outlook? And also, if you zoom out, could you double click on the rationale for these deals and maybe any early customer design engagements you've seen to date?

如果可以的話，我有兩個問題。Tim，你談到成長的四分之三來自 Photonics、Quantum、IP 與客製化矽。但你能否談談 ARC 與 Foton Technologies IVRT 對營收與營業費用（OpEx）的貢獻，以及 9 月的展望？另外，如果拉遠一點看，你能否更深入說明這些交易的策略理由，以及截至目前你們看到的任何早期客戶設計合作案？

Yeah. So let's -- I'll talk about rationale, and then I'll let Sam comment on how we're thinking about revenue for this year. So we've been very focused in our acquisition strategy on identifying capabilities that our customers value and that links to our manufacturing road map but also linked to what they tell us around gaps that the industry today is not meeting. So let me take kind of the MIPS and Synopsys story first, and I'll come back to Foton both very exciting in their own ways.

好的。那我們——我先談策略理由，然後我會請 Sam 補充我們對今年營收的看法。我們在併購策略上一直非常聚焦：找出客戶重視、且能連結到我們製造路線圖的能力，同時也對應他們告訴我們的、目前產業尚未滿足的缺口。我先從 MIPS 與 Synopsys 的故事談起，之後再回到 Foton——兩者各自都非常令人振奮。

Customer feedback on MIPS and then ARC has been excellent. I spent a lot of time personally with customers, especially since we've closed the ARC deal. And by the way, with that came 300 customers. Some of those were not GF customers before. So it gives us also new customers to engage with on those road maps. And these are very strategic discussions because these are about future architectures for their processor solutions. How can they add AI at the Edge? How can they do on device inference in the automotive space, the industrial space, the robotics space?

客戶對 MIPS、以及後來的 ARC 的回饋都非常好。我個人花了很多時間和客戶交流，尤其是在我們完成 ARC 交易之後。順帶一提，這也帶來了 300 家客戶。其中有些先前並不是 GF 的客戶。因此也讓我們有新客戶可以在這些路線圖上展開互動。而且這些都是非常具策略性的討論，因為它們關乎其處理器解決方案的未來架構。他們如何在邊緣端加入 AI？他們如何在車用、工業、機器人等領域做裝置端推論（on-device inference）？

And so it's bringing some really interesting discussions to bear, and it allows us as GF to engage much earlier in that design conversation than we would if it was just a conversation about manufacturing capacity and manufacturing process technology. It has another benefit, which is that it's also giving us very, very early input into our manufacturing road map.

因此，這帶來了一些非常有意思的討論，也讓我們 GF 能比以往更早介入設計端的對話；若只是談製造產能與製程技術，我們不會這麼早參與。它還有另一個好處：也讓我們能非常早期地獲得對我們製造路線圖的輸入。

And so now you have this let's say, symbiosis internally that we have an internal customer for what we're doing that is actually challenging us to push performance of next-generation technologies, particularly in CMOS business to the next level. So you're thinking about how do you do lower power inference at the edge and so on. So I'd say early, of course, for both of these, but very encouraging.

所以現在內部形成一種——姑且說——共生關係：我們正在做的事情在公司內部也有「內部客戶」，而這些內部客戶其實在挑戰我們把下一代技術的效能推到更高水準，特別是在 CMOS 業務上。例如你會思考如何在邊緣端做到更低功耗的推論等等。所以我會說，對這兩項來說當然都還在早期，但非常令人鼓舞。

We've talked about some of the early wins and partnerships in spaces like defense with Lockheed Martin, automotive players like Infineon, but there are very many more in the pipeline. So very encouraging for our IP software and custom silicon strategy. To talk briefly about IDR and Foton, we've had the chance to work with the Potion team for many years now as the IVR category has started to become more and more important. The way you should think about that is IVR is to power what CPO is to Photonics, right?

我們也談過一些早期的勝利與合作夥伴關係，例如國防領域與洛克希德馬丁（Lockheed Martin）、車用領域與英飛凌（Infineon）等，但管線中還有非常多。因此對我們的 IP 軟體與客製化矽策略而言非常令人鼓舞。簡要談一下 IDR 與 Foton：我們多年來一直有機會與 Foton 團隊合作，因為 IVR 類別開始變得越來越重要。你可以這樣理解：IVR 對電源而言，就像 CPO 對 Photonics 一樣，對吧？

Think about how do you build a much more wafer-level integrated to min to deliver power closer to chip and to be able to do things that today exist in much more kind of traditional power modules, much more like the industrial and automotive power modules of today.

想想看：如何打造更偏向晶圓層級整合的方案，以把電力更靠近晶片供應，並能做到一些目前仍存在於較傳統電源模組中的功能——更像是今天工業與車用的電源模組那樣。

This is moving to a much more kind of, let's say, wafer-level solution that is higher and higher performance. that is essential for next-generation data center power, given how hungry these XPUs PPUCPUs are all for power. Portion team, very engaged with a number of our existing customers. So it's a very natural transition to bring that team on board, accelerate those engagements and again, early back from the likes of existing kind of fabless and IDMs, but also hyperscalers who are engaged there are very, very positive, bringing that capability into GF.

這正在走向一種更偏向——姑且說——晶圓層級的解決方案，效能越來越高。這對下一代資料中心供電至關重要，因為這些 XPU、PPU、CPU 都非常耗電。Foton 團隊與我們許多既有客戶互動非常深入。因此把該團隊納入公司、加速這些合作案，是非常自然的轉變；而且來自既有無晶圓廠（fabless）與 IDM 客戶、以及參與其中的超大規模業者的早期回饋都非常、非常正面，因為這等於把這項能力帶進 GF。

And Karl, maybe if I just jump on the second part of your question as it relates to some of the financial profile. And for all the reasons that Tim outlined, these investments are incremental, they're strategic. And actually, in the case of the MIPS acquisition as well as the ARC IP business from Synopsys, they are revenue generative from day one.

Karl，或許我也補充你問題的第二部分，關於一些財務輪廓。基於 Tim 所概述的所有原因，這些投資都是增量的、具策略性的。而且實際上，就 MIPS 的收購以及從 Synopsys 取得的 ARC IP 業務而言，它們從第一天起就能帶來營收。

At the outside of this year, we expected that we'd be seeing about $60 million to $100 million of incremental revenue through from that MIPS acquisition during the course of 2026. That remains the expectation. But what's changed over the course of the last quarter is that we closed the acquisition of the Synopsys ARC IP business.

以今年年底的角度來看，我們原先預期在 2026 年期間，透過這筆 MIPS 收購可帶來約 6,000 萬至 1 億美元的增量營收。這仍然是我們的預期。但在過去一季中改變的是：我們完成了對 Synopsys ARC IP 業務的收購。

So the midpoint of that range, as you can kind of infer from my prepared remarks has moved up from the $60 million to $100 million to $100 million to [$120 million, $30 million] of revenue growth our expectation in terms of the skew of that incremental revenue from that recent acquisition is sort of on-third, two-thirds skewed from third quarter and fourth-quarter perspective.

因此，該區間的中點，如你從我事先準備的發言中大致可以推知，已從原先的 6,000 萬至 1 億美元，上移至 1 億美元至 [1.2 億美元、3,000 萬美元] 的營收成長；就那筆近期收購所帶來的新增營收在季度間的分布而言，我們的預期大致是三分之一、三分之二，分別偏向第三季與第四季。

As it relates to the R&D and the fall-through ultimately through to EPS. Look, these are R&D intensive businesses. They're also highly accretive from a gross margin point of view. Overall, we expect that the increased in OpEx and particularly the acceleration from an R&D point of view to largely be covered by that incremental revenue we see coming through both of those acquisitions. So we feel quite good about it from that perspective.

至於研發以及最終流入到 EPS 的落差（fall-through）。你看，這些都是研發密集型的業務。同時，從毛利率角度來看，它們也具有很高的增厚性（accretive）。整體而言，我們預期營業費用（OpEx）的增加，特別是研發面的加速，將在很大程度上由我們看到的、來自這兩項收購的新增營收所覆蓋。因此，從這個角度我們感覺相當不錯。

Very clear. For my follow-up, if I may, could you discuss what portion of those seven customers on your scale platform are working on NPL or your packaged optics? And I guess how should we think about the timing of your NPL opportunity?

非常清楚。我可以追問一下嗎？你能否談談你們 scale 平台上的那七家客戶中，有多少是在做 NPL 或你們的封裝式光學（packaged optics）？另外，我們應該如何看待你們 NPL 機會的時程？

Yes. Maybe just to take a step back, and I think there's obviously a year ago, the industry wasn't talking a lot about NPL now it's talking a lot about it. I think the reason is that you see a comfort level for a number of players moving from, let's say, traditional pluggable infrastructure to NPL, and that's because they have synergies in terms of things like the SerDes.

可以。也許先退一步看；我想很明顯，一年前產業界還不太談 NPL，但現在談得很多。我認為原因在於，你看到不少業者對於從傳統可插拔（pluggable）基礎架構轉向 NPL 的接受度提高，因為在 SerDes 等方面存在協同效應。

So the system is, let's say, a smaller transition versus moving to the full co-packaged optics. We think both those form factors as well as pluggables will continue to exist in the data center. Obviously, as you get more into scale up networking, that's where you need to move to smaller form factors like NPO and CPO and to really address all of those linkages within a scale-up network. So we think all of those will continue to exist.

因此，相較於全面轉向共封裝光學（co-packaged optics），這個系統可以說是一個較小的轉換。我們認為，這兩種外形規格（form factors）以及可插拔產品，未來都會持續存在於資料中心。顯然，當你更深入到 scale-up 網路時，就需要轉向像 NPO 與 CPO 這樣更小的外形規格，才能真正涵蓋 scale-up 網路內的所有連結。所以我們認為這些都會持續存在。

Scale supports NPO and CPO. So a lot of the dimensions that go into bringing our scale solution together and EIC bonded to a pick with a micro-optic with a fiber attached unit and necessary for both NPO and for CPO. Fundamentally, the mechanical difference is that the NPO is bonded to the Board, whereas the CPO sits within the package, right?

Scale 支援 NPO 與 CPO。因此，將我們的 scale 解決方案整合起來所涉及的許多面向——例如 EIC 與 PIC 的鍵合、搭配微型光學元件與帶光纖的組件——對 NPO 與 CPO 都是必要的。從根本上說，機械上的差異在於：NPO 是鍵合到電路板（Board）上，而 CPO 則是位於封裝（package）內，對吧？

And there are differences, therefore, to let's say, more of the things like the (inaudible) architecture, as I mentioned, but left to the mechanics of how things are done. So look, we see very good momentum on NPO. I'd say the scale engagements we have cut across both. And actually, I'd say even many of the customers are doing both because they have both an NPO that they're ramping sooner and a CPO that they're ramping a little bit later. We still maintain the view that '27, we'll see the beginning of NPO ramp in '28. We'll see the beginning of CPO ramp, and that's been quite consistent over the last few quarters.

因此也會有差異，例如我提到的（聽不清）架構等，但這些更多是取決於實作的機械方式。總之，我們看到 NPO 的動能非常好。我會說，我們的 scale 專案合作同時涵蓋兩者。而且我甚至會說，很多客戶兩者都在做，因為他們有一個較早量產爬坡的 NPO，同時也有一個稍晚量產爬坡的 CPO。我們仍維持這樣的看法：在 2027 年會看到 NPO 開始爬坡；在 2028 年會看到 CPO 開始爬坡；而這在過去幾個季度一直相當一致。

Mehdi Hosseini, Susquehanna International Group.

Mehdi Hosseini，Susquehanna International Group。

I also have a couple of follow-ups on co-infrastructure. Tim, can you help us understand what is the contribution of SiGe into your overall optical revenue mix? And as we migrate to NPL and assuming the PIC itself becomes a catalyst to what extent should I expect some synergy between Sipho and SiGe? And I do have a follow-up.

我也有幾個關於共封裝基礎架構（co-infrastructure）的追問。Tim，你能幫我們理解 SiGe 在你們整體光學營收組合中的貢獻度嗎？另外，隨著我們遷移到 NPL，並假設 PIC 本身成為催化劑，在多大程度上我應該期待 Sipho 與 SiGe 之間會有一些協同效應？我還有一個追問。

Yeah, it's a great question. And let me talk about SiGe just for a little bit since you picked it out. Just to wind the clock back, right? How do we have such an important position in SiGe? IBM Microelectronics, I think, is on record for inventing SiGe.

是的，這是個很好的問題。既然你特別點到 SiGe，我先花點時間談一下。先把時間往回拉，對吧？我們為什麼能在 SiGe 上有這麼重要的地位？我想 IBM Microelectronics 有公開記錄顯示，他們發明了 SiGe。

IBM Microelectronics is part of GF today. And so we've had team members building SiGe Solutions for a long time now. And so that's always been an important part of our portfolio. What you're seeing in Siggi is that the acceleration is driven by, let's say, two trends happening at the same time. One is the shift to optical networking means you are moving more data through a different kind of link.

IBM Microelectronics 如今是 GF 的一部分。因此，我們的團隊成員長期以來一直在打造 SiGe 解決方案。所以這一直是我們產品組合中重要的一環。你在 SiGe 看到的加速成長，是由兩個同時發生的趨勢所驅動。其一是轉向光網路，意味著你會透過不同類型的連結來傳輸更多資料。

But as you push to higher and higher bandwidth what you could do previously in CMOS at say, 50 gig per lane, 100-gig per lane, at 200 gig per lane, and definitely a 400 gig per lane, you cannot do. And so what we're seeing is people are breaking out those TIAs, those transimpedance amplifiers and also incrementally also breaking out the drivers and doing them in high-performance analog solutions like SiGe.

但當你推進到越來越高的頻寬時，以前在 CMOS 上能做到的——例如每通道 50G、100G——到了每通道 200G，尤其是每通道 400G，就做不到了。因此我們看到，大家把那些 TIA（跨阻放大器，transimpedance amplifiers）拆分出來，而且也逐步把驅動器（drivers）拆分出來，改用像 SiGe 這類高效能類比解決方案來實作。

And so in a way, what we're seeing in our SiGe business is actually growth driven by both the switch to optical, but also the increased bandwidth requirements within those optical solutions. That's driving significant growth within the space. Just to give you a dimension, our SiGe business is actually larger than our Silicon Photonic business today.

所以某種程度上，我們在 SiGe 業務上看到的成長，同時受到「轉向光」以及「光學解決方案內部頻寬需求提升」兩者驅動。這正在推動該領域的顯著成長。給你一個量級概念：我們的 SiGe 業務目前其實比我們的矽光子（Silicon Photonic）業務還大。

So it's actually a meaningful part of our data center business overall. And like Silicon Photonics has very strong growth trajectories because it serves the same underlying trend. And perhaps with even that multiplier effect playing out even more so as we're expanding the performance of those pluggable technologies. You'll also find high-performance analog solutions in near and co-packaged optics as well. So those will also have components of those depending on the architecture.

因此它確實是我們整體資料中心業務中相當有意義的一部分。而且像矽光子一樣，它也有非常強勁的成長軌跡，因為服務的是相同的底層趨勢。也許當我們擴展那些可插拔技術的效能時，這種乘數效應會更加明顯。你也會在近封裝與共封裝光學中看到高效能類比解決方案。因此，視架構而定，那些也會包含其中的一些元件。

Our SiGe Solutions, we build them today in Burlington, Vermont. We're expanding capacity there. That's part of what is also increasing our output within '26 and definitely into '27. But we're also qualifying 300-millimeter SiGe in Singapore, which will bring additional capacity. The economics of 300-millimeter, which obviously is very good. but also higher performance. And so as you move to new platforms, you improve your FT or f MAX performance. And so we think we can continue that great tradition started by IBM of leading the industry in terms of our SiGe solutions for the market.

我們的 SiGe 解決方案目前在佛蒙特州伯靈頓（Burlington, Vermont）生產。我們正在擴充那裡的產能。這也是我們在 2026 年、並且在 2027 年更明顯地提升產出的一部分。但我們也正在新加坡認證（qualifying）300 毫米 SiGe，這將帶來額外產能。300 毫米的經濟性顯然非常好，同時也能帶來更高的效能。因此，當你轉向新平台時，你可以改善你的 fT 或 fMAX 效能。所以我們認為，我們可以延續 IBM 開創的優良傳統，在面向市場的 SiGe 解決方案上持續引領產業。

Okay. Great. I feel like an or just focusing on optical, but I'm going to move on. Within the com infrastructure, there's also [salarated] prom and SpaceX had their first quarterly call last night. So help us understand, right now, we're just focused on a transceiver and optical solution. But I see there's also a synergy. So what kind of a substrate for satellite com and remind us from an Analyst Day how [SATCOM] drive double-digit growth here. Hopefully, I'm in line with more assumptions.

好的。很好。我覺得我好像一直只聚焦在光學，但我會往下個主題。在通訊基礎架構（com infrastructure）方面，還有 [salarated] prom，而 SpaceX 昨晚也開了他們第一次季度電話會議。所以請幫我們理解一下：目前我們只聚焦在收發器（transceiver）與光學解決方案。但我看到這裡也有協同效應。那麼在衛星通訊（satellite com）方面是什麼樣的基板（substrate）？另外也請你提醒我們，在分析師日（Analyst Day）上提到 [SATCOM] 如何在此帶動雙位數成長。希望我的假設方向是對的。

Yes. So SATCOM look, continues to be a strong growth business for us. And the reasons for that, I think, are fairly clear. The transition to LEO deployment, really driven by SpaceX originally and a couple of other players now ramping as well. It's very clear to see.

是的。所以你看，SATCOM 對我們而言仍然是一項強勁成長的業務。我認為其原因相當清楚。向 LEO（低地球軌道）部署的轉型，最初確實是由 SpaceX 推動，現在也有其他幾家業者開始加速。這一點非常明顯。

And as a consumer, once you take a flight and you still link on the flight, you kind of don't want to go back to how it used to be. So I think it's clear why those solutions are taking effect. Remember, you're beaming a signal, 300 miles into space, a signal that normally would have gone 2 or 3 miles to a base station or a cell tower. And so you need higher performance RF, you need beamforming, you need other kind of wireless technologies.

而作為消費者，一旦你搭過飛機，且在飛行途中仍能連上網，你就會有點不想回到以前那種狀態。所以我想很清楚為什麼這些解決方案正在發揮作用。記住，你是在把訊號打到太空中 300 英里遠的地方；而這個訊號原本通常只需要傳 2 到 3 英里到基地台或行動通訊塔。因此你需要更高效能的 RF，你需要波束成形（beamforming），也需要其他類型的無線技術。

These are very core strength areas for GF. Some of these technologies are in SOI today. Some of them have SOI like characteristics that are done without SOI as well. And we're seeing more and more trajectory of new RF technologies playing into SATCOM going forward. Even areas like RF GaN, we think will play a very important role in SATCOM going forward because you're not just increasing the number of units deployed, but you're talking about faster and faster bandwidth.

這些都是 GF 的核心強項領域。其中一些技術目前已在 SOI 上。也有一些具備類似 SOI 特性的技術，即使不採用 SOI 也能實現。而我們看到，未來會有越來越多新 RF 技術的發展軌跡，將應用到 SATCOM。甚至像 RF GaN 這樣的領域，我們認為未來在 SATCOM 會扮演非常重要的角色，因為你不只是增加部署的單位數量，還在追求越來越快的頻寬。

And by the way, this is not just for consumers. You're hearing about industrial deployment, corporate deployment and so on. And there I say, when you put data centers in space, you're going to need a lot more bandwidth to bring that capability back to us for people to use. So I think very strong secular driver of growth for us. Obviously, we're starting from a relatively small base compared to other markets, given this is relatively new, but we definitely see it growing very well long term.

順帶一提，這不只是面向消費者。你也會聽到工業部署、企業部署等等。而在那裡我會說，當你把資料中心放到太空，你就需要更多頻寬把這種能力帶回來，讓人們可以使用。所以我認為這對我們而言是非常強的長期結構性成長驅動因素。當然，鑑於這相對較新，與其他市場相比我們目前的基期相對較小，但我們確實看到它長期會成長得很好。

Timothy Arcuri, UBS.

Timothy Arcuri，瑞銀（UBS）。

So Sam, I had a multipart question. So the segment guidance implies like December is up in the 10% range. So my question is, A, is that right? And then can you give us any sense for September of how the guidance shakes out by segment, even if you just give qualitative comments on that?

所以 Sam，我有一個多段式問題。分部指引看起來暗示 12 月大概有 10% 左右的成長。所以我的問題是，A，這樣理解對嗎？另外，你能否就 9 月的情況，提供一些按分部拆分後指引的概況，即使只是定性評論也可以？

Yes. Happy to take that, Tim, and you're right in terms of your overall inference. We said that certainly for the remaining part of this year, just given some of the dynamics around mobile, we think that's sort of download teens for the full year. Meanwhile, automotive, despite having a slight softer in 2Q, that was really a function of some of the customer shipment timing.

是的。Tim，很樂意回答，你的整體推論是對的。我們說過，至少在今年剩餘期間，考量到行動端的一些動態，我們認為全年大概是下滑十幾個百分點（download teens）。同時，汽車業務儘管在第二季略有走弱，那其實主要是某些客戶出貨時點的影響。

Our expectation is that come back in second half. So sort of low double-digit growth on auto is very consistent with what we've been guiding throughout this year. IoT has actually been an interesting update from our perspective. And what we're seeing is that where there has been some softness in Smart mobile, some of those customers within mobile have actually reallocated their demand into IoT, a lot of commonality between some of those connectivity applications that you see between the two.

我們預期下半年會回升。因此，汽車端的低雙位數成長，與我們今年以來一直給出的指引非常一致。IoT 從我們的角度看其實是個有趣的更新。我們看到，在智慧型行動（Smart mobile）有些疲軟之處時，行動領域的一些客戶其實把部分需求重新分配到 IoT；兩者在某些連線應用上有很多共通性。

So that's what's driving the pickup in IoT to sort of the 10% to 15% level that I indicated year-over-year. And then obviously, we spent quite a bit of time on coming for in data center, so I won't reiterate that, but that's really the change quarter-on-quarter, that movement to kind of 50% to 60% year-over-year growth. So you can sort of infer from that Tim, what that means for the third quarter into fourth quarter, I think your math is about right in terms of that implied pickup into the fourth quarter. And then look, as it relates to some of the dynamics within 3Q and 4Q, it's largely consistent with, I think, what we're expecting to see from a trend perspective and what I just outlined on the end market split as well.

所以這就是推動 IoT 年增回升到我提到的約 10% 到 15% 水準的原因。然後很明顯地，我們在資料中心的成長上已經花了不少時間說明，所以我不再重複；但那確實是季度之間的主要變化，也就是年增成長移動到大約 50% 到 60% 的區間。所以 Tim，你可以由此推算第三季到第四季的含意；就你提到第四季隱含回升的數學推算，我認為大致是對的。至於第三季與第四季內部的一些動態，整體上與我們從趨勢角度的預期，以及我剛才概述的終端市場拆分，基本一致。

And maybe, Tim, if I can just add. If we zoom out for a second and think about kind of are we seeing more and more traction in those end markets? I think we haven't spoken a lot about design wins across the board, but we continue to see record design win momentum entering into 2026 versus '25.

Tim，也許我再補充一下。如果我們把視角拉遠一點，想想我們是否在這些終端市場看到越來越多的動能？我想我們沒有花很多時間談整體的設計導入（design wins），但我們持續看到創紀錄的設計導入動能，進入 2026 年相較於 2025 年更是如此。

So again, that traction remains really strong. I think the other thing that's very important, we've worked very hard to make our manufacturing footprint as flexible as possible because there are always going to be perturbations between different end markets in terms of timing. You see obviously a very strong story in the data center. You see mobile is obviously going in a different direction given the memory shortage. We've been able to reuse capacity very well this year to enable us to capture some of those upsides given the flexibility of our manufacturing footprint.

所以這股動能依然非常強。我認為另一個非常重要的點是：我們非常努力讓製造佈局（manufacturing footprint）盡可能具備彈性，因為不同終端市場在時點上總會出現擾動。你顯然看到資料中心是一個非常強的故事。你也看到行動端因為記憶體短缺而朝著不同方向發展。今年我們能夠很好地重複利用產能，憑藉製造佈局的彈性，讓我們得以掌握其中一些上行機會。

Great. And then just as a quick follow-up. So on SIPO, we do hear about some competitors are getting aggressive in going after that business as well. And the customer base is fairly concentrated. So how to think about that? Like can customers multisource across different suppliers? How feasible is that?

很好。那再快速追問一下。關於 SIPO，我們確實聽到一些競爭對手也在積極搶攻這塊業務。而客戶基礎相對集中。所以該怎麼看待這件事？例如客戶能否在不同供應商之間多源採購（multisource）？這在可行性上如何？

Yes. I think -- look, the fact that there's competition is a sign that there's strong support for this secular trend, and I think that's not a bad thing from our point of view. What I can say is customers are also coming to us and saying, I don't have enough secured and I need to secure more and how can I help? And some of that is also mitigating geopolitical risk that they see as well. And look, our strategy in any platform is try to work as broadly as we can across the industry, so that we're not betting on the success of player A over player B, but we're mitigated if different people in different sockets.

是的。我想——你看，出現競爭其實是一個訊號，代表這個長期結構性趨勢獲得強力支持；從我們的角度看，這不見得是壞事。我能說的是，客戶也會來找我們並表示：我目前鎖定的供應量不夠，我需要鎖定更多，我能怎麼做、你們能怎麼協助？其中一部分也與他們看到的地緣政治風險緩解有關。而且你看，我們在任何平台上的策略，都是盡可能廣泛地與產業合作，這樣我們不會押注於 A 玩家勝過 B 玩家；即使不同的人在不同的插槽（sockets）中勝出，我們也能降低風險。

Actually, we have more than 40 customers today in Sipo, and that, of course, includes some early-stage companies, but some of the early companies are really doing great things in terms of ramping a new solutions. So I don't think we have a customer concentration concern at this stage and you're going to see even new entrants look at all of the, let's say, I would say, smaller fabless companies that historically were saying copper will last for a long time.

其實，我們目前在 Sipo 已經有超過 40 家客戶；當然其中包含一些早期公司，但有些早期公司在新解決方案的量產爬坡方面做得非常出色。所以我認為現階段我們沒有客戶集中度的疑慮，而且你也會看到新的進入者——看看所有那些我會說較小的無晶圓（fabless）公司，過去一直認為銅互連還能維持很長一段時間。

All of them to a tea have done a photonic strategy made a photonics acquisition, and they're entering into that space. quite aggressively. So I don't think we have a customer concentration concern. And I just think we have durable demand. And like I mentioned earlier, our conversations with the hyperscalers are very helpful to vet what they see kind of as an end consumer of those applications. And again, that reinforces the durability of the demand.

他們全都不約而同地制定了光子（photonics）策略、進行了光子相關併購，並且相當積極地進入這個領域。所以我不認為我們有客戶集中度的疑慮。我也認為我們的需求具有持久性。而且如我先前提到的，我們與超大規模業者（hyperscalers）的對話非常有幫助，能用他們作為這些應用的終端使用者的角度來驗證他們所看到的情況。而這也再次強化了需求的持久性。

CJ Muse, Cantor Fitzgerald.

CJ Muse，Cantor Fitzgerald。

I guess first question on CID implicit in your guide roughly an exit rate of $350 million for this segment. And so curious, based on what you see today in terms of design wins, how do you see kind of the growth into calendar '27? I know you don't want to give specific guidance, but should we be thinking about very strong growth off of that kind of new level? Or is there kind of a digested period? .

我想第一個問題是關於你們指引中隱含的 CID，大致是這個部門退出時的營收水準約 3.5 億美元。因此我很好奇，基於你們目前在設計導入（design wins）方面所看到的情況，你們如何看待成長延伸到 2027 年曆年？我知道你們不想提供具體指引，但我們是否應該把它視為在這個新水準之上會有非常強勁的成長？還是會有一段消化期？.

Yes. Happy to take that, CJ. And you're kind of getting to the right rough numbers going out of this year. the commentary that we provided as part of our Investor Day and Analyst Day a couple of months ago still stands as we see the opportunities to come in for data center. Obviously, it's expected to come in stronger during 2026, but there was a good reason we indicated that [30-plus percentage] year-over-year growth that we're targeting through our long-term model.

是的。CJ，我很樂意回答。你提出的今年結束時的粗略數字大致是對的。我們在幾個月前的投資人日與分析師日所提供的評論仍然成立，因為我們看到資料中心的機會正在到來。顯然，預期在 2026 年會更強，但我們之所以指出在長期模型中所鎖定的年增 [30% 以上]，是有充分理由的。

The plus is sort of ties into what Tim said earlier around some of the growth that we're seeing in demand, the increases that we're making to support that demand into our capacity and just the continued ramp in customer expansion as well. So look, I would still stand by our long-term model that we shared a couple of months back. in terms of that 30-plus point percentage growth going into 2027 and beyond.

這個「加分項」也呼應了 Tim 先前提到的：我們看到需求成長、我們為支援該需求而在產能上做的提升，以及客戶擴張的持續爬坡。所以你看，我仍會支持我們幾個月前分享的長期模型，也就是在 2027 年及之後維持 30% 以上的成長。

CJ, if I could it's very helpful. Demand is clearly very strong. And so the question is more about the rate and pace of manufacturing productivity and capacity expansion. That's obviously a conversation we continue to have with customers. One of the advantages for us is that we can meaningfully inflect our capacity within our existing fab footprint. We don't need to build other fabs and some of our other players in the industry are building other fabs, which obviously is a longer lead time to bring that capacity online.

CJ，如果可以補充一下，這點非常有幫助。需求顯然非常強勁。因此問題更多是在於製造生產力與產能擴張的速度與節奏。這當然也是我們持續與客戶討論的議題。對我們而言的一個優勢是：我們可以在既有晶圓廠（fab）版圖內，實質性地提升產能。我們不需要再興建其他晶圓廠；而產業中的一些其他業者正在興建新廠，這顯然需要更長的前置時間才能把產能上線。

So we're bringing within our four walls. And just to give you a sense, if we think of our long-term plan, just take for Photonics, we could 10 times our photonics capacity within our current four walls plan on a global basis. And so we have a lot of flexibility about when we do that based on the demand and the partnerships we have with customers.

所以我們是在既有廠區內完成擴產。給你一個概念：若以我們的長期規劃來看，單就光子（Photonics）而言，在目前既有廠區的全球規劃下，我們可以把光子產能提升到 10 倍。因此，我們在何時進行擴產上有很大的彈性，取決於需求以及我們與客戶的合作夥伴關係。

Very helpful. And then a follow-up on gross margins. It sounds like the story here in calendar '26 is really all about mix. And so curious given your commentary around selective price increases and how you're going to continue to look at that into calendar '27, how are you thinking about the prioritization of drivers between kind of mix, pricing, and utilization? And is there a framework for us to think about incremental gross margins from here?

非常有幫助。接著追問毛利率。聽起來 2026 年曆年的故事主要都在於產品組合（mix）。因此我想了解，考量你們提到的選擇性漲價，以及你們將如何在 2027 年曆年持續評估這件事，你們如何看待 mix、定價與產能利用率之間的驅動因素優先順序？我們是否有一個框架可以用來思考從現在起的增量毛利率？

Yes. Look, maybe I'll start with the mix point, C.J., because it remains the single biggest driver. And in actual fact, we're still at some of the early innings of those mix shifts that we talked about from a capacity point of view. Clearly, the demand has been pulling through well during the course of 2026. But when you think about the CapEx that we indicated for this year, that's sort of 15% to 20% range, my expectation is that we'll be up to the higher end of that range.

是的。CJ，也許我先從 mix 談起，因為它仍然是最大的單一驅動因素。事實上，從產能角度來看，我們先前談到的那些 mix 轉移仍處於相對早期的階段。顯然，在 2026 年的過程中，需求拉動表現良好。但當你想到我們今年所指出的資本支出（CapEx）約在 15% 到 20% 的區間時，我預期我們會落在該區間的高端。

One of the single biggest drivers to support incremental investments into capacity to meet that growing demand. Now as you'd expect, the time lag between when you incur those CapEx dollars to when you install a tool, qualify it, ramp it there's a lag to be able to then support that demand. My point being is that mix will continue to be a significant driver as we outlook within the model over the course of the next couple of years. And really, the two to three other factors I'd point to.

其中一個最大的驅動因素，就是為了滿足持續成長的需求而對產能進行增量投資。如你所預期，從支出 CapEx 到設備安裝、認證（qualify）、再到爬坡（ramp）之間存在時間落差，之後才能支援該需求。我的重點是：在未來幾年、我們對模型的展望中，mix 仍將持續是重要的驅動因素。另外我還會指出兩到三個因素。

One, yes, utilization, but somewhat impacted. If you think about where things are at from a cycle point of view we still got a point of utilization to be able to grow into just with our installed capacity today. So that's one dynamic. And then the other is really around continued improvements from a cost and a productivity point of view.

第一，是的，產能利用率（utilization），但會受到一些影響。如果從景氣循環的角度來看，以我們目前已安裝的產能而言，利用率仍有提升空間可供成長。這是一個動態。第二則是持續在成本與生產力方面的改善。

We have been focusing maniacally on our structural cost improvements. We've seen that come through in terms of our relative cash cost per Marcelo. It's an area of continued focus for the teams as well. So really across that range of metrics is where we see the opportunity to continue expanding margin and get towards that 40% exit run rate that we indicated in the 2028 time frame.

我們一直非常專注於結構性成本改善。我們已看到這反映在我們相對的現金成本（cash cost）上（每 Marcelo）。這也仍是團隊持續聚焦的領域。因此，正是在這一系列指標上，我們看到持續擴大毛利率的機會，並朝向我們所指出的 2028 年時點 40% 退出（exit）運行率目標邁進。

And beyond that, to Tim's earlier comments, is where we expect to see continued ramps in custom silicon and activities under our IP software custom silicon business as well as the increased ramp from co-packaged optics and the broader Silicon Photonics offerings as well. So that's how we think about it over the next couple of years around some of those margin ramp drivers.

此外，呼應 Tim 先前的評論，我們預期客製化矽（custom silicon）將持續爬坡，以及我們的 IP 軟體客製化矽業務下的相關活動也會增加；同時，共封裝光學（co-packaged optics）與更廣泛的矽光子（Silicon Photonics）產品組合也會加速放量。這就是我們對未來幾年、部分毛利率爬坡驅動因素的看法。

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.

今天節目的問答環節到此結束。我想把節目交回給 Eric Chow，請他做進一步的結語。

Thank you, Jonathan. Thank you, everyone, for joining today. We're very glad to see you, and we will see you at the Goldman Sachs Conference on September 8. Thank you.

謝謝你，Jonathan。也謝謝各位今天的參與。很高興見到大家，我們將在 9 月 8 日的高盛（Goldman Sachs）會議上再見。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、先生參與今天的電話會議。本節目到此結束。您現在可以掛線。祝您有美好的一天。