科磊 (KLAC) 2026 Q1 法說會逐字稿

Good afternoon. My name is Stephanie, and I'll be your conference operator today. At this time, I'd like to welcome everyone to the KLA Corporation September quarter 2025 post-Earnings conference call.

午安.我叫史蒂芬妮，今天我將擔任你們的會議接線生。在此，我謹代表 KLA 公司歡迎各位參加 2025 年 9 月季度財報電話會議。

(Operator Instructions)

（操作說明）

Thank you. I will now turn the call over to Kevin Kessler, Vice President of Investor Relations and Market Analytics for KLA. Please go ahead.

謝謝。現在我將把電話交給KLA投資者關係和市場分析副總裁凱文凱斯勒。請繼續。

Welcome to the September 2025 quarterly earnings call. I'm joined by our CEO, Rick Wallace; and our CFO, Bren Higgins. We will discuss today's results as well as our December quarter outlook, which was released after the market close and is available on our website along with the supplemental materials. We are presenting today's discussion and metrics on a non-GAAP financial basis unless otherwise specified. All full year references made refer to calendar years. The earnings materials contain a detailed reconciliation of GAAP to non-GAAP results.

歡迎參加2025年9月季度財報電話會議。與我一同出席的還有我們的執行長里克華萊士和財務長布倫希金斯。我們將討論今天的業績以及我們在收盤後發布的 12 月季度展望，該展望和補充資料可在我們的網站上查閱。除非另有說明，我們今天的討論和指標均以非GAAP財務準則為基礎。所有提及的全年均指日曆年。盈利資料中包含 GAAP 與非 GAAP 結果的詳細調整表。

KLA's IR website also contains future events, presentations, corporate governance information and links to our SEC filings. Our comments today are subject to risks and uncertainties reflected in the disclosure of risk factors in our SEC filings. Any forward-looking statements, including those we make on the call today, are also subject to those risks, and KLA cannot guarantee those forward-looking statements will come true. Our actual results may differ significantly from those projected in our forward-looking statements. We will begin the call with Rick providing commentary on the business environment in our quarter, followed by Bren with financial highlights and our outlook.

KLA 的投資者關係網站還包含未來活動、簡報、公司治理資訊以及我們向美國證券交易委員會提交的文件的連結。我們今天的評論受到風險和不確定性的影響，這些風險和不確定性已反映在我們向美國證券交易委員會提交的文件中揭露的風險因素中。任何前瞻性陳述，包括我們今天在電話會議上所做的陳述，也都存在這些風險，KLA 無法保證這些前瞻性陳述會實現。我們的實際結果可能與我們前瞻性聲明中預測的結果有重大差異。我們將首先由 Rick 對我們本季的商業環境進行評論，然後由 Bren 介紹財務亮點和我們的展望。

Before I turn the call over to Rick, I wanted to provide a save the date for our Investor Day. It has been rescheduled for Thursday, March 12, 2026 in New York. Now over to Rick.

在將電話交給里克之前，我想提醒大家預留我們投資者日的日期。活動改期至 2026 年 3 月 12 日星期四在紐約舉行。現在輪到瑞克了。

Thank you, Kevin. To kick off our call today, I'll cover a few highlights from our quarter that showcase how the company is benefiting from the growing relevance of process control and AI infrastructure investment and our momentum in advanced packaging. KLA delivered strong results across the board in the September quarter with revenue of $3.21 billion and non-GAAP diluted EPS of $8.81, GAAP diluted EPS was $8.47. This performance demonstrates how KLA process control leadership has expanded beyond leading-edge R&D investment to address all growth markets in WFE, including high bandwidth memory and advanced packaging.

謝謝你，凱文。在今天的電話會議開始時，我將介紹我們季度的一些亮點，這些亮點展示了公司如何受益於製程控制和人工智慧基礎設施投資日益增長的重要性以及我們在先進封裝領域的良好勢頭。科萊恩在9月的季度業績全面強勁，營收達32.1億美元，非GAAP攤薄後每股收益為8.81美元，GAAP攤薄後每股收益為8.47美元。這項業績表明，科萊恩在製程控制領域的領先地位已從前沿研發投入擴展到WFE（晶圓前端）的所有成長市場，包括高頻寬記憶體和先進封裝。

Accelerating investment in scaling AI infrastructure is fueling technology development investment across the leading edge driving more designs, increased complexity, shorter product cycles and higher-value wafers. Alongside this growth, the industry is also seeing rising demand for advanced packaging. In this complex environment of rapid AI technology development, process control accelerates time to results by resolving process integration challenges during the fab and ramp-up phase to optimize time to market for a diverse mix of semiconductor designs.

加速投資於擴大人工智慧基礎設施規模，正在推動尖端技術開發投資，從而帶來更多設計、更高複雜性、更短的產品週期和更高價值的晶圓。隨著產業成長，對先進包裝的需求也不斷增加。在人工智慧技術快速發展的複雜環境中，製程控制透過解決晶圓製造和產能提升階段的製程整合挑戰來加快實現成果的時間，從而優化各種半導體設計的上市時間。

KLA's leading-edge customers are also challenged to optimize yield and limit process variability in high-volume production environment, resulting in increasing process control intensity. In this increasingly complex semiconductor device technology landscape, we're seeing rapid growth in demand for KLA's advanced packaging portfolio, which has emerged as a meaningful market for the company as heterogeneous device integration has become more complex. KLA's advanced packaging systems revenue continues to gain momentum through a combination of intensity gains and market share improvements across our portfolio.

KLA 的領先客戶也面臨著在大批量生產環境中優化產量和限制製程變異性的挑戰，導致製程控制強度不斷增加。在日益複雜的半導體裝置技術領域，我們看到市場對 KLA 先進封裝產品組合的需求正在迅速增長，隨著異構裝置整合變得越來越複雜，這已成為該公司的重要市場。KLA先進包裝系統的收入持續成長，這得益於我們產品組合中強度提升和市場份額改善的綜合作用。

For calendar year 2025, we expect advanced packaging related revenue to exceed $925 million, up approximately 70% year-over-year. KLA Service business also continues to build our strong growth. Services grew to $745 million in the September quarter, up 6% sequentially and 16% year-over-year. Consistency and resiliency are hallmarks of the KLA service business.

我們預計 2025 年先進封裝相關收入將超過 9.25 億美元，年增約 70%。KLA 服務業務也持續保持強勁成長動能。9 月季度服務收入成長至 7.45 億美元，季增 6%，年增 16%。穩定性和韌性是KLA服務業務的標誌。

Finally, the September quarter was strong on both cash flow and capital returns front. Strong cash flow in the quarter was at a record of $1.066 billion. Over the past 12 months, free cash flow was $3.9 billion, with free cash flow margin of 31%. Total capital return in the September quarter was $799 million comprised of $545 million in share repurchases and $254 million in dividends. Total capital return over the past 12 months was $3.09 billion.

最後，9 月當季現金流和資本回報方面均表現強勁。本季強勁的現金流達到創紀錄的 10.66 億美元。過去 12 個月，自由現金流為 39 億美元，自由現金流利潤率為 31%。9 月季度總資本回報為 7.99 億美元，其中包括 5.45 億美元的股票回購和 2.54 億美元的股息。過去 12 個月的總資本回報為 30.9 億美元。

In summary, KLA business is both enabled and benefits from today's technology inflections and the growth drivers related to AI as well as from growth in advanced packaging. KLA's business has gone from being primarily indexed to leading-edge R&D investments in foundry logic customers to now addressing all growth markets in WFE, including memory, advanced packaging and leading edge and legacy node logic. As we look ahead over the next several years, the long-term secular trends driving semiconductor industry demand, investments in WFE and advanced packaging are compelling and represent a relevant performance opportunity for KLA.

總而言之，KLA 的業務既受益於當今的技術變革和與人工智慧相關的成長驅動因素，也受益於先進包裝的成長。KLA 的業務已經從主要依靠對代工邏輯客戶的前沿研發投資，發展到現在面向 WFE 的所有成長市場，包括記憶體、先進封裝以及前沿和傳統節點邏輯。展望未來幾年，推動半導體產業需求、WFE 和先進封裝投資的長期趨勢十分顯著，這對 KLA 而言是一個重要的業績成長機會。

In this dynamic growth environment, our consistent execution reflects the resilience of the KLA operating model, the strength of our global team and our disciplined approach to capital allocation focused on long-term investment and maximizing total shareholder value.

在這種充滿活力的成長環境中，我們始終如一的執行力體現了 KLA 營運模式的韌性、我們全球團隊的實力以及我們以長期投資和最大化股東總價值​​為重點的嚴謹資本配置方法。

With that, I'll turn the call over to Bren to discuss the quarter's financial highlights.

接下來，我將把電話交給布倫，讓他來討論本季的財務亮點。

Thanks, Rick. KLA September quarter results reflect double-digit year-over-year growth and improved profitability. Revenue was $3.21 billion, above the guidance midpoint of $3.15 billion. Non-GAAP diluted EPS was $8.81 and GAAP diluted EPS was $8.47, each above the midpoint of the respective guidance ranges. Gross margin was 62.5%, 50 basis points above the midpoint of guidance, driven by a stronger product mix and manufacturing efficiencies. Non-GAAP operating expenses were $618 million. Operating expenses included $360 million in R&D and $258 million in SG&A. Non-GAAP operating margin was 43.2%. The Other income and expense net was a $28 million expense with upside to guidance principally driven by a favorable mark-to-market adjustment on a strategic supplier investment. The quarterly effective tax rate was 14.1%.

謝謝，里克。KLA 9 月季度業績反映出兩位數的年成長和獲利能力的提高。營收為32.1億美元，高於先前預期的31.5億美元中位數。 （非GAAP準則）稀釋後每股收益為 8.81 美元，GAAP 稀釋後每股收益為 8.47 美元，均高於各自預期範圍的中點。毛利率為 62.5%，比預期中位數高出 50 個基點，主要得益於更強勁的產品組合和更高的生產效率。非GAAP營運費用為6.18億美元。營運費用包括研發費用 3.6 億美元及銷售、管理及行政費用 2.58 億美元。非GAAP營業利益率為43.2%。其他收入和支出淨額為 2,800 萬美元，高於預期，主要得益於對策略供應商投資的有利市值調整。季度實際稅率為14.1%。

Net income was $1.17 billion. GAAP net income was $1.12 billion. Cash flow from operations was $1.16 billion, and free cash flow was $1.07 billion. The breakdown of revenue by reportable and end markets and major products and regions can be found within the shareholder letter and slides. Moving on to the balance sheet.

淨利為11.7億美元。 （GAAP準則）淨收入為11.2億美元。經營活動產生的現金流量為 11.6 億美元，自由現金流量為 10.7 億美元。股東信和幻燈片中提供了按可報告市場、終端市場、主要產品和地區劃分的收入明細。接下來來看資產負債表。

We ended the quarter with $4.7 billion in total cash, cash equivalents and marketable securities and $5.9 billion in debt. The company has a flexible and attractive bond maturity profile supported by investment-grade ratings from all three major rating agencies. A cornerstone of KLA's business is consistent strong free cash flow generation. driven by one of the best operating models in the industry and a predictable, highly differentiated service business. This helps drive a comprehensive capital return strategy that includes consistent dividend growth and increase in share repurchases over the long term.

本季末，我們持有現金、現金等價物和有價證券總額為 47 億美元，負債總額為 59 億美元。該公司擁有靈活且具吸引力的債券到期結構，並獲得了三大評級機構的投資等級評級。KLA業務的基石在於持續強勁的自由現金流，這得益於業界最佳的營運模式之一以及可預測且高度差異化的服務業務。這有助於推動全面的資本回報策略，包括長期持續的股息成長和股票回購的增加。

Our actions this year emphasize our commitment to capital returns and our confidence in KLA's long-term shareholder value accretion. On April 30, 2025, we announced the 16th consecutive annual dividend increase, up 12% to $1.90 per share per quarter or an annualized dividend of $7.60 per share. Along with this action, we also announced a $5 billion share repurchase authorization.

我們今年的行動凸顯了我們對資本回報的承諾，以及我們對KLA長期股東價值成長的信心。2025 年 4 月 30 日，我們宣布連續第 16 年提高年度股息，每季每股股息成長 12%，達到 1.90 美元，或年化股息每股 7.60 美元。同時，我們也宣布了一項50億美元的股票回購授權。

Turning to the outlook. It continues to be driven by increasing investment in leading edge logic, HBM and advanced packaging. Growth of advanced packaging supporting heterogeneous chip integration has led to a new meaningful served market for KLA. What was once a rounding error in wafer fab equipment is now according to KLA internal estimates an approximately $11 billion market, growing faster than core WFE. This is particularly true as chip density shrink and the processing required for package increase risk for our customers.

展望未來。持續成長的動力來自於對尖端邏輯、HBM 和先進封裝技術的不斷增加投資。支援異質晶片整合的先進封裝技術的發展，為KLA帶來了一個新的、有意義的服務市場。曾經在晶圓製造設備領域只是一個四捨五入的誤差，如今根據 KLA 內部估計，它已發展成為一個價值約 110 億美元的市場，其增長速度超過了核心 WFE。隨著晶片密度降低和封裝所需加工製程的增加，這對我們的客戶來說尤其如此，風險也隨之增加。

For KLA, this creates a new served available market that will augment the company's revenue growth over the next several years. The market and technology road map for leading edge WFE supporting high-performance compute is driving relative inflections for process control. This opportunity, coupled with the evolving complexity of advanced packaging, supports an even broader market opportunity for KLA.

對於 KLA 而言，這將創造一個新的服務市場，並在未來幾年內促進公司的收入成長。支援高效能運算的前沿 WFE 的市場和技術路線圖正在推動製程控制的相對轉變。這一機遇，加上先進封裝技術的日益複雜化，為KLA提供了更廣泛的市場機會。

As we approach the close of calendar 2025, we continue to expect mid- to high single-digit growth in WFE, modestly improved from our previous outlook discussed last quarter. Growth in 2025 is being driven principally by increasing investment in both leading-edge foundry logic and memory to support growing AI and premium mobile demand, partially offset by lower demand from domestic China. Given KLA's business momentum, expanding market share opportunities and higher process control intensity is the leading edge across all segments. We remain on track to outperform the WFE market in 2025.

隨著 2025 年即將結束，我們繼續預計 WFE 將實現中高個位數成長，比上個季度討論的先前展望略有改善。2025 年的成長主要得益於對尖端代工邏輯和記憶體投資的增加，以支援不斷增長的人工智慧和高端行動需求，但部分成長被中國國內需求的下降所抵消。鑑於 KLA 的業務發展勢頭，擴大市場份額的機會和提高流程控制強度是其在所有細分市場的領先優勢。我們仍有望在 2025 年跑贏 WFE 市場。

The advanced packaging market is also expected to grow more than 20% compared to last year. Finally, customer discussions have become more constructive on expectations for calendar year 2026 to be a growth year for the industry with a broader spending profile in 2025 for both WFE and advanced packaging. While it is still too early to provide precise calendar 2026 revenue guidance, our view today is that first half revenue levels will be roughly flat to modestly up compared to the second half of calendar 2025, with accelerating growth in the second half of the calendar year. This outlook is inclusive of the revenue impact related to additional market access loss related to certain customers in China resulting from extended export controls from the US government.

預計先進封裝市場也將比去年成長超過20%。最後，客戶討論變得更加積極，他們預期 2026 年將是該行業的成長之年，並預計 2025 年 WFE 和先進封裝的支出規模將擴大。雖然現在提供 2026 年日曆年的具體收入預測還為時過早，但我們目前的看法是，與 2025 年下半年相比，上半年的收入水平將大致持平或略有增長，而下半年將加速增長。這項展望包含了由於美國政府延長出口管製而導致中國某些客戶市場准入受限所帶來的收入影響。

We estimate the revenue impact on the December quarter and calendar 2026 to be approximately $300 million to $350 million for KLA. For 2026, this impact is spread roughly evenly across the first and second half of the calendar year. KLA's unique product portfolio differentiation and value proposition are focused on enabling technology transitions, accelerating process node capacity ramps and ensuring yield entitlement and high-volume production. The market environment and the complexity of our customers' technology road maps are compelling and bring challenges and opportunities for KLA to continue its relative performance.

我們估計，對 KLA 2026 年 12 月季度和日曆年的營收影響約為 3 億至 3.5 億美元。2026 年，這種影響大致平均分佈在日曆年的上半年和下半年。KLA獨特的產品組合差異化和價值主張專注於實現技術轉型、加速製程節點產能提升，並確保良率和大量生產。市場環境和客戶技術路線圖的複雜性為KLA帶來了挑戰和機遇，使其能夠繼續保持相對優異的表現。

In this industry environment, KLA remains focused on supporting customers, investing for the future, executing product road maps and driving productivity across the enterprise. KLA's December quarter guidance is as follows: Total revenue is expected to be $3.225 billion, plus or minus $150 million. Foundry logic revenue from semiconductor customers is forecasted to be approximately 59% and memory is expected to be approximately 41%, semi process control systems revenue to semiconductor customers. Within DRAM -- within memory, DRAM is expected to be about 78% and NAND the remaining 22%. As always, these business mix approximations pertains solely to our semiconductor customers and do not fully reflect our total semiconductor process control systems revenue.

在當前的產業環境下，KLA 仍然專注於支援客戶、投資未來、執行產品路線圖並提高整個企業的生產力。KLA 對 12 月季度的業績預期如下：總收入預計為 32.25 億美元，上下浮動 1.5 億美元。預計半導體客戶的代工邏輯收入約為 59%，記憶體收入約 41%，半導體製程控制系統收入也來自半導體客戶。在 DRAM 記憶體中，DRAM 預計佔 78%，NAND 佔 22%。與以往一樣，這些業務組合估算僅涉及我們的半導體客戶，並不完全反映我們半導體製程控制系統的總收入。

Gross margin is forecasted to be 62%, plus or minus 1-percentage-point based on relatively consistent factory output versus the September quarter and product mix revenue expectations. Operating expenses are forecasted to be approximately $635 million in the December quarter as we continue to make product development and infrastructure investments to support expected revenue growth. Given our expectations for company growth and product development road map requirements, we will maintain our operating expense trajectory.

根據與 9 月季度相比相對穩定的工廠產量和產品組合收入預期，預計毛利率為 62%，上下浮動 1 個百分點。由於我們將繼續進行產品開發和基礎設施投資以支持預期的收入成長，預計 12 月季度的營運支出約為 6.35 億美元。鑑於我們對公司成長和產品開發路線圖要求的預期，我們將維持目前的營運費用水準。

Our business model is designed to deliver 40% to 50% incremental non-GAAP operating margin leverage on revenue growth over the long run. Other model assumptions include other income and expense net of approximately $32 million expense for the December quarter, the effective tax rate assumption has risen slightly to 14%, reflecting the impact of recent global tax changes.

我們的商業模式旨在從長遠來看，透過營收成長實現 40% 至 50% 的非 GAAP 營業利潤率提升。其他模型假設包括其他收入和支出，扣除 12 月季度約 3,200 萬美元的支出後，有效稅率假設略微上升至 14%，反映了近期全球稅收變化的影響。

For the December quarter, GAAP diluted EPS is expected to be $8.46, plus or minus $0.78 and non-GAAP diluted EPS of $8.70 plus or minus $0.78. EPS guidance is based on a fully diluted share count of approximately 132 million shares.

預計12月季度GAAP攤薄後每股收益為8.46美元，上下浮動0.78美元；非GAAP攤薄後每股收益為8.70美元，上下浮動0.78美元。每股盈餘預期基於約1.32億股的完全攤薄後股份數量。

In conclusion, our near-term revenue guidance shows modest growth and is consistent with our views from the start of the year of relative top line stability. We expect to meaningfully outperform the mid- to high single-digit WFE growth rate in 2025, driven by rising process control intensity, inclusive of the significant growth of the advanced packaging market. KLA focuses on delivering a differentiated product portfolio that addresses customers' technology road map requirements which are driving our longer-term relevance and growth expectations.

總之，我們近期的營收預期顯示將實現溫和成長，這與我們年初以來對營收相對穩定的看法一致。我們預計，在製程控制強度不斷提高（包括先進封裝市場的顯著成長）的推動下，到 2025 年，WFE 的成長率將顯著超過中高個位數。KLA 致力於提供差異化的產品組合，以滿足客戶的技術路線圖需求，從而推動我們實現長期發展和成長預期。

KLA's business is well positioned for today's technology inflections and growth drivers. We are encouraged by the customer engagement that informs our business forecast. Long-term secular trends driving semiconductor industry demand and investments in WFE and advanced packaging are compelling and represent a relative performance opportunity for KLA over the next several years.

KLA 的業務已做好充分準備，迎接當今的技術變革和成長動力。客戶的積極參與為我們的業務預測提供了依據，這令我們倍感鼓舞。推動半導體產業對晶圓前端和先進封裝的需求和投資的長期趨勢十分強勁，並代表著未來幾年科樂集團相對的業績成長機會。

In addition, the growing investment in custom silicon, particularly among hyperscalers developing their own custom chips have led to a proliferation of unique device designs and increased demand on our customers to deliver performance, volume and time to market. As design complexity and diversity grow, so does the need for advanced process control. As a result, KLA has seen growth in process control intensity as each new chip design requires rigorous inspection, metrology and yield optimization solutions. KLA is uniquely positioned to benefit from these trends as we expand our market leadership and deliver differentiated value to our customers. That concludes our prepared remarks. Let's begin Q&A.

此外，對客製化矽的投資不斷增長，尤其是超大規模企業開發自己的客製化晶片，導致獨特的設備設計層出不窮，也對我們的客戶提出了更高的要求，要求他們提供性能、產量和上市時間。隨著設計複雜性和多樣性的增加，對先進製​​程控制的需求也隨之增加。因此，隨著每款新晶片設計都需要嚴格的檢驗、計量和良率最佳化解決方案，KLA 的製程控制強度也隨之提升。KLA憑藉其獨特的優勢，能夠從這些趨勢中獲益，從而擴大我們的市場領導地位，並為客戶提供差異化的價值。我們的發言稿到此結束。讓我們開始問答環節。

I'm sorry, I was just going to pass it over to you to start the process. Thank you.

抱歉，我正打算把它交給你開始處理。謝謝。

(Operator Instructions)

（操作說明）

Harlan Sur with JPMorgan.

哈蘭·蘇爾任職於摩根大通。

Congratulations on the strong quarterly execution. Last quarter, you had this early view given your lead times, customer discussions on calendar '26 being a growth year. You reiterated that today, but talked about being more constructive on that growth rate. So it was another day -- with another 90 days of visibility given that leading-edge design starts are continuing to expand at a rapid pace, the recent and significant AI data center infrastructure announcements, has the magnitude on the WFE growth outlook improved? Or is it more just confidence level on the growth that you were thinking about 90 days ago and you mentioned the broader spending profile on WFE and advanced packing. Can you guys just elaborate on that a little bit?

恭喜貴公司季度業績表現出色。上個季度，鑑於你們的交貨週期，你們對「2026 年是成長年」的討論已經提前看到了這一點。你今天重申了這一點，但談到了要對成長率採取更具建設性的態度。又是一天過去了——鑑於前沿設計開工量持續快速增長，以及近期重大的人工智慧資料中心基礎設施公告，未來 90 天的可見性是否有所提高？ WFE 成長前景的規模是否有所改善？或者，這更多的是指你90天前所考慮的增長的信心水平，以及你提到的WFE和先進包裝方面的更廣泛的支出情況。你們能再詳細解釋一下嗎？

Sure, Harlan. I'll start. Thank you for the comments. I don't know if it's really a strengthening outlook as much as it's just -- we're getting closer to it. Customers, particularly our long-standing customers and their lead time expectations, we're starting to get more constructive about exact timing. We're encouraged by what we're seeing certainly for KLA at the leading edge with a broadening level of investment. We think that's going to be positive on leading edge foundry logic.

當然，哈蘭。我先來。感謝您的評論。我不知道這是否真的代表前景更光明，或許只是──我們離目標越來越近了。對於客戶，特別是我們的老客戶及其交貨時間期望，我們開始在確切時間安排方面採取更具建設性的態度。我們看到 KLA 在產業前沿不斷擴大投資規模，這無疑令人鼓舞。我們認為這對領先的晶圓代工邏輯來說將是一個積極因素。

DRAM is also constructive. And with the investments in HPM, that's been very process control intensive. And so that's been a really good sign as well. Flash market is, I think, continues to grow. The rest of the legacy market, I'm not so sure there's much growth there, and I think we'll have a little bit of a correction in China.

DRAM 也具有建設性作用。而對 HPM 的投資，使得流程控制變得非常繁重。所以，這也是一個非常好的跡象。我認為閃購市場仍在持續成長。至於其他傳統市場，我不太確定它們是否還有成長空間，我認為中國市場會出現一些調整。

Obviously, we're feeling the effects of some new control that's impacting our view into next year. But we were expecting China to normalize anyway. So I think as we look at it all, we're pretty constructive on WFE growth, rising in capital intensive -- process control intensity. And packaging has a lot of momentum, both in terms of intensity, but also share. So we feel pretty good about what's in front of us, and we'll have a lot more to say specifically about growth in the industry and our expectations for KLA beyond what we said in the comments. We'll have a lot more to say when we report for December and January.

顯然，我們正在感受到一些新的管控措施的影響，這些措施正在影響我們對明年的展望。但我們原本就預期中國會走向正常化。所以我認為，從整體來看，我們對 WFE 的成長持相當正面的態度，認為其資本密集度高，流程控制強度大。包裝產業發展勢頭強勁，無論是在發展強度還是市場份額方面都是如此。所以我們對眼前的情況感到非常樂觀，除了我們在評論中提到的之外，我們還會就行業成長和我們對 KLA 的期望發表更多具體意見。我們將在發布12月和1月的報告時詳細介紹。

One thing, Harlan, just to add to Ben's comment. What we do see and kind of feel is body language from customers pretty strong in terms of wanting to make sure they're securing slots. So we're having kind of conversations with people not wanting us to get away from them because they're worried that they may not be able to achieve their objectives if they don't bind us up. And I suspect that's across the other equipment guys, too.

哈蘭，我還要補充一點，關於本的評論。我們看到和感受到的是，顧客的肢體語言非常強烈地表明他們想要確保自己能預訂到座位。所以我們現在正和一些不想讓我們擺脫他們的人進行對話，因為他們擔心如果不束縛住我們，他們可能無法實現他們的目標。我懷疑其他設備管理員也存在同樣的問題。

Got it. And I appreciate that. And then specifically on advanced foundry and logic, in addition to the increased process control intensity, as the industry moves from 2-nanometer to less than 2-nanometer right there, there's an added dynamic, I feel like where your foundry customers are standing up fabs in totally new geographies, right? So more uncertainty on yield ramps, systematic defects, like different type of workforce, right? .

知道了。我對此表示感謝。具體來說，在先進晶圓代工和邏輯領域，除了製程控制強度不斷提高之外，隨著產業從 2 奈米發展到小於 2 奈米，還出現了一種新的動態，我覺得你們的晶圓代工客戶正在全新的地區建立晶圓廠，對吧？所以，產量提升、系統性缺陷以及不同類型的勞動力等方面都存在更多不確定性，對吧？。

And then on the advanced logic side, the large guy here is more focused on building out a world-class foundry business, which means way more focus on yield and manufacturability versus their historical trend. I'm Wondering if these additional dynamics are driving the potential for incremental process control spend as you look into next year?

而在先進邏輯方面，這家大公司更專注於打造世界級的代工業務，這意味著與以往的趨勢相比，他們更加重視良率和可製造性。我想知道，在您展望明年時，這些額外的動態因素是否會影響流程控制的支出成長？

Well, sorry, I do think that there -- as the customers are dealing with the new design rules and some of them, especially that are maybe back in it, if you will, trying to be leading edge, they're benchmarking what do they need for process control, and we're seeing kind of very constructive conversations around that. So I think that's true. I think it's kind of filling out the rest of the players, if you will, in terms of how they're thinking about investment and process control. So yes, I'd say that, that strengthens, if there are more players doing more leading edge in more locations, that's going to be accretive to overall intensity.

嗯，抱歉，我的確認為——由於客戶正在應對新的設計規則，其中一些客戶，特別是那些可能重新回到這個領域，試圖保持領先地位的客戶，他們正在對標流程控制所需的標準，我們看到圍繞這一點展開了非常建設性的對話。所以我認為這是真的。我認為，從投資和流程控制的角度來看，這在某種程度上完善了其他參與者的認知。所以，是的，我認為，如果更多的選手在更多的地方進行更多前沿技術的應用，這將增強整體的競爭強度。

Yeah. One of the themes over the last couple of years has obviously been significant investment augmented by China and legacy design rules. But when you think about leading edge, leading edge was tremendously efficient, right, with most of the investment being really driven by one of our customers. I think as you start to see a broadening out there, it creates more opportunities for leading-edge engagement, process control intensity as a lot of the strategic investment happens to support what is an accelerating growth opportunity for our customers. So I think we're encouraged by that profile as we move forward.

是的。近兩年來的一個明顯主題是，中國加強了投資力度，並沿用了傳統的設計規則。但當你想到領先技術時，你會發現領先技術效率極高，對吧？而且大部分的投資其實都是由我們一位客戶推動的。我認為，隨著市場範圍的擴大，它將為前沿參與和流程控制強度創造更多機會，因為許多策略投資都是為了支持我們客戶不斷加速的成長機會。所以我認為，這樣的發展前景讓我們倍受鼓舞。

Vivek Arya, Bank of America.

Vivek Arya，美國銀行。

On the foundry logic side, you are, I think, guiding it to decline to 59% from 74% of sales, I believe, so a decline of over $300 million sequentially. I'm curious what's causing this drop? How much of this is the China restriction? How much of this is the China impact? And is this kind of just a 1 quarter lumpiness? Or is there more to read into it as we look into the first half of next year?

我認為，在晶圓代工邏輯方面，你們正在引導其銷售額從 74% 下降到 59%，這意味著季減超過 3 億美元。我很想知道是什麼原因導致了這次下跌？其中有多少是受中國限制？其中有多少是中國的影響？這難道只是四分之一的凹凸不平嗎？或者，展望明年上半年，我們還有更多值得解讀的地方？

Yeah. So Vivek, for our semiconductor customers on the mix side, on the leading edge, it's upticking in the December quarter, but it's being offset by a reduction in China. China was elevated in September at 39%. And just for a reference point, our expectations for the total year for China, and I have been pretty consistent with this for the last 9 months or so as we thought it would be somewhere in 30% plus or minus range. So it was a little bit elevated versus the annual trend.

是的。所以 Vivek，對於我們半導體客戶而言，在混合配置方面，也就是在尖端技術方面，12 月的季度有所增長，但被中國市場的下滑所抵消。中國9月升至39%。僅供參考，我們對中國全年的預期，在過去的9個月左右的時間裡一直相當一致，我們認為會在正負30%的範圍內。所以它比年度平均略高一些。

So as China comes down, part of that is you have leading edge going up. And then you also have memory as -- and particularly in DRAM, we see that upticking in the December quarter. So there's some moving parts there that's what's happening. As it relates to the recent export controls, I would say the impact on the December quarter is fairly immaterial to the company overall and that we were able to move slots around. We've got certain products where customers get in the queue, and so we can pull business forward. It's a lot of different customers. But obviously, over the long term, that's lost business. And so as we said in the prepared remarks, we think that's about $300 million to $350 million between now and the end of '26. So hopefully, that gives you a little color on the moving parts.

所以，隨著中國經濟下滑，部分原因是領先優勢正在上升。此外，記憶體方面也有所成長——尤其是在DRAM方面，我們看到12月份季度記憶體需求上升。所以這裡面有很多環節在相互作用，這就是事情的來龍去脈。關於最近的出口管制，我認為對公司整體而言，12 月季度的影響相當微不足道，而且我們能夠調整發貨安排。我們有一些產品需要顧客排隊購買，這樣我們就可以提前完成交易。顧客群非常多元化。但很顯然，從長遠來看，這會造成業務損失。正如我們在準備好的演講稿中所說，我們認為從現在到 2026 年底，這大約是 3 億至 3.5 億美元。希望這能讓你對各環節有所了解。

C.J Muse, Cantor Fitzgerald.

C.J.繆斯，坎托·費茲傑拉。

I guess I was hoping to focus on gross margins. You guided down 50 bps. I'm assuming that's just product mix. I'd love to hear your thoughts there. And then, Bren, you're highlighting, again, the 40% to 50% incremental operating margins. If we are in a world where WFE continues to grow kind of in a double-digit world over the next couple of years, should we be thinking that you're at the higher end of that range, even kind of greater contribution from the higher-margin silicon.

我原本希望專注在毛利率。你引導下調了 50 個基點。我猜這只是產品組合的問題。我很想聽聽你的看法。然後，布倫，你再次強調了 40% 到 50% 的增量營業利潤率。如果未來幾年 WFE 繼續以兩位數的速度成長，我們是否應該認為 WFE 已經處於這個範圍的較高水平，甚至高利潤矽晶片的貢獻會更大？

Yeah, C.J., on the guide down, you're right, it's about 50 bps, and it's mostly related to just mix adjustments in the quarter. As I've said over the last couple of quarters, there is a tariff impact that we're dealing with, which is more or less consistent quarter-to-quarter. That's roughly 50 to 100 basis points of the impact. So yes, we're guiding 62 and output is relatively consistent. So it's really a mix issue.

是的，C.J.，你說得對，在指南中，大約是 50 bps，這主要與四分之一處的混音調整有關。正如我在過去幾個季度所說，我們正在應對關稅的影響，而且這種影響每季都或多或少地保持穩定。這大約相當於 50 到 100 個基點的影響。是的，我們指導 62 個單位，產量也相對穩定。所以這實際上是一個混合問題。

In terms of how we were thinking about running the company and over the long run, certainly our 40% to 50% long-standing incremental operating margin target does drive how we size the company. Gross margin obviously is a factor in that. And so we have to think about where gross margins are trending as we consider that. We outperformed that target pretty significantly as revenue has grown in the mid-teens. I'd call it above trend line growth in 2025. So as we move forward, I think the easiest way to think about that is if you're more or less a trend line, we're more or less in the middle of the target range.

就我們對公司營運和長期發展的思考而言，我們長期以來設定的 40% 至 50% 的增量營業利潤率目標，無疑會影響我們對公司規模的規劃。毛利率顯然是其中一個因素。因此，當我們考慮這個問題時，我們必須思考毛利率的趨勢。我們的業績遠超預期目標，營收成長了兩位數以上。我認為到 2025 年，這將高於趨勢線成長。所以，展望未來，我認為最簡單的理解方式是，如果你大致上是一條趨勢線，那麼我們大致處於目標範圍的中間位置。

And if growth levels are above that high single-digit trend line, we should outperform it growth levels below, we'll probably underperform the target a bit. But that's how we're going to size the company over time, and our performance over longer periods of time is obviously very consistent with that.

如果成長水準高於這個高個位數的趨勢線，我們應該會超過它；如果成長水準低於這個趨勢線，我們可能會略低於目標。但我們將以這種方式衡量公司未來的發展規模，而我們長期的業績也顯然與此非常一致。

Joe Quatrochi, Wells Fargo.

喬·夸特羅奇，富國銀行。

I appreciate the qualification or quantification on the advanced packaging WFE. I was curious just to think about just the advanced packaging process control intensity. I think just based on some of the things you put out there or talked about in the past, it's like high teens. Is that the right way to think about it? And then how do you think about where that goes over time?

我非常感謝您提供先進封裝WFE的定性或量化資訊。我很好奇，想了解先進包裝製程的控制強度。我覺得，根據你過去發表的一些言論或談論的內容來看，你現在的年齡大概相當於十幾歲的青少年。這是正確的思考方式嗎？那麼，你如何看待它隨著時間的推移會如何發展呢？

No, I wouldn't say it's that high. I think if you look at KLA's share of WFE, I mean, one of the interesting things and as I said in the prepared remarks, is it wasn't much of a factor for us in our business. you don't have to go back more than just a few years, and the percent per KLA was in the 1% range. And now if you take our views on 2025 at $11 billion or so, we're approaching 6%. So we've seen an escalation here in terms of intensity as the requirements have changed fundamentally related to the high-performance computing on the logic side and the memory side.

不，我不會說它有那麼高。我認為，如果你看一下科樂美在WFE中的份額，你會發現一個有趣的現象，正如我在準備好的發言稿中所說，它對我們的業務影響不大。你只要回顧過去幾年的數據，科樂美每家公司的市佔率就只有1%左右。如果按照我們對 2025 年 110 億美元左右的預期，我們離 6% 的目標還差得很遠。因此，我們看到，隨著邏輯端和記憶體端的高效能運算需求發生根本性變化，強度也隨之增加。

So we think that, that continues over time. I don't think you're going to see that kind of that slope of growth. But we do expect that as density shrink and processes become more complex than it does play to the need for more more advanced systems. And of course, we have our front-end portfolio that we can use to address this interesting market. So I think it's a new SAM for KLA. We have a lot of great drivers within that we think are driving process control and KLA share of the market, and we're augmenting that growth with this growth that we expect to see in advanced packaging that likely over time grows modestly faster than WFE.

所以我們認為，這種情況會一直持續下去。我不認為你會看到那種成長速度。但我們預計，隨著密度降低和製程變得更加複雜，這將促使人們需要更先進的系統。當然，我們還有前端產品組合可以用來開拓這個有趣的市場。所以我認為這是KLA的新型SAM。我們內部有很多強大的驅動因素，我們認為這些因素正在推動製程控制和KLA市場份額的成長，我們正在利用我們預期在先進封裝領域看到的成長來增強這種成長，隨著時間的推移，先進封裝的成長速度可能會略快於WFE。

So it's a really encouraging opportunity. And I think as we start to move up the value chain in terms of new capability required, and I think it creates an opportunity for us to drive something in the neighborhood or better than general corporate averages on margins?

所以這確實是一個非常令人鼓舞的機會。我認為，隨著我們開始在價值鏈上向上發展，需要新的能力，這將為我們創造機會，使我們的利潤率達到甚至超過一般企業的平均水平？

Tom O'Malley, Barclays.

湯姆·奧馬利，巴克萊銀行。

I wanted to ask a bigger one that people have been trying to the earnings period here. I understand that it wasn't in the preamble. But Lam went out and talked about $100 billion of AI spend is roughly equivalent to $8 billion in WFE or additional spend. And they talked about most of that being related to memory. Do you agree with that statement? Or do you have any qualification for how you would look at that ratio?

我想問一個更大的問題，這個問題是大家一直在嘗試了解的，與這裡的獲利期有關。我知道它沒有寫在序言裡。但林先生卻公開表示，1,000億美元的人工智慧支出大致相當於80億美元的勞動支出或其他額外支出。他們談到，其中大部分都與記憶有關。您同意這個說法嗎？或者，您對如何看待這個比例有什麼看法？

I think in general, this is Rick. I think in general, if you think about what goes into a data center, the percent that is memory, the percent that is GPUs or logic and then the rest, you're probably at about -- you take $100 billion, then you could say that half of that would be semiconductor related and the intensity on that kind of gets you to that run rate.

我覺得總的來說，這就是里克。我認為總的來說，如果你想想資料中心的組成，記憶體佔比，GPU 或邏輯晶片佔比，剩下的部分，大概就是——假設有 1000 億美元，那麼可以說其中一半都與半導體相關，而半導體產業的強度也正是造成這種規模的原因。

But I don't think it's necessarily 50-50 in terms of memory and in terms of the logic side. So our view is you're pretty close. And then we would -- as we were just talking out in packaging, so we get closer to $10 billion on the $100 billion because of the investment that's not just semiconductor but packaging. And we think our opportunity in that is pretty good because, again, those are all the high challenging process control elements, kind of everything that we've been talking about larger die, more valuable die, more HBM is really challenging from a process control, getting more so and then packaging. So we're in general agreement, we would add back in the packaging part, and we think our participation in there is above our average intensity for the rest of the industry.

但我認為在記憶體和邏輯方面，兩者未必是五五開。所以我們認為你的說法相當接近正確答案。然後，正如我們剛才在包裝領域討論的那樣，由於不僅在半導體領域，而且在包裝領域的投資，我們距離 1000 億美元的目標更近了 100 億美元。我們認為我們在這方面的機會相當不錯，因為，再說一遍，這些都是極具挑戰性的製程控制要素，我們一直在談論的所有內容，更大的晶片、更有價值的晶片、更多的 HBM，從製程控制的角度來看都是真正的挑戰，而封裝也是如此。所以我們總體上達成共識，要把包裝部分重新納入考量，我們認為我們在該領域的參與程度高於行業平均。

Timothy Arcuri UBS.

提摩西‧阿庫裡，瑞銀集團。

Bren, Rick was talking about customers starting to want to get in line for. So I would imagine bookings were pretty good. So can you give us RPO? I know it was $7.9 billion last quarter. Where did it end this quarter?

布倫，里克當時正在談論顧客開始想要排隊購買。所以我覺得預約狀況應該相當不錯。那麼，你們能提供RPO（招募流程外包）方案嗎？我知道上個季度是79億美元。本季最終結果如何？

Tim, we don't -- we changed our disclosures, so we're not disclosing that anymore. But what I will tell you is that if you look at our lead times, our expectations for our lead times, as I said last quarter and they've been converging after a couple of years of elevated backlog related to a number of greenfield projects that have now shipped through. And if you look at the composition of our business going forward, really tied to our -- some of our long-standing customers that tend to operate in 6 months to kind of lead time windows.

提姆，我們不再披露這些資訊了——我們更改了披露方式，所以我們不再披露這些資訊了。但我可以告訴你們的是，如果你看看我們的交貨週期，以及我們對交貨週期的預期，正如我上個季度所說，在經歷了幾年與許多新建項目相關的積壓訂單之後，這些預期已經趨於一致，而這些新建項目現在已經交付完畢。如果你看看我們未來的業務組成，你會發現它實際上與我們的一些長期客戶息息相關，這些客戶的營運週期通常為 6 個月到 10 個月。

Our lead times have converged and I think, normalized between 7 and 9 months. Now if you go back and look at 2020, 2021, even go back historically for KLA, it used to be about 6 months. Now our customer base is broader today. And I think there's a combination of a little bit more new fab activity that will push those up. But the context that we provided in terms of our expectations for growth next year and how that plays through in terms of KLA's first half and second half is supported by an order flow that is consistent with those lead times.

我們的交貨週期已經趨於一致，我認為已經穩定在 7 到 9 個月之間。現在，如果你回顧 2020 年、2021 年，甚至回顧 KLA 的歷史，你會發現以前大約是 6 個月。如今我們的客戶群更加廣泛了。我認為，一些新的時尚活動將會推動這些數字上升。但是，我們針對明年成長預期以及這些預期如何影響 KLA 上半年和下半年所做的解釋，都得到了與這些交貨週期相符的訂單流的支持。

So I think that, that is how you should think about it in the 10-K each year, we will provide our -- as we have historically, we'll provide our backlog. And so that will give you an anchor point in terms of backlog on an ongoing basis. But 7 to 9 months, and it looks pretty consistent in that range as we go forward.

所以我認為，你應該這樣考慮每年的 10-K 表格，我們會像以往一樣提供我們的積壓工作。這樣一來，就能為你持續評估積壓工作量提供一個基準點。但預計需要 7 到 9 個月，而且隨著時間的推移，這個時間範圍似乎會保持穩定。

I guess you did give it that last quarter, Bren, is that like new this quarter?

布倫，我猜你上個季度確實把它賣掉了，這季度它是不是跟新的一樣？

Yeah. We changed our disclosure, Tim, and we highlighted that we were going to make that change back in the March quarter earnings results. We said when we started the new fiscal year beginning July 1, that disclosure would change. There were a number of reasons for that, the primary reason being the inconsistency in how that disclosure was being interpreted and reported across our industry. So we have aligned more closely with the disclosure that our peers have, and that's what we're going to do here going forward.

是的。提姆，我們修改了揭露方式，並且我們在三月的季度收益報告中強調了我們將做出這項改變。我們在7月1日開始的新財年說過，資訊揭露方式將會改變。造成這種情況的原因有很多，主要原因是整個產業對這種揭露的解讀和報告方式不一致。因此，我們更緊密地與同儕的資訊揭露保持一致，這也是我們今後要做的事。

Okay. Then I guess as my second question. So you said last quarter, Bren, that China was going to be for you, down 10% to 15% this year. But even to get down 10%, I have to have China down like $250 million Q-on-Q in December. Is that the right number that China is going to be back to like 30%, 31% in December?

好的。那麼，我的第二個問題是什麼呢？布倫，你上個季度說過，今年中國市場對你來說將會下降 10% 到 15%。但即使要下降 10%，我也必須讓中國在 12 月的季度環比下降 2.5 億美元。中國12月的感染率會回升到30%、31%左右，這個數字對嗎？

Yeah. In the December quarter, I think China will be high 20s. So yes, you're in the range. We'll see how the quarter finishes up. But I think in the high 20s is how modeling and then it translates into maybe 30%, maybe 31%, very consistent with the way I've talked about it all year long.

是的。我認為中國在12月當季的成長率將達到20%以上。是的，你在範圍內。我們拭目以待本季最終結果如何。但我認為，根據模型預測，這個數字會接近 30%，甚至 31%，這與我全年所說的非常一致。

And the other thing I'd say is, as you look at 2026, we think it's probably it comes down into the mid-20s. And obviously, some of that is driven by the export restriction, but also some general normalization that we've been talking about that would be coming. So we think it's likely settle somewhere in the mid-20s as we look at 2026, at least how we see it today.

另外我想說的是，展望 2026 年，我們認為它可能會降到 20 多歲。顯然，部分原因是出口限制，但也與我們一直在討論的即將到來的整體正常化進程有關。因此，我們認為到 2026 年，它可能會穩定在 20 多度左右，至少我們目前是這麼認為的。

Krish Sanker with TD Cowen.

Krish Sanker 與 TD Cowen 合作。

I have a two-part question. One is in the past, you've spoken about 2-nanometer gate all-around as 100 basis points improvement in share for you a, is that true -- still the case? And number 2 is, I think if I remember right, the advanced packaging share is about 50%. If I do the math on that $11 billion and $925 million, it seems like advanced packaging, WFE intensity is around mid-teens. Do you think advanced back-end inspect -- process control intensity is getting higher than front end? Or do you think it's still below

我有一個包含兩個部分的問題。一是過去您曾說過，2 奈米閘極可以為您帶來 100 個基點的市佔率提升，這是真的嗎——現在還是這樣嗎？第二點，沒記錯的話，先進包裝的份額約為 50%。如果我計算這 110 億美元和 9.25 億美元，似乎先進的封裝技術，WFE 強度大約在十幾個百分點左右。你認為後端高階檢測－製程控制的強度是否正在超過前端？或者你認為它仍然低於這個水平？

I think it's still below. I mean, just the first part of your question, we haven't -- look N2 has been a more intensive process control intensive node. Obviously, you have an architecture change that's that's quite significant. The other issue is that as you think about larger die that creates opportunities for more process control intensity. So -- and I think the litho scaling and litho layers with larger die design or HBC designs tends to drive more litho layers in the process as well.

我認為它仍然在下面。我的意思是，就你問題的第一部分而言，我們還沒有——你看，N2 一直是一個過程控制更密集的節點。顯然，你們的架構發生了相當大的變化。另一個問題是，隨著晶片尺寸的增加，對製程控制強度的要求也隨之提高。所以——而且我認為，隨著晶片尺寸增大或採用 HBC 設計，光刻製程的縮小和光刻層數的增加也會導致製程中出現更多的光刻層數。

So when we look at N2 overall versus N3, we do see an improvement in overall intensity. There is a share element to it, too. We're encouraged by some of the share movement we're seeing. But overall, that's how we see it. As it relates to advanced packaging, I think our logic share is higher than our memory share. We'll have more to say in terms of articulating how this this market breaks up.

所以當我們比較 N2 和 N3 的整體情況時，我們確實可以看到整體強度有所提高。這其中也包含著股份共享的因素。我們對目前看到的一些股價波動感到鼓舞。但總的來說，我們就是這麼看的。就先進封裝而言，我認為我們的邏輯電路份額高於記憶體電路份額。我們稍後會詳細闡述這個市場是如何劃分的。

But as I look at KLA share of the advanced packaging market, we're about 6%. And if you look at KLA's share of the PC market, it's closer to 8%. So the process control intensity is not as high and sampling rates are pretty elevated these days. And we'll see over time if that changes. Certainly, the need for more capability will be a requirement here moving forward. But we'll see how those things trade off over time as our customers move forward.

但就 KLA 在先進封裝市場的佔有率而言，我們約佔 6%。如果你看KLA在PC市場的份額，它接近8%。因此，如今過程控制強度並不高，取樣率也相當高。隨著時間的推移，我們將看到這種情況是否會改變。當然，未來對更高能力的需求將是不可或缺的。但隨著客戶的發展，我們將觀察這些因素如何權衡取捨。

Then I think you also mentioned that some of your front-end tools are being used to back end. I'm just kind of curious, especially on the macro inspection or inspection side, is there an overkill, like even ASNO spoke about introducing new tool for advanced packaging, would you consider introducing new tools for back-end packaging? Or are you going to use the front end tools? .

然後我想你還提到過，你們的一些前端工具也被用於後端開發。我只是有點好奇，尤其是在宏觀檢測或檢測方面，是否存在過度設計的情況，例如，ASNO 甚至談到了為高級包裝引入新工具，您是否會考慮為後端包裝引入新工具？還是你打算使用前端工具？。

No, I mean -- it's funny you say that because that was our initial reaction when the customers want us to put our front-end portfolio in the back end, we said, "Oh, are you sure?" Because the cost of that are you sure you need it, and they were quite certain that they needed it because of the -- just think about the cost of yield failure in packaging and how much it's worth to ensure that that's not happening. So that's what we've really seen in terms of that. They're not -- it's obviously not our most advanced tools, but it is.

不，我的意思是——你說得真有意思，因為當客戶要求我們將前端產品組合放在後端時，我們最初的反應是：“哦，你們確定嗎？”因為那東西的成本很高，你確定你需要它嗎？他們非常確定他們需要它，因為——想想包裝良率失敗的成本，以及確保這種情況不會發生的價值。這就是我們在這方面實際看到的情況。它們當然不是——顯然這不是我們最先進的工具，但它確實是。

They are systems that we use in the front end and that would have been considered the most advanced tools you go back a few years. So absolutely, they're the ones that pulled us in. I mean we did not come to them. It was almost the other way they said, we want you to to provide this capability. So when we think about the road map for packaging, and remember what we're talking about advanced packaging, we're in many ways, early innings because there's still other technology inflections that are going to go into HBM over the next several years as the rest of the market catches up with these advanced packages.

這些是我們前端使用的系統，在幾年前，它們會被認為是當時最先進的工具。所以毫無疑問，是他們吸引我們過來的。我的意思是，我們並沒有主動去找他們。他們幾乎是反過來說的，他們說我們希望你們提供這種能力。所以，當我們思考包裝的發展路線圖，並記住我們談論的是先進包裝時，我們在許多方面還處於早期階段，因為在未來幾年裡，隨著市場其他部分趕上這些先進包裝，還會有其他技術變革融入 HBM 中。

So we're really talking about a pretty small percentage of available packages being inspected at this high level. Now over time, people will learn more about it, and they won't inspect it the frequency. So that's why we see the growth will continue, but it won't continue at the rate we've had for the last couple of years, but it should outpace overall WFE growth. And our share position is great for 2 reasons. One is that we've got this capability. But beyond that, unlike our competitors that are coming from the back end, we have road maps and a lot of customers have tremendous value in that road map ability.

所以，實際上只有很小一部分可用的包裹會接受這種高水準的檢查。隨著時間的推移，人們會對此了解更多，他們就不會再去檢查它的頻率了。所以這就是為什麼我們看到成長將會繼續，但不會像過去幾年那樣快速成長，但應該會超過整體 WFE 的成長速度。我們的持股地位非常有利，原因有二。第一點是我們具備這種能力。但除此之外，與那些從後端開始的競爭對手不同，我們有路線圖，許多客戶都非常重視這種路線圖能力。

Charles Shi with Needham.

史志強與李約瑟。

I noticed that based on your Q4 guidance, the KLA process control revenue in DRAM is probably going to grow 50 percentage year-on-year. I don't think the overall DRAM WFE is growing that much. And I think historically, people don't really think the KLA as a DRAM house more of a leading-edge logic house. So wonder what's happening this year? Why is it growing this much faster than specifically, is it kind of tied to the EUV insertion DRAM? And how do we think about 2026 DRAM side of the process control growth?

我注意到，根據您第四季度的業績指引，KLA DRAM 製程控制收入可能會比去年同期成長 50%。我認為DRAM WFE的整體成長幅度並沒有那麼大。而且我認為從歷史上看，人們並不真正把科樂美視為一家DRAM公司，而更像是一家領先的邏輯晶片公司。想知道今年會發生什麼事嗎？為什麼它的成長速度比預期快得多？這是否與 EUV 光刻技術有關？那麼，我們該如何看待 2026 年 DRAM 製程控制的成長呢？

Yeah. It's an interesting observation. For those of us who have been around for a long time, we remember when DRAM was actually leading technologically and was the biggest market for inspection for KLA. And what happened was, for many years, there was a bit of a holiday in terms of design rules in DRAM and the need for process control and what the use case was.

是的。這是一個有趣的觀察。對於我們這些入行很久的人來說，我們還記得 DRAM 曾經在技術上處於領先地位，並且是 KLA 檢測的最大市場。多年來，DRAM 的設計規則、製程控制的需求以及應用場景方面都出現了一段相對寬鬆的時期。

But when you get into what's going on with -- especially around the high-bandwidth memory, and the challenges that people have relative to the new design rules, we're actually seeing, in some cases, higher sensitivity requirements in some cases for DRAM on some layers that we're even seeing in logic. So we've had a bit of a reversal of some -- in some areas, and we've seen big adoption of early on as people are developing these processes and then realizing they don't have a lot of process margin. So that's the other thing, is there -- and when they start EUV, then they're using our systems for precheck.

但是，當你深入了解正在發生的事情——尤其是在高頻寬記憶體方面，以及人們在新設計規則方面面臨的挑戰時，我們實際上看到，在某些情況下，DRAM 在某些層上的靈敏度要求更高，甚至在邏輯層上也是如此。因此，在某些領域，情況出現了一些逆轉——我們看到，隨著人們開發這些流程，早期就得到了廣泛的採用，然後他們意識到他們的流程餘地並不大。所以還有一點——當他們開始使用 EUV 時，他們會使用我們的系統進行預檢查。

So you see a lot of applications happening, and that's really what's been driving this increase. We thought it would happen. Years ago, we were hoping it would happen sooner, but it's definitely we're seeing leadership in some areas in terms of the need for process control as they retool these DRAM facilities to deal with some of the new market requirements.

所以你會看到很多申請湧現，而這正是推動這一成長的真正原因。我們早就料到了。幾年前，我們希望這種情況能更早發生，但現在我們確實看到，在某些領域，隨著 DRAM 工廠進行改造以應對一些新的市場需求，在製程控制方面，一些企業展現出了領導地位。

Look, the introduction of EUV was certainly a factor in terms of process control intensity. We think overall, probably changed at about 1 point. But then if you look at the requirements for HBM, we think it's increased at another point or so. Rick talked about a lot of the issues. The other thing you have to keep in mind is that the reliability requirements in a stack of DRAM chips and an HBM device, the device is only as good as the weakest DRAM.

你看，EUV 的引入無疑是製程控制強度的因素。我們認為總體上可能變化了大約 1 個點。但是，如果你看一下 HBM 的要求，我們認為它在另一個方面有所提高。里克談到了很多問題。你還需要記住一點，在 DRAM 晶片和 HBM 設備堆疊中，可靠性要求很高，整個設備的性能取決於其中性能最差的 DRAM。

And so the performance requirements, the process variability that the customer can accept, you can't spin these devices that go into an HBM integration. So there are a number of things that are happening there that are positive for process control intensity.

因此，性能要求、客戶可以接受的製程變異性，你不能隨意更改這些用於 HBM 整合的設備。因此，那裡正在發生一些對過程控制強度有利的事情。

Got it. Maybe a quick follow-up. I think going back a couple of years ago, you guys talked about the delayed pellicleization what that means to the KLA mask inspection portfolio. I thought the thesis was that without pellicle, your existing mask inspection plus print check probably works the best. But with the pellicle, maybe actinic works better.

知道了。或許需要快速跟進。我想回顧幾年前，你們討論過延遲薄膜化對 KLA 口罩偵測組合的影響。我原以為論文的論點是，如果沒有薄膜，你現有的掩模檢測加上列印檢查可能是最好的方法。但是對於表膜來說，光化法可能會更好。

I know this has been an ongoing discussion for many years, but we are hearing recent -- some recent reporting out of Taiwan, talking about your leading foundry customers potentially converting a fab into a pellicle fab and wonder what that means to your overall strategy on mask inspection. If you shed some light on that?

我知道這個問題已經討論了很多年，但我們最近聽到了一些來自台灣的報道，說你們的主要代工廠客戶可能會將一家晶圓廠改造成薄膜晶圓廠，我們想知道這對你們的整體掩模檢測戰略意味著什麼。您能對此做一些解釋嗎？

Yeah. So rather than going to specific strategies, specific customers have, I can tell you we're in conversations with all the leading mask manufacturers in the fabs in terms of what is their strategy relative to reticle qualification and requalification because it's a critical area. And and although we don't have all the pieces in that, we have many of the pieces because there's many points along the way, whether you're calling the reticles in the fab, you're doing recall or you're doing verification and print check. So we're heavily involved in those conversations.

是的。因此，與其討論具體客戶的具體策略，不如告訴大家，我們正在與晶圓廠所有領先的掩模製造商進行對話，了解他們關於光罩認證和再認證的策略，因為這是一個關鍵領域。雖然我們還沒有掌握其中的所有信息，但我們掌握了很多信息，因為過程中有很多環節，無論是在製造廠調用光刻膠，還是進行召回或進行驗證和打印檢查。所以我們積極參與了這些討論。

As you know, the challenges with pellicleization and the trade-off is throughput. And so that's always the issue of using pellicles as you give up some of the live performance. There have been advancements and we're well positioned to support those. But when we talk about a record year in our reticle business, obviously, people are buying with the future in mind as they do that, and we continue to see growth going forward, we feel pretty good about our ability to participate as we go forward in terms of any scenario that plays out. But I can tell you we're very heavily involved in customers with those conversations.

如您所知，薄膜化面臨的挑戰和權衡之處在於產量。所以，使用薄膜總是會面臨一個問題，就是它會犧牲一些現場表演的效果。已經取得了一些進展，我們有能力支持這些進展。但當我們談到瞄準鏡業務創紀錄的一年時，顯然，人們在購買時會考慮到未來，而且我們繼續看到成長，我們對未來在任何情況下都能參與其中的能力感到非常滿意。但我可以告訴你，我們與客戶的這些對話非常深入。

Chris Caso, Wolfe Research.

克里斯‧卡索，沃爾夫研究公司。

I guess the first question is regarding the commentary on '26 because you have a little more detail about what you're seeing first half versus second half. I assume that it's a combination of advanced logic and DRAM driving that second half? And is that just simply a function of where your lead times are that some of the improvements we've seen over the past couple of months are just now flowing through in orders given where the lead times are right now.

我想第一個問題是關於 '26 的評論，因為你們對上半場和下半場的對比有更詳細的描述。我猜想後半部是由先進的邏輯電路和DRAM共同驅動的？這是否僅僅與你們的交貨週期有關？鑑於目前的交貨週期，我們在過去幾個月中看到的一些改進現在才體現在訂單中。

Yeah. As we said in the prepared remarks, and I think the first half is maybe flat to slightly up, and we'll see as we move forward, where that ends up. But at least that's how it looks today. And I think you'll see growth accelerate more into the second half. Lead time discussions, we're talking about slots, but I think there's some facility dynamics also that are influencing some of the timing. But would expect advanced logic and the broadening of the investment that I mentioned to be a driver into the first half.

是的。正如我們在準備好的演講稿中所說，我認為上半年可能會持平或略有上升，至於最終結果如何，我們拭目以待。但至少現在看來是這樣。我認為下半年成長速度將會加快。在提前期討論中，我們談到了場地位置，但我認為也有一些場地動態因素影響了時間安排。但我認為，我提到的先進邏輯和投資範圍的擴大將成為上半年的驅動力。

But there's continued momentum on the DRAM front, too. So we're pretty encouraged by what we're seeing there. Obviously, it'll be offset by some weaker numbers out of China, but the leading edge dynamics are encouraging.

但DRAM領域也保持持續的發展動能。所以，我們對目前看到的情況感到非常鼓舞。顯然，中國的一些疲軟數據會抵消這一成長，但領先優勢的動態令人鼓舞。

Just a question on gross margins as we go into '26 also. And with some of the mix changes, particularly with China probably coming down as a percentage of the mix, anything we should think about with respect to gross margins as we start modeling through '26?

進入2026年，我還想問毛利率的問題。考慮到產品組合的一些變化，特別是中國產品在產品組合中所佔比例可能會下降，我們在開始模擬 2026 年的毛利率時，應該考慮哪些因素？

Yeah, Chris, I'll give a little bit more specific guidance on margins and operating expense expectations based on our revenue picture at next quarter. Gross margins for KLA are generally impacted, I'd say, almost exclusively impacted by what we sell, not so much who we sell to. And where we sell it. So it really is a factor of how it's impacting certain product types that we have a pretty extensive portfolio of products, the broadest in our segment of the industry.

是的，克里斯，我會根據下個季度的收入情況，對利潤率和營運費用預期給出更具體的指導。我認為，KLA 的毛利率通常幾乎完全取決於我們銷售的產品，而不是我們的銷售對象。以及我們的銷售地點。所以，這確實取決於它對某些產品類型的影響，而我們擁有相當廣泛的產品組合，是業界同類產品中最廣泛的。

And depending on what you're buying, it can carry different margin profiles. Certain parts of the market, like packaging tend to carry a more dilutive stream, as I mentioned earlier, I think, over time, that goes from being a headwind to a tailwind. Service tends to grow, has a dilutive gross margin, but we believe an accretive operating margin. So you do have some of the moving parts. The tariff impact, when you compare year-to-year, we really that's more second half dynamic this year. And my hope is given some of the things that we have going on in the company in terms of assessing how we can try to mitigate that exposure that, that becomes less of a headwind over time.

根據你購買的商品不同，其利潤率也會有所不同。正如我之前提到的，某些市場領域，例如包裝領域，往往會帶來更稀釋性的市場波動。我認為，隨著時間的推移，這會從逆風變成順風。服務業往往具有成長潛力，但會稀釋毛利率，不過我們相信其營業利潤率會提高。所以，你的確有一些活動部件。與去年同期相比，關稅的影響在今年下半年更為顯著。我希望，鑑於公司目前正在進行的一些評估工作，例如如何減輕這種風險，隨著時間的推移，這種不利因素會逐漸減少。

I think structurally, we're in a world where we'll be dealing with higher tariffs, but I do think there are things we do in terms of how we operate the company, where there's some ROI in terms of just how we move parts around the world and how we reduce the leakage and drawback scenarios as we understand and track different parts attributes that help reduce some of that. So I think there are a number of dynamics at play that will become less of a headwind over time.

我認為從結構上看，我們身處一個將面臨更高關稅的世界，但我確實認為，我們在公司運營方面所做的一些事情，例如我們如何在全球範圍內運輸零部件，以及如何減少洩漏和退稅情況，都能帶來一定的投資回報。因為我們了解並追蹤不同的零件屬性，這有助於減少其中的一些問題。所以我認為，目前有很多因素在起作用，隨著時間的推移，這些因素帶來的阻力會逐漸減少。

And depending on the growth of the business, obviously, that will happen as volume will have an impact, too. So I want have more to say about, I think those are some of the context behind the different moving parts.

當然，隨著業務成長，銷量也會產生影響，這種情況顯然會發生。所以我想再多說幾句，我認為這些是不同組成部分背後的一些背景資訊。

Shane Brett, Morgan Stanley.

Shane Brett，摩根士丹利。

I wanted to follow up on Charles' earlier question, but your memory customers have been talking of CapEx growth into 2026. But just given how strong this December quarter DRAM guide is, how should we think about your memory growth expectations into next year relative to this really strong December quarter?

我想就查爾斯之前提出的問題做個後續討論，但是你們的記憶體客戶一直在談論到 2026 年的資本支出成長。但鑑於 12 月季度 DRAM 市場指引如此強勁，我們應該如何看待明年記憶體市場的成長預期，並將其與 12 月季度的強勁表現進行比較呢？

Yeah, I apologize. I know that Charles asked that and we didn't answer that question. There are definitely some timing factors that are influencing process control, timing relative to other products. As I look at where we're at, I mean, this year has been a very strong year in DRAM for the company. I would expect next year to be a growth year as well. And I think a lot of these announcements, I think, as we start to see that play out and we'll see whether that is more of a second half dynamic into next year and how much that sort of carries forward. But what is clear is, across all of our customers, I expect them to spend more and to see growth in our DRAM investment from our customers into next year.

是的，我很抱歉。我知道查爾斯問過這個問題，但我們沒有回答。確實有一些時間因素會影響製程控制，例如與其他產品的時間關係。縱觀公司目前的狀況，今年對公司而言，DRAM業務是表現非常強勁的一年。我預計明年也將是成長之年。我認為，隨著我們開始看到這些公告陸續發布，我們將看到這是否會成為明年下半年的動態，以及這種動態將在多大程度上延續下去。但可以肯定的是，我預計我們所有客戶的支出都會增加，明年客戶在 DRAM 的投資也會成長。

(Operator Instructions)

（操作說明）

Edward Yang, Oppenheimer.

愛德華楊，奧本海默。

Rick, Bren, most of my questions have been already answered, but maybe you could talk about your outlook for foundry-related revenue opportunities outside the dominant Taiwanese customer. I think at least one major foundry has historically underinvested in yield improvement tools. Maybe that tune is changing. So are you seeing any change in engagement there?

Rick、Bren，我的大部分問題都已經得到了解答，但或許你們可以談談在台灣以外的代工相關收入機會的前景。我認為至少有一家大型晶圓代工廠在提高良率的工具方面歷來投入不足。或許這種風向正在改變。那麼，您認為這方面的參與度有任何改變嗎？

Yeah. I would say that we're encouraged by, as we said earlier, we're encouraged by the broadening of investment that we're seeing at the leading edge as we go into next year.

是的。正如我們之前所說，我們對明年即將到來之際，前沿領域投資範圍的擴大感到鼓舞。

Yeah. The conversation qualitatively, the conversations we have with customers that are looking to, as you mentioned, not the leader that are looking to do advanced logic, we do have a lot of conversations around what they're asking our advice, what do they need to be successful, especially since they maybe haven't been pressing the latest nodes. And so there's a lot of conversation we have about the specific things they're trying to accomplish. It depends on the dynamics, it depends on their mix. It depends on their die size. It depends on what their expectations are for how fast they want to ramp.

是的。從定性角度來看，我們與那些尋求（正如您所提到的）高級邏輯的客戶進行的對話，而不是與那些尋求高級邏輯的領導者進行的對話，我們確實進行了很多關於他們向我們尋求建議的內容、他們需要什麼才能成功等方面的對話，尤其是在他們可能還沒有關注最新節點的情況下。因此，我們經常就他們試圖實現的具體目標進行討論。這取決於具體情況，取決於它們的混合比例。這取決於他們的模具尺寸。這取決於他們對成長速度的預期。

But it is, for sure, a lot of conversations we're having. And as Bren says, we feel pretty good about those discussions. I think for a lot of people, the process control part, they haven't really fully understood how the world has changed in the last few nodes. And so those conversations are ongoing.

但可以肯定的是，我們確實進行了許多對話。正如布倫所說，我們對這些討論感到非常滿意。我認為對很多人來說，在過程控制方面，他們還沒有真正完全理解世界在過去幾個節點中發生了怎樣的變化。因此，這些對話仍在繼續。

And just as a quick follow-up. Yesterday, the leading AI accelerator company talked about a $0.5 trillion backlog. We're seeing these flurry of deals involving the major AI lab players. From your vantage point, can you level set for us, is the semi cap industry position to serve that scale of demand? Or is the ecosystem discounting some of these projections as aspirational.

最後再補充一點。昨天，領先的人工智慧加速器公司談到了高達 0.5 兆美元的積壓訂單。我們看到，各大人工智慧實驗室紛紛達成交易。從您的角度來看，您能否為我們評估一下，半導體產業能否滿足如此大規模的需求？或者，生態系統是否將其中一些預測視為理想化的願景而忽視？

Yeah. I don't -- I guess I would maybe pass it a little bit differently in the sense that I think the semiconductor industry has been prudent in terms of adding capacity. And right now, when we talk to -- when we try to reconcile the external discussions about CapEx and would that translate into in terms of wafers, not enough wafers will be available to achieve those objectives in the time frame. And our view is that it's not necessarily a bad thing. It means it's not going to -- it's unlikely to overheat if those forecasts remain intact because there's more gating factors than there are people making announcements. It's easier to make an announcement about investment in the data center than it is to build a new fab. So I think it's going to take some time for the industry to absorb and support the capacity demands implied by all the public announcements.

是的。我不這麼認為——我想我可能會以不同的方式看待這個問題，因為我認為半導體產業在增加產能方面一直很謹慎。而現在，當我們與外界討論資本支出以及這是否會轉化為晶圓時，我們發現，在規定的時間內，沒有足夠的晶圓可供使用，以實現這些目標。我們的觀點是，這未必是件壞事。這意味著它不會過熱——如果這些預測保持不變，它不太可能過熱，因為限制因素比發佈公告的人數要多。宣布對資料中心進行投資比建造一座新的晶圓廠要容易得多。所以我認為，業界需要一些時間來消化和滿足所有公開聲明所隱含的產能需求。

Blayne Curtis, Jefferies.

布萊恩·柯蒂斯，傑富瑞。

I just want to go back to the DRAM comments. You talked about the increasing capital intensity. I'm just kind of curious about kind of the conversations. Obviously, massive numbers have been thrown out there. Just kind of curious, I think someone asked us prior about what -- was that always the plan to have DRAM up this much? Or have things been pulled in? And then I guess, just if you could elaborate a little bit more color on those conversations you said where they're looking for capacity, like what's holding it up? Is it just they need fab space? Or are they unsure about the demand? Just any color there would be great.

我只想回到DRAM的評論部分。你談到了資本密集度不斷提高的問題。我只是有點好奇大家在聊些什麼。顯然，已經有大量的數字被拋了出來。我只是有點好奇，我想之前有人問過我們——是不是一直都有把 DRAM 容量提升到這麼高的計劃？還是事情已經被控制住了？然後，我想，您能否再詳細闡述您剛才提到的那些關於他們尋求產能的對話，例如是什麼阻礙了產能的提升？他們只是需要製造空間嗎？還是他們對市場需求沒有把握？任何顏色都可以。

Yeah, I would say that you definitely feel there's more urgency on the DRAM front in terms of timing. We'll have to see how that plays out in terms of slots as we move into next year given our lead times, what we can accommodate and obviously, we try to work closely with our customers on that front. But there's certainly. And I think from a pricing point of view, I think the dynamics that are driving HBM pricing overall, there's definitely a sense of urgency from our customer base. And as I said earlier, across the top 3, I do expect higher investment levels next year than we're seeing here in '25.

是的，我認為在DRAM方面，時間安排上確實更加緊迫。考慮到我們的提前期、我們能夠滿足的需求，以及我們顯然會在這方面與客戶密切合作，我們還需要看看明年檔期安排會如何變化。但肯定有。我認為從定價的角度來看，推動 HBM 整體定價的因素，肯定是我們的客戶群的緊迫感。正如我之前所說，我預計明年前三大投資領域的投資水準將高於我們今年（2025年）所看到的水準。

I think the other way to think about it, if you think that there's 3 components, primary components that go into supporting these AI infrastructure build-outs, you have obviously the GPUs and the accelerators around those, you have the packaging and then you have the memory. I think you're going to see over time that you go in phases of which one seems to be short supply. And we kind of went through a phase of that with packaging being behind. And now I think the realization by a lot of our customers is they might -- there might be more opportunity in memory than they thought, and those are accelerated from what they even told us a few months ago.

我認為換個角度來看，如果你認為支撐這些人工智慧基礎設施建設的主要組成部分有三個，那麼顯然包括 GPU 和圍繞它們的加速器、封裝以及記憶體。我認為隨著時間的推移，你會發現供應會經歷幾個階段，每個階段似乎都會短缺。我們曾經歷過一段包裝滯後的時期。現在我認為，我們許多客戶都意識到，儲存領域可能蘊藏著比他們想像中更大的機遇，而且這種趨勢比他們幾個月前告訴我們的還要加速。

So I think that that's part of what you're seeing. I think everybody is a bit amazed by the number of applications that are being realized using AI. Even inside of KLA, we keep coming up with new ways to leverage the technology. And I don't think we're the only ones doing that. So I think the memory guys are right now feeling well, there's more opportunity if they could add capacity. And those are kind of the conversations because they realize that it takes longer, as I said, to ramp the supply chain than it does to make these announcements about CapEx. Thank you.

所以我覺得這就是你所看到現象的一部分。我認為大家都會對人工智慧的應用數量之多感到驚訝。即使在 KLA 內部，我們也不斷想出利用這項技術的新方法。而且我認為我們不是唯一這樣做的。所以我覺得儲存設備部門現在感覺不錯，如果他們能增加產能，機會就更多了。而這些正是人們討論的話題，因為他們意識到，正如我所說，擴大供應鏈規模所需的時間比宣布資本支出所需的時間還要長。謝謝。

Jim Schneider, Goldman Sachs.

吉姆‧施耐德，高盛集團。

Maybe just one follow-up relative to the earlier question about the diversification in your leading-edge logic and foundry customer base. Is that something that's more on the inquiry level at this point? Or are you actually seeing that either in your order book or in the form of forecast from those customers at this stage?

或許可以就先前關於貴公司尖端邏輯晶片和代工客戶群多元化的問題做一個後續問題。這是否屬於現階段需要探究的問題？或者，您是否在現階段的訂單簿或客戶預測中實際看到了這一點？

I would say we are seeing it, and it's certainly as it informs our views of next year. We're seeing it in the order forecast. And the -- as Rick talked about earlier, I think there's a lot of collaborative discussion about how we can help navigate and ramp and drive time to results in this capacity.

我認為我們正在看到這種情況，而且這無疑會影響我們對明年的看法。我們在訂單預測中看到了這一點。正如里克之前提到的那樣，我認為有很多合作討論，探討我們如何在這個方面幫助引導、加速並加快取得成果的速度。

And then maybe just as a quick follow-up. Maybe you can help us just refresh your expectations about kind of confirming mid-teens is the right level for service growth in 2025 and maybe give us a sense about whether that could accelerate next year?

然後或許可以簡單地跟進一下。或許您可以幫助我們重新評估您對 2025 年服務成長的預期，確認 15% 左右的成長率是否合適，並讓我們了解明年成長速度是否會加快？

Yeah. We've seen service actually pick up a little bit here. It's been a little stronger than we expected. Obviously, we've had some FX benefits, but we've also had some strengthening as utilization rates have gone up. We've seen some strengthening in some of our billable business. So I think our service growth will be in our target range of 12% to 14% this year. And I would expect right now as I look at next year, that will be in the same range as well. So I think we feel pretty good about where that is. both based on the growth expected in the installed base, the life time increase in the tools, the incremental value that's coming from the complexity in the systems and how that's affecting the pricing as it relates to contracts, the opportunities in the acquired businesses that we've acquired over the last few years to drive their service business and new requirements that we're seeing, this is a factor in '25 as well. But as you think about packaging, a different service model for packaging, but also for DRAM, where where the utilization rates to some of the earlier questions have been higher and higher expectations of performance of our system.

是的。我們發現這裡的服務確實有改善。比我們預想的要強一些。顯然，我們從外匯市場獲得了一些好處，但隨著利用率的上升，我們也獲得了一些利好。我們看到部分計費業務有所增強。所以我認為我們今年的服務成長率將達到12%至14%的目標範圍。展望明年，我預計情況也會大致相同。所以我覺得我們對目前的情況相當樂觀。這既基於預期安裝基礎的成長，也基於工具使用壽命的延長，也基於系統複雜性帶來的增量價值及其對合約定價的影響，以及過去幾年我們收購的企業在推動其服務業務方面的機會，還有我們看到的新需求，這些因素也會影響到2025年的業績。但當你考慮封裝時，你會發現封裝需要不同的服務模式，DRAM 也是如此，因為先前一些問題的使用率越來越高，我們對系統效能的期望也越來越高。

So I think there's some new opportunities for growth there that, frankly, if you go back a couple of years, I don't think we fully anticipated, both on the packaging front but also on the high bandwidth memory front supporting HBC.

所以我認為這裡存在一些新的成長機會，坦白說，如果你回顧幾年前，我認為我們並沒有完全預料到這些機會，無論是在封裝方面，還是在支援 HBC 的高頻寬記憶體方面。

Timm Schulze-Melander, Rothschild & Co.

蒂姆·舒爾茨-梅蘭德，羅斯柴爾德公司

Actually, I just had one with respect to your outlook for next year. You talked about this broadening in demand for 2026. Could I just ask, could you just paint some color around to what extent that already bakes high NA engagements or whether that would be an upside to your outlook for '26?.

其實，我只是想問你對明年的展望。您談到了2026年需求的擴大。請問，您能否具體說明一下，這在多大程度上已經預示著北美地區的高參與度，或者這是否會成為您對2026年前景的利好？

One thing, look, on the R&D front, there's a lot of collaboration with customers, how that translates into revenue is not part of the outlook. It's most of what the engagement is, is on ramping the continuation of the N2 ramp, our 2-nanometer ramp across our customer base. And so isn't influenced by the adoption of high NA in that time frame.

有一點要注意，在研發方面，我們與客戶進行了許多合作，但這些合作如何轉化為收入並不在考慮範圍之內。此次合作的大部分內容，都是為了繼續推進 N2 製程的推廣，也就是我們在客戶群中推廣 2 奈米製程。因此，它不受該時間段內高NA採用率的影響。

Brian Chin, Stifel.

Brian Chin，Stifel。

I appreciate that. Maybe a question here. I think there was a reference earlier about the potential for some acceleration in second half of next calendar year. I'm sure it's not one single thing, but how much of that second half outlook is tied to new clean room space availability, knowing that some of your tools, probably some of the first that go into a greenfield fab.

我很感激。這裡或許有個問題。我認為之前曾提到過，明年下半年可能會出現加速成長的趨勢。我確信這並非單一因素造成的，但下半年的前景有多少與新的無塵室空間可用性有關？要知道，你們的一些工具，可能是新建工廠最先使用的工具之一。

Look, there are some issues, I think, with space constraints potentially. It's not, of course, on every customer, but I do think that that it could affect some timing as we move into the second half of next year and into '27 as you start to think about the next node ramping and some of the new fabs coming online in memory. So right now, I don't think space is going to be an issue. Obviously, that would depend on the strength of demand, yes, that could change. But for now, it's certainly a factor, but I don't think it's a big factor as it stands today, and we'll see how things go.

我認為，空間限制方面可能會存在一些問題。當然，這並非對每個客戶都會產生影響，但我認為，隨著我們進入明年下半年和 2027 年，當我們開始考慮下一個節點的產能爬坡以及一些新的晶圓廠即將上線時，這可能會影響一些時間表。所以目前來看，我認為空間不會是個問題。顯然，這取決於需求的強弱，是的，這種情況可能會改變。但就目前而言，這當然是一個因素，但我認為就目前的情況來看，這並不是一個很大的因素，我們拭目以待。

Maybe if I had a time for a quick follow-up. On advanced packaging, I think to date, a lot of your inspection business strength has been strongly tied to logic. How much of your continued optimism on market and KLA growth in package next year involves expansion opportunities in HBM packaging?

或許如果我有時間的話，可以快速跟進。在先進包裝領域，我認為到目前為止，你們的檢測業務優勢很大程度上與邏輯密切相關。您對明年包裝市場和關鍵產品領域成長的持續樂觀態度中，有多少是源自於 HBM 包裝領域的擴張機會？

I feel good about market share, both in logic and in memory on the packaging front. We've seen positive trends in both segments over the last couple of years, and I think that, that continues into next year.

我對市場份額感到滿意，無論是在邏輯電路還是記憶體封裝方面。過去幾年，這兩個領域都呈現出正面的發展趨勢，我認為這種趨勢將延續到明年。

Thank you, Brian, and thank you, everybody, for your interest in KLA. We appreciate your time today, and we'll be in touch. I'll turn it back over to the operator for any closing instructions.

謝謝布萊恩，也謝謝大家對KLA的關注。感謝您今天抽出時間，我們會與您聯繫。我會把它交還給操作員，聽取任何關門指示。

Thank you. And this does conclude today's KLA Corporation quarter 2025 post earnings call. Please disconnect your line at this time, and have a wonderful day.

謝謝。至此，KLA公司2025年季度財報電話會議結束。請您暫時斷開電話，祝您有美好的一天。