科林研發 (LRCX) 2026 Q2 法說會逐字稿

Good day, and welcome to the Lam Research Corporation December 2025 earnings conference call. (Operator Instructions) Please note that this event is being recorded.

大家好，歡迎參加 Lam Research Corporation 2025 年 12 月的財報電話會議。（操作說明）請注意，本次活動正在錄影。

I would now like to turn the conference over to Ram Ganesh, Vice President of Investor Relations. Please go ahead.

現在我將把會議交給投資者關係副總裁拉姆·加內什。請繼續。

Thank you, and good afternoon, everyone. Welcome to the Lam Research quarterly earnings conference call. With me today are Tim Archer, President and Chief Executive Officer; and Doug Bettinger, Executive Vice President and Chief Financial Officer.

謝謝大家，大家下午好。歡迎參加Lam Research季度財報電話會議。今天陪同我出席的有總裁兼執行長提姆·阿徹，以及執行副總裁兼財務長道格·貝廷格。

During today's call, we will share our overview on the business environment and we'll review our financial results for the December 2025 quarter and our outlook for the March 2026 quarter. The press release detailing our financials results was distributed a little after 1:00 PM Pacific Time. The release can also be found on the Investor Relations section of the company's website along with the presentation slide that accompanied today's call.

在今天的電話會議上，我們將分享我們對商業環境的概述，並回顧我們 2025 年 12 月季度的財務業績以及我們對 2026 年 3 月季度的展望。詳細介紹我們財務表現的新聞稿已於太平洋時間下午 1 點剛過發布。該新聞稿以及今天電話會議的簡報也可在公司網站的投資者關係部分找到。

Today's presentation and Q&A include forward-looking statements that are subject to risks and uncertainties reflected in the risk factors disclosed in our SEC public filings. Please see accompanying slides in the presentation for additional information.

今天的演示和問答環節包含前瞻性陳述，這些陳述受到風險和不確定性的影響，這些風險和不確定性已在我們的美國證券交易委員會公開文件中披露。更多資訊請參閱簡報中的配套幻燈片。

Today's discussion of our financial results will be presented on a non-GAAP financial basis unless otherwise specified. A detailed reconciliation between GAAP and non-GAAP results can be found in the accompanying slides in the presentation. This call is scheduled to last until 3:00 PM Pacific Time. A replay of this call will be made available later this afternoon on our website.

除非另有說明，我們今天對財務表現的討論將以非GAAP財務報表為基礎。簡報中的幻燈片詳細列出了 GAAP 和非 GAAP 結果之間的對比。本次通話預計將持續到太平洋時間下午3點。今天下午晚些時候，我們將在網站上提供本次通話的錄音回放。

And with that, I'll hand the call over to Tim.

那麼，我就把電話交給提姆了。

Thank you, Ram, and good afternoon, everyone. We ended calendar year 2025 on a strong note, delivering December quarter revenues ahead of the midpoint of our guidance. Gross margins, operating margins, and EPS all exceeded the high end of the range. Our performance demonstrates continued strong execution in an accelerating semiconductor demand environment.

謝謝你，拉姆，大家下午好。我們在 2025 年日曆年結束時表現強勁，12 月季度的營收超過了我們預期的中位數。毛利率、營業利潤率和每股盈餘均超過了預期範圍的上限。我們的業績表明，在半導體需求加速成長的環境下，我們依然保持著強勁的執行力。

At our Investor Day event last year, we outlined our tremendous opportunity to expand our market and gain share at every successive technology node. Vertical scaling of device and packaging architectures is driving higher deposition and etch intensity and moving the market to our strengths. The vision we shared was to more than double Lam's revenue and profit over the next five years. Today, we are well on our way.

在去年的投資者日活動上，我們概述了我們在每個技術節點上拓展市場和獲得市場份額的巨大機會。裝置和封裝架構的垂直擴展正在推動更高的沉積和蝕刻強度，並將市場推向我們的優勢。我們共同的願景是，在未來五年內，讓林氏的收入和利潤翻倍以上。今天，我們已走在正確的道路上。

With the industry ramping capacity and adopting new technologies to meet the demands of the AI transformation, Lam's deposition and etch capabilities are proving to be key enablers in the transition to gate all around transistors, backside power distribution, higher performance materials, and 3D advanced packaging. We prepared for this moment, launching an array of new products and advanced services targeted at broadening our exposure across DRAM, leading-edge foundry-logic, and NAND.

隨著產業產能不斷提升，並採用新技術來滿足人工智慧轉型的需求，Lam 的沉積和蝕刻能力正在成為向全包圍閘極電晶體、背面電源分配、更高性能材料和 3D 先進封裝過渡的關鍵推動因素。我們為這一刻做好了準備，推出了一系列新產品和先進服務，旨在擴大我們在 DRAM、領先的晶圓邏輯和 NAND 領域的業務範圍。

We've also invested in expanding our manufacturing and R&D footprint to increase operational velocity in response to strong customer demand. In 2025, we achieved record revenues of more than $20 billion and expanded our served available market, or SAM share, of WFE into the mid-30% range. This marks solid progress toward our multi-year goal of being in the high 30s.

我們也投資擴大了製造和研發規模，以提高營運速度，從而響應客戶的強勁需求。2025 年，我們實現了超過 200 億美元的創紀錄收入，並將我們在 WFE 中的服務可用市場份額 (SAM 份額) 擴大到 30% 左右。這標誌著我們在實現多年目標（即達到30%以上）方面取得了堅實進展。

Our shipped share of WFE grew by well over 1 percentage point year on year, and our CSBG business hit key milestones, with the size of our installed base topping 100,000 chambers and revenue growing faster than the increase in installed base units. We are proud of these accomplishments, but there's even more to come.

我們的 WFE 出貨份額年增超過 1 個百分點，我們的 CSBG 業務也達到了關鍵里程碑，我們的安裝基礎規模超過 10 萬個腔室，收入增長速度超過了安裝基礎單元的增長速度。我們為這些成就感到自豪，但未來還有更多精彩值得期待。

The AI transformation is driving industry spending higher. In 2025, WFE came in close to $110 billion. Our initial 2026 view is for WFE to be in the $135 billion range, with the growth in spending remaining constrained by a shortage of available cleanroom space. Chip makers have been public about their efforts to alleviate constraints, but they've also commented on sold-out conditions persisting, indicating the magnitude of the challenge.

人工智慧轉型正在推動產業支出增加。到 2025 年，WFE 的規模將接近 1,100 億美元。我們最初預計到 2026 年，WFE 的支出將達到 1,350 億美元左右，但支出成長仍將受到可用無塵室空間短缺的限制。晶片製造商已公開表示正在努力緩解供應限制，但他們也評論說，供應短缺的情況依然持續，這表明挑戰的嚴峻性。

We expect WFE this year to be weighted to the second half with robust growth in investments across all three device segments led by DRAM and leading-edge foundry-logic. All indications are that we are still in the early stages of the AI buildout and end markets are signaling a strong appetite for greater compute and storage capability at both the device and package level.

我們預計今年 WFE 的成長將主要集中在下半年，所有三個裝置領域的投資都將強勁成長，其中 DRAM 和領先的晶圓邏輯裝置將引領這一成長。種種跡象表明，我們仍處於人工智慧建設的早期階段，終端市場對設備和封裝層面的運算和儲存能力有強烈的需求。

In foundry-logic, customers are accelerating migration to nodes employing gate all-around transistors. If you recall, we've previously said that the transition to gate all-around equates to roughly $1 billion in incremental Lam SAM for every 100,000 wafer starts per month capacity. Given the 3D nature of gate all-around structures, we are well positioned with our deposition and etch portfolio to gain share within this segment.

在晶圓代工邏輯領域，客戶正在加速向採用環柵電晶體的節點遷移。如果你還記得的話，我們之前說過，全面採用閘極設計相當於每月每 10 萬片晶圓產能增加約 10 億美元的 Lam SAM 增量成本。鑑於閘極環繞結構的三維特性，憑藉我們的沉積和蝕刻產品組合，我們有能力在這個細分市場中獲得份額。

In addition, customers are integrating more functionality with advanced packaging. We previously estimated advanced packaging would make up a mid-single-digit percentage of overall foundry-logic equipment spend. As additional devices, including those for mobile applications, adopt more complex packaging schemes, we see this number moving higher.

此外，客戶正在將更多功能整合到先進的包裝中。我們先前估計，先進封裝將佔代工邏輯設備總支出的個位數百分比。隨著包括行動應用裝置在內的更多裝置採用更複雜的封裝方案，我們看到這個數字還在不斷上升。

In high bandwidth memory, or HBM, advanced packaging is critical for the transition to HBM4 and 4E and stacking of up to 16 layers. Lam is in an excellent position given our market leadership and electroplating and TSV etch. We expect our overall advanced packaging business to grow more than 40% in 2026, outperforming our view of WFE growth in this space.

在高頻寬記憶體（HBM）中，先進的封裝技術對於向 HBM4 和 4E 過渡以及最多 16 層的堆疊至關重要。憑藉市場領導地位以及電鍍和TSV蝕刻技術，Lam公司處於非常有利的地位。我們預計到 2026 年，我們整體的先進封裝業務將成長超過 40%，超過我們對 WFE 在該領域成長的預期。

Finally, in NAND, demand is growing faster than previously expected as new use cases for high-capacity SSDs emerge. Non-volatile context memory layers to enable large-scale AI inference have the potential to add incremental growth in NAND bit demand. For every 2 million to 3 million accelerators sold, we estimate an incremental 1 point increase in overall NAND bit demand growth. Lam has the industry's largest installed base of NAND systems and is well positioned to outperform as the NAND market inflects higher. Against this backdrop of strong semiconductor demand and accelerated technology transitions, we are seeing increased momentum for our newly launched products.

最後，隨著高容量固態硬碟新應用場景的出現，NAND快閃記憶體的需求成長速度超過了先前的預期。用於實現大規模 AI 推理的非揮發性上下文儲存層有可能增加 NAND 位元需求。我們估計，每售出 200 萬至 300 萬個加速器，NAND 快閃記憶體的整體位需求成長就會增加 1 個百分點。Lam 擁有業界最大的 NAND 系統裝置容量，並且隨著 NAND 市場走高，它已做好充分準備，有望取得優異成績。在半導體需求強勁和技術轉型加速的背景下，我們看到新推出的產品動能日益強勁。

Akara, our latest generation conductor etch system, has doubled its installed base over the past year, with production tool of record wins for EUV and high aspect ratio etch applications in advanced DRAM and foundry-logic. Critical dimensions in foundry-logic are shrinking by roughly 10% to 20% node to node. Similarly, in DRAM, aspect ratios are increasing with each node and process complexity is set to grow even more with future moves to 4F2 and 3D DRAM.

Akara 是我們最新一代的導體蝕刻系統，在過去一年中，其安裝量翻了一番，並在先進的 DRAM 和代工邏輯裝置的 EUV 和高縱橫比蝕刻應用領域贏得了創紀錄的生產工具訂單。晶圓代工邏輯的關鍵尺寸正在逐節點縮小約 10% 到 20%。同樣，在 DRAM 中，縱橫比隨著每個節點的推進而增加，隨著未來向 4F2 和 3D DRAM 的過渡，工藝複雜度必將進一步增加。

Consequently, multiple customers have chosen Akara for its unmatched ability to etch the smallest dimensions at very high aspect ratios while maintaining profile control and reducing variability across the wafer. This is achieved through new innovations, including our DirectDrive solid-state power delivery hardware and TEMPO plasma pulsing. In next-generation gate-all-around devices, we expect the number of applications using Akara to grow by roughly 2 times, including wins for critical front-end silicon etch applications.

因此，許多客戶選擇 Akara，因為它能夠在保持輪廓控制和減少晶圓上的變化的同時，以極高的縱橫比蝕刻最小尺寸，具有無與倫比的能力。這是透過新的創新實現的，包括我們的 DirectDrive 固態電源傳輸硬體和 TEMPO 等離子脈衝技術。在新一代全環柵元件中，我們預計使用 Akara 的應用數量將成長約 2 倍，其中包括關鍵的前端矽蝕刻應用。

In DRAM, we already have wins with Akara for the 1C node that are set to ramp this year and expect growing momentum as the applications using Akara expand nearly 3 times in a subsequent 1D node. As we look out over a multi-year period, the unprecedented AI ramp demands greater speed and agility across the ecosystem. Our customers are moving faster at every stage of process development and manufacturing, so we have increased the velocity of our execution across the board.

在 DRAM 領域，我們已經憑藉 Akara 贏得了 1C 節點的訂單，這些訂單將於今年開始量產，並且隨著使用 Akara 的應用在後續的 1D 節點中增長近 3 倍，我們預計增長勢頭將更加強勁。展望未來幾年，前所未有的 AI 發展速度要求整個生態系統具備更高的速度和靈活性。我們的客戶在流程開發和製造的各個階段都在加快步伐，因此我們也全面提高了執行速度。

We are strengthening our supply base, automating logistics and ramping high-volume manufacturing. Over the last four years, we have nearly doubled our overall manufacturing capacity. And in 2025, we launched state-of-the-art automated warehouses that enable greater production efficiency. These investments have proved critical in a fast-ramping market environment, and we're set to expand our footprint further to meet the demand we see over the next several years.

我們正在加強供應鏈，實現物流自動化，並提高大規模生產能力。過去四年，我們的整體生產能力幾乎翻了一番。2025年，我們推出了最先進的自動化倉庫，從而提高了生產效率。事實證明，這些投資在快速成長的市場環境中至關重要，我們將進一步擴大業務範圍，以滿足未來幾年我們看到的需求。

Our product sales and support teams are also executing to the accelerated pace of customer demand. Over the course of 2025, Lam was recognized with nearly 40 supplier awards, highlighting in many cases our fast tool installations and outstanding production ramp support. Looking forward, we see Lam's Equipment Intelligence solutions and Dextro cobots leading the way to the autonomous fab with predictive and automated maintenance and precise global fleet matching. Dextro continued to gain momentum in 2025, expanding to cover six different Lam tool types.

我們的產品銷售和支援團隊也以加快的速度回應客戶需求。在 2025 年期間，Lam 獲得了近 40 項供應商獎項，其中許多都突顯了我們快速的工具安裝和出色的生產爬坡支援。展望未來，我們看到 Lam 的設備智慧解決方案和 Dextro 協作機器人將引領自主晶圓廠的發展，實現預測性和自動化維護以及精確的全球車隊匹配。2025年，Dextro持續保持成長勢頭，業務範圍擴大到涵蓋六種不同的層壓工具類型。

And finally, in an environment where inflections are more complex and innovation timelines are compressing, we have transformed our R&D capabilities to help us stay ahead. We are utilizing velocity labs located close to our customers to screen new materials, new hardware, and new process regimes at a rate not previously possible. We are also leveraging Lam's digital twins capabilities to shorten product development cycles and converge on next-generation tool and process solutions with greater efficiency.

最後，在情況更加複雜、創新時間線不斷縮短的環境下，我們轉變了研發能力，以幫助我們保持領先地位。我們正在利用位於客戶附近的 Velocity Labs，以前所未有的速度篩選新材料、新硬體和新工藝方案。我們也利用 Lam 的數位孿生技術來縮短產品開發週期，並以更高的效率實現下一代工具和製程解決方案。

Wrapping up, the growth we envisioned for Lam at our investor event one year ago is materializing faster than we anticipated. We are making progress against our SAM expansion, share gain, and profitability objectives. And with the demand environment continuing to accelerate, we are elevating our focus on scaling the company, delivering for customers, and outperforming in the AI era.

總結起來，我們在一年前的投資者活動上為 Lam 設想的成長目標，正在以比我們預期更快的速度實現。我們在實現 SAM 擴張、市佔率提升和獲利目標方面取得了進展。隨著需求環境持續加速成長，我們將更重視擴大公司規模、為客戶提供服務，並在人工智慧時代取得優異成績。

Thank you, and here's Doug.

謝謝，這位是道格。

Great. Thank you, Tim. Good afternoon, everyone. Thank you for joining our call today during what I know is a super busy earnings season. Before I get into the details, I'd like to say that we were quite pleased with the strong execution across the company in calendar year 2025, which translated into record top and bottom line financial performance.

偉大的。謝謝你，提姆。大家下午好。感謝您在今天這個財報季非常繁忙的時候參加我們的電話會議。在詳細介紹之前，我想說，我們對公司在 2025 年日曆年的出色執行力感到非常滿意，這轉化為創紀錄的營收和利潤財務表現。

In calendar year 2025, revenue was a record coming in at $20.6 billion, which was up 27% year over year. CSBG revenue also reached a record of $7.2 billion. Gross margin was 49.9%, the highest result as a combined company for the full year since the Novellus merger back in 2012. Gross profit increased 31% year over year to $10.3 billion.

2025 年，公司營收創下歷史新高，達到 206 億美元，年增 27%。社區服務補助金 (CSBG) 收入也達到了創紀錄的 72 億美元。毛利率為 49.9%，這是自 2012 年與 Novellus 合併以來，合併後公司全年毛利率最高的一次。毛利年增 31%，達到 103 億美元。

We also had record operating margin of 34.1% and operating profit dollars of $7 billion, which was up 41% year over year. Diluted earnings per share were $4.89, which was up [46%] (corrected by company after the call) year over year. Looking at it, we delivered leverage from the top to the bottom of the P&L in 2025.

我們的營業利益率也創下歷史新高，達到 34.1%，營業利益達到 70 億美元，較去年同期成長 41%。稀釋後每股收益為 4.89 美元，年增 46%（該公司在電話會議後進行了更正）。從這個角度來看，我們在 2025 年實現了從損益表頂部到底部的全面槓桿效應。

Let me turn to the December quarter results. Our revenue was above the midpoint of guidance, while gross margin, operating margin, and earnings per share all exceeded the high end of our guided range. Revenue for the December quarter was a record coming in at $5.34 billion. This marked our 10th consecutive quarter of revenue growth. The deferred revenue balance at quarter end came in at $2.25 billion, down sequentially due to an approximately $500 million reduction in those customer advanced down payments.

接下來我來看看十二月季度的業績。我們的收入高於預期的中點，而毛利率、營業利潤率和每股收益都超過了預期範圍的上限。12 月季度營收創歷史新高，達 53.4 億美元。這標誌著我們連續第十個季度實現營收成長。截至季末，遞延收入餘額為 22.5 億美元，季減，原因是客戶預付的首付減少了約 5 億美元。

From a market segment perspective, Foundry accounted for 59% of our systems revenue in the December quarter, slightly down sequentially, but up from 35% in the December 2024 period. This underscores the success of our strategic focus and execution in Foundry. Foundry's strength came from investments at the leading edge in addition to mature node spending that we saw in China.

從市場區隔的角度來看，晶圓代工業務在 12 月季度占我們系統收入的 59%，較上季略有下降，但高於 2024 年 12 月期間的 35%。這凸顯了我們在 Foundry 的策略重點和執行上的成功。Foundry 的優勢來自於對尖端技術的投資，以及我們在中國看到的成熟節點支出。

Memory was 34% of systems revenue, in line with the prior quarter. Within memory, we generated record DRAM revenue accounting for 23% of systems revenue, which was up from 16% in the September quarter. Investments in high-bandwidth memory continue to remain strong, driven by movement to HBM3E and 4. We also saw traditional node migrations to the 1B and 1C nodes, enabling the transition to DDR5.

記憶體業務收入佔系統總收入的 34%，與上一季持平。在記憶體業務方面，DRAM 收入創歷史新高，佔系統收入的 23%，高於 9 月季度的 16%。受 HBM3E 和 HBM4 的發展趨勢推動，對高頻寬記憶體的投資仍然強勁。我們也看到了傳統節點向 1B 和 1C 節點的遷移，從而實現了向 DDR5 的過渡。

Non-volatile memory contributed 11% of our systems revenue, down from 18% in September quarter. This trajectory was in line with our expectations for customer plans coming into the year.

非揮發性記憶體貢獻了我們系統營收的 11%，低於 9 月季度的 18%。這一發展軌跡與我們對年初客戶計畫的預期相符。

Despite the quarterly decline, NAND revenues grew strongly for Lam in what was the first-half-weighted calendar year 2025. As we enter 2026, we see solid end market demand as customers prepare for the next stage of AI-driven growth in NAND.

儘管季度有所下降，但 Lam 在 2025 年上半年（按加權日曆年計算）的 NAND 收入實現了強勁增長。進入 2026 年，我們看到終端市場需求強勁，因為客戶正在為 NAND 領域人工智慧驅動的下一階段成長做好準備。

And finally, the Logic and Other segment came in at 7% of systems revenue in the December quarter, slightly up sequentially.

最後，邏輯和其他業務板塊在 12 月季度佔系統收入的 7%，較上季略有成長。

Let's turn to the regional breakdown of our total revenue. China came in at 35%, which was a decrease from the prior quarter level of 43% but slightly higher than our original expectations. This was due to updates in the affiliate rule and the resulting timing of shipments from that. The next largest geographic concentrations were Taiwan coming in at 20%, up sequentially from 19%; and Korea had 20%, up sequentially from 15%.

讓我們來看看總收入的區域細分。中國佔比為 35%，低於上一季的 43%，但略高於我們最初的預期。這是由於聯盟規則的更新以及由此導致的貨物發貨時間變化所致。其次，台灣的集中度為 20%，比上一輪的 19% 上升；韓國的集中度為 20%，比上一輪的 15% 上升。

Customer Support Business Group generated approximately $2 billion in revenue for the December quarter, up 12% sequentially on an increase in Reliant systems. We were 14% higher than the same period in 2024, primarily a growth in spares. CSBG obviously remains a key part of our growth strategy with our expanding installed base and innovation and advanced services.

客戶支援業務集團在 12 月季度創造了約 20 億美元的收入，環比成長 12%，這主要得益於 Reliant 系統的成長。與 2024 年同期相比，我們成長了 14%，這主要是由於備件成長所致。隨著我們不斷擴大的用戶基礎以及創新和先進服務，CSBG 顯然仍然是我們成長策略的關鍵組成部分。

NAND spending enabled record upgrade revenue in 2025, up more than 90% year over year. In the 13 years since we brought Lam and Novellus together, I'd like to remind everybody that CSBG has grown every year except for one.

NAND快閃支出推動2025年升級營收創歷史新高，較去年同期成長超過90%。自從我們把 Lam 和 Novellus 結合在一起以來的 13 年裡，我想提醒大家，CSBG 除了有一年之外，每年都在成長。

Let's look to profitability. Gross margin in the December quarter was 49.7%, which exceeded the high end of our guided range on better-than-expected customer mix. Sequentially, gross margin was about 1 percentage point lower reflecting a customer mix that was less favorable than what we saw in September. Operating expenses for December came in at $827 million, which is roughly flat sequentially. R&D accounted for 68% of the total operating expenses.

讓我們來看看獲利能力。12 月季度的毛利率為 49.7%，由於客戶組合好於預期，超過了我們預期範圍的上限。與 9 月相比，毛利率環比下降約 1 個百分點，反映出客戶組合不如 9 月有利。12 月份的營運支出為 8.27 億美元，與上一季基本持平。研發費用佔總營運費用的 68%。

We continue investing to maintain our leadership with a differentiated product portfolio for our customers with innovations like Vantex, Akara, Halo, and Dextro. The December quarter operating margin was 34.3%, exceeding the high end of our guidance. The non-GAAP tax rate for the quarter came in at 13.2%, generally in line with our expectations.

我們將繼續投資，透過 Vantex、Akara、Halo 和 Dextro 等創新產品，為客戶提供差異化的產品組合，以維持我們的領先地位。12 月季度的營業利潤率為 34.3%，超過了我們預期的上限。本季非GAAP稅率為13.2%，基本上符合我們的預期。

We continue to see the tax rate in the low- to mid-teens for calendar 2026. Other income and expense for the December quarter was approximately $10 million in income compared with $8 million in income in the September quarter. Slight fluctuation in OI&E was primarily the result of gains in our venture portfolio, partly offset by lower interest income. As we've talked about in the past, you should expect to see variability in OI&E quarter-to-quarter.

我們預計 2026 年的稅率將繼續保持在十幾到五幾的水平。12 月季度的其他收入和支出約為 1000 萬美元，而 9 月季度的收入為 800 萬美元。OI&E 的輕微波動主要是由於我們創投組合的收益，但部分被較低的利息收入所抵銷。正如我們之前討論過的，您應該預料到營運成本和支出會逐季度出現波動。

Capital return in the December quarter, we allocated approximately $1.4 billion towards share buybacks through open market share repurchases. Our average buyback price in the quarter was approximately $154 per share.

在 12 月季度的資本回報中，我們透過公開市場股票回購，分配了約 14 億美元用於股票回購。本季我們的平均回購價格約為每股 154 美元。

In calendar year 2025, we repurchased approximately 39 million shares at an average price of $104 per share. We also paid $328 million in dividends in the quarter. In calendar year 2025, we returned 85% of our free cash flow. Our plans remain to return at least 85% of free cash flow to our shareholders over time.

2025 年，我們以每股 104 美元的平均價格回購了約 3,900 萬股股票。本季我們還支付了 3.28 億美元的股息。2025 年，我們返還了 85% 的自由現金流。我們的計劃仍然是隨著時間的推移，將至少 85% 的自由現金流返還給股東。

The December quarter diluted earnings per share were $1.27, which came in above the guidance range. The diluted share count was 1.26 billion shares, which was a reduction from the September quarter and consistent with our guidance. We have $5.1 billion remaining on our board authorized share repurchase plan.

12 月季度稀釋後每股收益為 1.27 美元，高於預期範圍。稀釋後的股份數量為 12.6 億股，較 9 月季度有所減少，與我們的預期一致。我們董事會批准的股票回購計畫中還剩下 51 億美元。

Let me pivot to the balance sheet. Cash and cash equivalents totaled $6.2 billion at the end of the December quarter, a decrease from $6.7 billion at the end of the September quarter. The reduction in cash is attributed to capital return as well as CapEx spending. As we look ahead, our strong cash position and continued free cash flow enable us flexibility to potentially simply repay the $750 million March 2026 notes when they mature.

讓我們把話題轉到資產負債表上來。截至 12 月季末，現金及現金等價物總額為 62 億美元，較 9 月季末的 67 億美元減少。現金減少的原因包括資本回報和資本支出。展望未來，我們強勁的現金狀況和持續的自由現金流使我們能夠靈活地在 2026 年 3 月到期時直接償還 7.5 億美元的票據。

Days sales outstanding was 59 days in the December quarter, a decrease from 62 days in the September quarter. Inventory turns improved to 2.7 times from 2.6 times in the prior quarter and up from 1.5 times a little over two years ago.

12 月季度應收帳款週轉天數為 59 天，較 9 月季度的 62 天有所下降。庫存週轉率從上一季的 2.6 次提高到 2.7 次，也高於兩年多前的 1.5 次。

As a company, we remain focused on asset utilization, and we were pleased by the sustained progress we continue to make. Our non-cash expenses in the December quarter included approximately $89 million for equity compensation, $91 million for depreciation, and $13 million in amortization. Capital expenditures for the December quarter was $261 million, which was up $76 million from the September quarter. Spending was driven by investments in manufacturing capacity, R&D and lab infrastructure that supports our technology road map and customer needs.

作為一家公司，我們始終專注於資產利用率，並且對我們持續取得的進展感到滿意。我們在 12 月季度的非現金支出包括約 8,900 萬美元的股權激勵、9,100 萬美元的折舊和 1,300 萬美元的攤銷。12 月季度的資本支出為 2.61 億美元，比 9 月季度增加了 7,600 萬美元。支出主要用於投資製造能力、研發和實驗室基礎設施，以支援我們的技術路線圖和客戶需求。

We also purchased a new building in Arizona to support the growing industry footprint there. This capital spending remains consistent with our global strategy of expanding capabilities close to where our customers are.

為了支持該地區不斷增長的產業規模，我們也在亞利桑那州購買了一棟新建築。這項資本支出與我們擴大服務能力、貼近客戶所在地的全球策略保持一致。

Looking forward, we continue to expect capital expenditure to be in the 4% to 5% of revenue range. We ended the December quarter with approximately 19,700 regular full-time employees, which was an increase of approximately 300 people from the prior quarter. Headcount increases were primarily within the field organization to support customer growth as well as in R&D to support our long-term product roadmap.

展望未來，我們預期資本支出仍將佔收入的 4% 至 5%。截至 12 月季度末，我們擁有約 19,700 名正式全職員工，比上一季增加了約 300 人。人員增加主要集中在現場組織，以支援客戶成長，以及研發部門，以支援我們的長期產品路線圖。

Let's turn to our non-GAAP guidance for the March 2026 quarter. We're expecting revenue of $5.7 billion, plus or minus $300 million. We're expecting gross margin of 49%, plus or minus 1 percentage point. We're expecting to see slight headwinds from customer mix. We're forecasting operating margins of 34%, plus or minus 1 percentage point. You'll see the normal seasonal uptick in operating expenses in the March quarter. And finally, we're forecasting earnings per share of $1.35, plus or minus $0.10, based on a share count of approximately 1.26 billion shares.

讓我們來看看我們對 2026 年 3 月季度的非 GAAP 業績指引。我們預計營收為 57 億美元，上下浮動 3 億美元。我們預期毛利率為 49%，上下浮動 1 個百分點。我們預期客戶結構的變化會帶來一些不利影響。我們預測營業利益率為 34%，上下浮動 1 個百分點。三月季度，營運費用會出現正常的季節性成長。最後，我們根據約 12.6 億股的股票數量，預測每股收益為 1.35 美元，上下浮動 0.10 美元。

So let me wrap up. We delivered a record year in 2025, reflecting strong execution and broad-based strength across our product portfolio. As we look into 2026, we expect meaningful year-over-year growth supported by sustained demand in AI-driven markets and continued investment in capacity. We agree with the prevailing view that much of the market will be undersupplied in 2026 due to clean room space constraints. In line with that, we see 2026 as a second-half-weighted year.

那麼，我就總結一下吧。2025 年，我們取得了創紀錄的業績，這反映了我們強大的執行力和產品組合的廣泛實力。展望 2026 年，我們預計在人工智慧驅動市場的持續需求和產能的持續投資的支持下，將實現顯著的同比增長。我們同意目前普遍的觀點，即由於無塵室空間限制，到 2026 年，市場將出現供應不足的情況。基於此，我們認為 2026 年是下半年表現較為突出的年份。

With our strong balance sheet, an expanding installed base and the strength of our product portfolio, we remain confident in Lam's ability to continue to outperform and deliver long-term value for our customers and shareholders.

憑藉我們強勁的資產負債表、不斷擴大的裝機量和強大的產品組合，我們仍然對 Lam 能夠繼續取得優異業績並為我們的客戶和股東創造長期價值充滿信心。

Operator, that concludes our prepared remarks. Tim and I would now like to open up the call for questions.

操作員，我們的發言到此結束。提姆和我現在想開始接受提問。

(Operator Instructions) Tim Arcuri, UBS.

（操作員說明） Tim Arcuri，UBS。

Doug, I had a question about WFE this year. So you said we're going to be constrained because of this fab readiness. Is it possible to say how much? I know you're guiding WFE to $135 billion this year. I mean, if we use semiconductor revenue and you assume sort of a normal WFE intensity number, it seems like you can get to like $150 billion. So maybe the constraints are costing the industry like $15 billion.

道格，我有個關於今年WFE的問題。所以你說我們會因為晶圓廠的準備工作而受到限制。能否說明具體金額？我知道你正帶領WFE今年實現1,350億美元的營收。我的意思是，如果我們以半導體收入為基準，並假設WFE強度處於正常水平，那麼似乎可以達到1500億美元左右。所以這些限制可能給該產業造成了 150 億美元的損失。

Is it possible to give a number in terms of how much the constraints are kind of costing in terms of WFE this year so we can kind of pro forma that out?

能否給出一個數字，說明今年這些限制對 WFE 造成的成本影響，以便我們進行預估？

Yeah, Tim. I knew somebody was going to ask that question. I should have anticipated it was going to be you. Listen, it's hard for us to put a number on it, and I'm going to decline to do that as we sit here right now. And the reason Tim is plans are somewhat fluid if we're honest, meaning people are trying to figure out how to get a little more cleanroom space, so to bring facilities online and bump things up a little bit. So I'm not -- we're not going to put a number on it.

是的，提姆。我知道一定會有人問這個問題。我早該想到會是你。聽著，我們很難給它下一個具體的數字，而且我現在也不打算這麼做。坦白說，提姆的計畫還有一些變數，這意味著人們正在想辦法獲得更多的無塵室空間，以便讓設施上線並稍微加快進度。所以，我們不會給它設定一個數字。

But I think it's safe to say, and Tim can comment on this as well. I think it sets up for '27 to also be a pretty good year as we think through this. I mean, the industry seems to be sold out for most of what it is supplying, everybody is talking about these multi-year agreements that they're working on. And I think that's largely a reflection of the fact that demand is very strong, and there's just not enough cleanroom out there.

但我認為可以肯定地說，蒂姆也可以就此發表評論。我認為，考慮到這一點，2027 年也可能是個相當不錯的年份。我的意思是，這個行業的大部分產品似乎都已售罄，每個人都在談論他們正在洽談的這些多年期協議。我認為這很大程度上反映了市場需求非常強勁，而無塵室供應不足的事實。

Yeah. I don't have much to add, Tim, other than to say that clearly, you've seen a large number of fab announcements. I mean, those fab announcements are capacity in '27, '28, and beyond. And so I think there's a view that the constraints this year are going to continue even out until many of those new fabs open up. And so I think we've given you a view of what we think WFE is, as Doug said, we're working on productivity improvements to get a little extra output for customers. That's how we support them. But fundamentally, it's a pretty big challenge.

是的。提姆，我沒什麼要補充的，只想說，很明顯，你已經看到了很多精彩的公告。我的意思是，這些驚人的消息意味著 2027 年、2028 年及以後的產能。因此，我認為有一種觀點認為，今年的限制措施將會持續下去，直到許多新的晶圓廠投入運作。所以，我想我們已經向大家展示了我們對 WFE 的看法，正如 Doug 所說，我們正在努力提高生產效率，以便為客戶帶來更多產出。這就是我們支持他們的方式。但從根本上講，這是一個相當大的挑戰。

Thanks a lot. Doug, and then you're guiding gross margin down a bit on up revenue. It sounds like it's predominantly related to China. So China was 35% in December.

多謝。道格，這樣一來，你就能在營收成長的同時，引導毛利率略微下降。聽起來主要與中國有關。所以中國12月的佔比是35%。

Is it going to come down? Like is that the reason why the gross margin is coming down? And if so, like what mixes in China will be for March?

它會倒塌嗎？這就是毛利率下降的原因嗎？如果真是這樣，那麼中國三月的配比會是什麼呢？

Yeah. I'm not going to give you a hard number, Tim. But yes, it's customer mix. It's going to be less rich in the March quarter. And I'll also remind everybody that this isn't a fixed cost business for us. So as volume goes up and down, the component that just benefits from revenue growing isn't that big. The mix component of both products as well as customer is an important item. So you're latched on to the right thing, Tim.

是的。提姆，我不會給你一個確切的數字。沒錯，這確實是客戶群構成的問題。三月份的季度收益會減少。我還要提醒大家，對我們來說，這不是固定成本企業。因此，隨著銷售量的波動，僅僅受惠於營收成長的部分並不大。產品和客戶的組合因素都非常重要。所以你找對了方向，提姆。

CJ Muse, Cantor Fitzgerald.

CJ Muse，坎托·菲茨杰拉德。

I guess to follow up on that last question, Doug, could you speak a bit about the work you're doing with the supply chain, bringing on manufacturing, ramping Malaysia, and how we should think about that in the context of gross margins as revenues ramp in the second half of calendar '26 and beyond?

我想接著上一個問題，Doug，你能談談你在供應鏈方面所做的工作，比如引入製造業，擴大馬來西亞的產能，以及在2026年下半年及以後隨著收入增長，我們應該如何看待這些工作對毛利率的影響嗎？

Yeah, CJ. I mean, we've been talking for a while. I think, about CapEx growing as a result of expanding manufacturing capability. Tim specifically talked about a doubling over the last four, five years. So it's been an item that we've been clearly very focused on.

是的，CJ。我的意思是，我們已經談了一段時間了。我認為，資本支出成長是由於製造能力的擴大所致。提姆特別提到，過去四、五年間，這個數字翻了一番。所以，這顯然是我們一直非常關注的一個面向。

We're ramping globally, and you're right, that Malaysia location is our biggest location as we sit here today, and we're trying to get more out of that in addition to everywhere that we are. But the mix component, CJ. is going to be more important than anything, at least in the near term, less than just volume ramping.

我們正在全球範圍內擴大規模，你說得對，馬來西亞是我們目前最大的分支機構，除了我們在其他所有地方的業務之外，我們還在努力從馬來西亞獲得更多收益。但混合成分，CJ，至少在短期內，比起單純提高銷量，更重要。

And then maybe a follow-up question on CSBG, I would think your customers are trying to get every bit that they can out. And so the uplift that you saw in the December quarter, and I assume continued strength into March. Is that something that will sustain throughout the whole calendar year and should drive better than kind of that 12% growth CAGR? Or does that take a pause at some point as the transition focuses on greenfield investments?

然後，關於 CSBG，我可能還有一個後續問題，我認為你們的客戶正在努力爭取他們能得到的每一分錢。因此，您在 12 月季度看到了成長，而且我預計這種強勁勢頭會延續到 3 月。這種趨勢能否持續一整年，並能比12%的複合年增長率帶來更好的成長？或者，隨著轉型重點轉向綠地投資，這種情況是否會在某個階段暫停？

Yeah. CJ, I'll let Doug speak to the specific numbers. But I think that what you should keep in mind is the CSBG, a lot of our growth, of course, is driven by customers' near-term actions and what we need to do to help them get maximum amount of output from the tools they have. But a lot of the growth really is we're transforming our service business to be much more oriented towards the use of equipment intelligence for predictive maintenance as a way of getting more output from tools and also the implementation of Dextro cobots for our automated maintenance.

是的。CJ，具體數字就讓Doug來解釋。但我認為你應該記住的是，CSBG，當然，我們的許多成長都是由客戶的近期行動以及我們需要做什麼來幫助他們從現有工具中獲得最大產出所驅動的。但實際上，我們許多成長都來自於我們正在將服務業務轉型，使其更加重視利用設備智慧進行預測性維護，從而從工具中獲得更高的產出，以及實施 Dextro 協作機器人進行自動化維護。

And both of those things not only will drive top-line growth but also margin profitability improvement just given the efficiency with which we deliver those services. And so I think there's a number of moving pieces that are all positive for CSBG.

這兩件事不僅會推動營收成長，而且鑑於我們提供這些服務的效率，還能提高利潤率。所以我認為有很多因素對CSBG來說是正面的。

Yeah. I would just add, CJ. We were pleased to see the chamber count number up. Obviously, we knew that was coming, and we're happy to share it with you. So that's one aspect of what we're continue to take advantage of going forward is very consistent, though, with what we articulated at that Investor Day back, I guess, almost a year ago.

是的。我還要補充一點，CJ。我們很高興看到密室數量增加。顯然，我們知道這一天終會到來，我們很樂意與你們分享。所以，我們將繼續利用的其中一個方面，與我們大約一年前在投資者日上闡述的內容非常一致。

And so I would still want you thinking about CSBG growing the same way we described it back then, which is a high single digit, maybe low double digit. We had a very strong December quarter. It was primarily a result of Reliant systems. That piece might be a little bit lumpy; it always is. But again, CSBG is just going to keep chugging along.

因此，我仍然希望你們按照我們當時所描述的方式看待 CSBG 的成長，即接近兩位數，甚至可能接近兩位數。我們十二月季度的業績非常強勁。這主要是 Reliant 系統造成的。那塊木頭可能有點凹凸不平；它總是這樣。但是，CSBG 還是會繼續穩定前進。

Atif Malik, Citi.

阿提夫‧馬利克，花旗銀行。

The first one for Tim. Tim, on the DRAM market, when do you see the volume adoption of 4F2 from 6F4? And can you talk about your SAM market share when you move to 4F2? I know you called out Akara in your prepared remarks.

這是給提姆的第一份。Tim，關於DRAM市場，你認為4F2何時會從6F4大規模普及？搬到 4F2 後，您能否談談您的 SAM 市場佔有率？我知道你在事先準備好的發言稿中點名批評了阿卡拉。

Yeah. So I think 4F2, I mean, obviously, there's still some question as to exact timing and customers have talked about it being something towards the end of the decade, probably full volume production. But clearly, we're engaged today with customers looking at the technical needs.

是的。所以我覺得 4F2，我的意思是，顯然，確切的時間安排還有一些疑問，客戶們也談到過，大概會在十年末期全面投產。但很顯然，我們今天正在與客戶溝通，了解他們的技術需求。

And we called out Akara. Akara is very well suited to the types of high aspect ratio, very small features that exist in 4F2 as well as other devices we've talked about, whether it's future at all around or even a foundry-logic nodes to CFET. And so Akara is -- it's sort of a foundational tool for us in terms of capabilities that are going to be important for all of these transitions.

我們點名批評了阿卡拉。Akara 非常適合 4F2 以及我們討論過的其他裝置中存在的高縱橫比、非常小的特徵，無論是面向未來的所有裝置，還是面向 CFET 的代工邏輯節點。因此，Akara 對我們來說是一個基礎工具，它所擁有的能力對於所有這些轉型都至關重要。

But you really should think about those kinds of technology transitions occurring after -- probably after this next big wave of fab openings. But again, back to the constraint question, in some ways, if this turns out to be as we believe a multi-year build-out of fabs, the fabs that come towards the end of that will be the fabs to benefit from the SAM expansion and share gain that we're going to see coming from these new products that we've talked about. So Anyway, I think Four F-Squared is probably on the back end of that, but there's a lot in between that will also drive our business.

但你真的應該考慮一下這類技術變革會在——可能是在下一波晶圓廠大規模投產之後——發生。但是，回到限制問題上來，在某種程度上，如果這最終像我們所認為的那樣是一個需要多年才能建成的晶圓廠，那麼在建設後期建成的晶圓廠將受益於我們即將看到的來自這些新產品的 SAM 擴張和市場份額增長。總而言之，我認為 Four F-Squared 可能處於這個過程的後期階段，但中間還有很多因素也會推動我們的業務發展。

Great. And then Doug on the NAND market. I know the memory dynamics were in line with your expectations in the December quarter, but NAND was down sequentially and DRAM up. Do you see the NAND makers slowing down technology migrations as they focus more on in terms of like minting money given how the supply shortages are materializing? And when do you see NAND new capacity additions coming on?

偉大的。然後是道格談NAND市集。我知道12月份的記憶體市場動態符合您的預期，但NAND快閃記憶體環比下降，DRAM快閃記憶體環比上升。鑑於供應短缺的情況日益嚴重，您是否認為 NAND 快閃記憶體製造商會放慢技術遷移的速度，轉而更加專注於創造利潤？您預計何時會推出新的NAND快閃記憶體容量？

Yeah, Atif. No, listen, NAND played out exactly as we saw it as the year began in 2025. And as we sit here looking at 2026, it will be a growth year for NAND. There's no question about that in our minds.

是的，阿提夫。不，聽著，NAND 的發展正如我們在 2025 年初所看到的那樣。展望 2026 年，這將是 NAND 快閃記憶體的成長之年。這一點我們毫不懷疑。

I think what I observed the memory customers doing at least to the extent that they have both NAND and DRAM right now anyway, is prioritizing DRAM over NAND a little bit because profitability there is somewhat better. I think you all understand that and know that. But that doesn't mean that people aren't focused on NAND.

我認為我觀察到的記憶體客戶（至少在他們目前同時擁有 NAND 和 DRAM 的情況下）正在優先考慮 DRAM 而不是 NAND，因為 DRAM 的盈利能力要好一些。我想你們都明白這一點，也都知道這一點。但這並不意味著人們不關注 NAND 快閃記憶體。

In fact, one of our largest customers announced a new fab that's going to be dedicated pretty well, not dedicated but heavily emphasizing NAND capacity. And so that's on the come line, we see that happening as we get into '26 growth that is happening. We're still sticking by. We think upgrades happen before real capacity additions. But you're going to see a combination of both and that $40 billion that we've been talking about, likely happens quicker than what we originally expected a year ago. So anyway, we feel quite good about where NAND is trending. Thanks, Atif.

事實上，我們最大的客戶之一宣布新建一座晶圓廠，該晶圓廠將主要用於（或非常專門）生產 NAND 閃存，但並非完全專用。所以，在即將到來的時期，我們看到，隨著 2026 年的增長，這種情況正在發生。我們依然堅守崗位。我們認為升級改造先於真正的產能提升。但你會看到兩者的結合，我們一直在談論的 400 億美元，可能會比我們一年前最初預期的更快實現。總之，我們對NAND的發展趨勢感到非常樂觀。謝謝你，阿提夫。

Vivek Arya, Bank of America Securities.

Vivek Arya，美國銀行證券。

So you're guiding WFE up 23%. I think last year, you said you gained a point of share. Do you expect to maintain or gain share this year, Tim? What are kind of the puts and takes around the different markets? And then specifically, what are you assuming for China contribution overall for WFE and what that means for Lam in calendar '26?

所以你預計 WFE 將上漲 23%。我記得去年你說過你們的市佔率增加了一個百分點。提姆，你預計今年市佔率會保持穩定還是有所成長？不同市場有哪些買賣交易？那麼具體來說，您認為中國對世界經濟論壇的整體貢獻是多少？這對 2026 年的林氏政治格局意味著什麼？

Yeah. Let me take the first part of it. I mean, I think what's important to remember from the longer-term plan we laid out at last year's Investor Day and things I said on the call today, we expect to gain share and expand our SAM with every successive technology node. And so to your point, do we plan to sustain or increase share? The answer is we plan to increase our share of WFE again this year.

是的。讓我來回答第一部分。我的意思是，我認為需要記住的是，從我們去年投資者日制定的長期計劃以及我今天在電話會議上所說的內容來看，我們期望隨著每一個技術節點的進步，都能獲得市場份額並擴大我們的銷售資產規模。所以，正如您所說，我們計劃維持還是提高市場佔有率？答案是，我們計劃今年再次增加在WFE的份額。

And what needs to take place is technology transitions need to keep occurring. And what we're seeing in the environment today is those are accelerating. That's a way for customers to get more output and more output of the types of devices that are strongly demanded by the AI environment. At the same time, those technology transitions are driving higher deposition and etch intensity which is pretty much our entire business. And so from that perspective, that's a real positive for us.

而我們需要做的，就是不斷進行技術轉型。而我們今天在環境中看到的現像是，這些現象正在加速發生。這樣一來，客戶就能獲得更多產出，以及更多人工智慧環境急需的設備類型。同時，這些技術變革正在推動更高的沉積和蝕刻強度，而這幾乎就是我們整個業務的核心。所以從這個角度來看，這對我們來說確實是個好消息。

And then we've talked about the success of our new products. I mean, we've refreshed our conductor etch product line. We previously refreshed our dielectric etch line. We've launched moly. Dry resist is gaining traction. We're strong in advanced packaging.

然後我們討論了新產品的成功情況。我的意思是，我們已經更新了導體蝕刻產品線。我們之前對介質蝕刻生產線進行了升級改造。我們已經推出了鉬。乾式防染技術正逐漸受到重視。我們在先進包裝領域實力雄厚。

Backside power is still to come, and it's going to be a driver for us. And so I think we have confidence that whatever the WFE is, if it's technology-driven as it looks like it will be, we will continue to expand SAM share of the WFE.

後場力量還有待提升，這將是我們前進的動力。因此，我認為我們有信心，無論 WFE 是什麼，如果它像現在看來那樣是由技術驅動的，我們將繼續擴大 SAM 在 WFE 中的份額。

Now as far as China goes, I think that we are looking at China being more kind of flattish year-on-year. And therefore, as the rest of the technology-driven part of the business grows, becoming a smaller percentage of our overall revenue.

至於中國，我認為中國經濟的年增速會比較平穩。因此，隨著公司其他技術驅動型業務的成長，其在我們總收入中所佔的比例越來越小。

And for my follow-up, I think in the past, you have given this $40 billion or so addressable opportunity to upgrade the installed base to higher layer accounts. I'm curious where are we now versus that $40 billion number? How much more to go?

至於我的後續問題，我認為過去你們曾給予這大約 400 億美元的潛在機會，將已安裝的用戶群升級到更高級別的帳戶。我很好奇我們現在的金額與400億美元的數字相比如何？還有多遠？

And given this emerging role of NAND or this enhanced role of NAND, I should say, in AI inference, is there a new number versus that $40 billion number that you had before?

鑑於 NAND 在人工智慧推理領域日益重要的作用，或者更確切地說，是 NAND 增強的作用，與之前提到的 400 億美元相比，現在是否有一個新的數字？

Yeah. I think we might wait until a little later in the year to refresh that number. But we've said a few times now, the specific wording we used at Investor Day was $40 billion over several years. We've now, I think, on almost every earnings call said that seems to be happening faster than expected. And today, I reiterated that, which was NAND is moving faster than we expected on the upgrade path. And I think we'll come out and look.

是的。我認為我們可能會等到今年晚些時候再更新這個數字。但我們已經多次說過，我們在投資者日上使用的具體措辭是幾年內投資 400 億美元。我認為，我們幾乎在每一次財報電話會議上都說過，這種情況似乎比預期發生得更快。今天，我重申了這一點，那就是 NAND 在升級道路上的發展速度比我們預期的要快。我想我們會出來看看。

And as Doug just mentioned, we're starting to see more interest in investing in NAND capacity, but it trades off. I mean, when you have clean room space, everybody has to make a decision as to where to use that today. But I think that as we move forward and we see the growth from AI inference and other use cases, NAND is going to take its place in the AI data infrastructure and memory infrastructure, and I think you'll see growth there. And so we're just executing to the customer demand today faster than we had previously expected, and anticipate more to come.

正如 Doug 剛才提到的，我們開始看到人們對投資 NAND 容量的興趣日益濃厚，但這需要權衡取捨。我的意思是，當你擁有無塵室空間時，每個人都必須決定今天該如何利用它。但我認為，隨著我們不斷向前發展，人工智慧推理和其他應用場景的成長，NAND 將在人工智慧資料基礎設施和記憶體基礎設施中佔據一席之地，我認為你會看到這方面的成長。因此，我們今天滿足客戶需求的速度比之前預期的要快，而且預計未來還會有更多需求成長。

Srini Pajjuri, RBC Capital.

Srini Pajjuri，RBC Capital。

I want to go back to the previous question. The 1 point of WFE share that you gained, maybe if you could help us understand if it is coming primarily in foundry and logic or if you're seeing that across the board. Because foundry and logic is where you, I think, you made the most progress in the last couple of years.

我想回到之前的問題。您獲得的 1 個百分點的 WFE 份額，能否請您幫助我們了解一下，這主要是來自晶圓代工和邏輯電路，還是普遍存在於所有領域？因為我認為，在過去幾年裡，你在程式碼建置和邏輯方面取得了最大的進展。

And then Doug said, you expect year-on-year growth to be meaningful this year. Just given your WFE expectation for 22% growth, I guess, should we model 22%-plus growth on the top line for the year?

然後道格說，你預計今年的年成長會很顯著。鑑於您對 WFE 22% 的成長預期，我想，我們是否應該將全年營收成長 22% 以上作為模式？

Yes. So let me take the first one. The share gains came from a combination of both NAND and foundry-logic. And again, it's -- you might think already we have very high share in NAND, but as technology transition occurs and layer count increases, we have an opportunity still to gain share of some of the new applications required to enable those higher layer counts. And so we gained share in NAND.

是的。那我先來回答第一個問題。市佔率的成長來自於NAND快閃記憶體和代工邏輯晶片的共同作用。再說一遍，你可能認為我們在 NAND 領域已經佔據了很高的份額，但是隨著技術轉型和層數的增加，我們仍然有機會在一些實現更高層數所需的新應用領域獲得份額。因此，我們獲得了NAND​​快閃記憶體的市場份額。

But a lot of our focus we talked about over the last number of years has been to launch products that allow us to gain share at the gate all-around nodes, more advanced foundry-logic and the transitions that are coming there. And also in advanced DRAM, we saw this year some of those foundry-logic share gains coming through in the numbers that you can see. So I'd say those primarily NAND and foundry-logic this year.

但過去幾年我們一直強調的重點是推出能夠讓我們在全能節點、更先進的代工邏輯以及即將到來的轉型中獲得市場份額的產品。此外，在先進的DRAM領域，我們今年也看到了代工邏輯市場份額的一些成長，這些成長體現在您可以看到的數據中。所以我認為今年主要集中在NAND快閃記憶體和代工邏輯晶片上。

And then -- sorry, the second question, can you just repeat the second question? What was the second question?

然後——抱歉，第二個問題，您能再說一次第二個問題嗎？第二個問題是什麼？

Yeah. So I guess my second question was about your expectation for the current year. I know you said. It's going to be second half weighted, yeah.

是的。所以我想我的第二個問題是關於您對今年的期望。我知道你說過。是的，下半年的權重會更高。

Yeah. No, no. Your comment was basically will we outperform the WFE that we just talked about. I guess that's the message to try and deliver us. We're going to expand SAM gaining share, and we're going to outperform WFE this year as our current view.

是的。不，不。你的評論基本上是：我們能否超越我們剛才討論的WFE？我想這就是他們想要傳達給我們的訊息。我們將擴大 SAM 的市場份額，並且根據我們目前的預期，今年我們的業績將超過 WFE。

Okay. Got it. And then one quickly on the Op margin. Doug, I think at the Analyst Day, you gave us the guidance for 34% to 35% at roughly $25 billion to $27 billion. And you're already there. I think you're around $23 billion run rate if I look at your March guidance.

好的。知道了。然後迅速在操作利潤率上再加一筆。道格，我想在分析師日上，你給我們的指導方針是成長 34% 到 35%，營收大約 250 億美元到 270 億美元。你已經到達那裡了。根據你們三月的業績指引，我認為你們的年化營收大約在 230 億美元左右。

So I guess my question is, as we go through the next few quarters, how should we think about the Op margin fall through? I know you're guiding OpEx a little bit of growth here, but I just want to understand how you should model OpEx going forward.

所以我想問的是，在接下來的幾個季度裡，我們該如何看待營業利潤率下降的問題？我知道你正在引導營運支出（OpEx）實現一些成長，但我只是想了解你未來應該如何對營運支出進行建模。

Yeah. No, we're pretty pleased with how we've performed. Clearly, we're ahead of the model, right? I mean, that model had -- '28 kind of model. And we're run rating at least on a percentage basis, what the model suggests that we're going to be able to do.

是的。不，我們對自己的表現相當滿意。顯然，我們領先於模型，對吧？我的意思是，那個型號是——28 年那種型號。至少從百分比來看，我們的運行評分與模型預測的結果一致。

Ram and I and Tim were debating a little bit. We probably later in the year need to come out and give you an update on that model. And I think we'll do that. Lots have changed in the last year or so. So stay tuned for that.

Ram、我和Tim稍微討論了一下。我們可能在今年稍後需要向大家報告這個模型的最新進展。我想我們會這麼做的。過去一年左右，發生了許多變化。敬請期待。

I think as we think about this year, frankly, this is a management team that prides itself on being able to deliver leverage through to the bottom line. We really did a great job with it last year which is why I went through all the kind of demonstration of what we did last year. We will be focused on delivering leverage as we go through this year as well. And like I said, we'll give you an update to that longer-term model probably later in the year.

坦白說，我認為當我們回顧今年時，這支管理團隊引以為傲的是能夠透過實際行動提升公司獲利。我們去年在這方面做得非常出色，所以我才詳細介紹了我們去年所做的一切。今年我們將繼續專注於發揮槓桿作用。正如我所說，我們可能會在今年稍後向您提供該長期模型的更新資訊。

Jim Schneider, Goldman Sachs.

吉姆‧施耐德，高盛集團。

Relative to your prior comments on NAND, understand there's a little bit more prioritization toward DRAM right now. But when do you expect that your customers are going to sort of pivot from NAND upgrades to more greenfield NAND capacity additions. We saw some announcements from at least one of your customers recently on that.

關於您之前對 NAND 的評論，需要說明的是，目前 DRAM 的優先順序更高一些。但您預計您的客戶何時會從 NAND 升級轉向更多全新的 NAND 容量擴充？我們最近從你們的至少一位客戶那裡看到了相關的公告。

So I'm curious about when you expect to sort of see that upgrade business turn into greenfield business. Could that be before the end of 2026? Or is it more of a 2027 event or maybe even later? Thank you.

所以我很好奇，您預計什麼時候才能看到這種升級改造業務轉變為全新的業務。會不會在2026年底前實現？或者更可能是 2027 年甚至更晚的事？謝謝。

Yeah, it's a great question. I think that what we're -- the way we view it right now is that because of the clean room space constraints, it's probably again, part of that multi-year build-out '27, '28. When cleanroom space is sufficiently available such that they can invest in additional NAND capacity in a big way. So that's probably our view right now.

是的，這是一個很好的問題。我認為我們現在看待這個問題的方式是，由於無塵室空間的限制，這可能又是 2027 年、2028 年多年建設的一部分。當無塵室空間足夠寬敞，可以大規模投資增加 NAND 快閃記憶體容量時。所以這大概就是我們目前的看法。

In the meantime, we talked about the acceleration of the technology transitions. You do get bit growth, you get more capacity of the higher-performing bits that are in strong demand at AI. And so I think that those are the decisions that people are making today is move ahead as quickly as possible with many of the key technology transitions. And so we're busy doing upgrades, and that's where our focus is right now. But greenfield will come eventually, and you've seen some of those initial announcements, and I think that's encouraging for all of us.

同時，我們也討論了技術轉型加速的問題。你的確可以獲得比特成長，獲得更多高效能比特的容量，而這些高效能比特在人工智慧領域需求旺盛。所以我認為，人們今天正在做出的決定就是盡快推動許多關鍵技術轉型。所以我們現在正忙著升級改造，這也是我們目前的重點。但綠地開發終究會到來，你們也看到了一些初步的公告，我認為這對我們所有人來說都是令人鼓舞的。

And then maybe just as a follow-on, I think we all can see the trends by Foundry, DRAM, and NAND. They're in play right now in terms of level of growth rate. But as we head into 2027 or the end of 2026, do you see the potential rank order of those growth rates sort of changing amongst those categories?

然後，或許可以順便提一下，我想我們都可以看到晶圓代工、DRAM 和 NAND 的發展趨勢。就目前的成長率而言，它們正處於發揮作用的狀態。但隨著我們進入 2027 年或 2026 年底，您是否認為這些類別中的成長率排名順序可能會改變？

Oh, man, Jim, that's a great question going in '27. We just, for the first time, give you a '26. You're asking '27. Listen, in '26, we're very confident everything is growing. It's unequivocal. And we're also very clear when we look everything is constrained, frankly, right? You're hearing it from every one of our customers when we talk about things, and they're talking about these multi-year agreements to kind of deliver the visibility into next year.

哦，老兄，吉姆，這是個關於27年的好問題。我們第一次為您提供 '26'。你問的是「27」。聽著，我們非常有信心，2026 年一切都在成長。這點毋庸置疑。而且我們也很清楚，當我們審視一切時，會發現一切都受到限制，坦白說，對吧？當我們談論這些事情時，我們所有的客戶都會提到這一點，他們都在談論這些多年期協議，以便對明年的情況有所了解。

Foundry-logic has grown a lot this year. DRAM has grown a lot this year. NAND has grown a little bit less, but still growing pretty well this year. At the end of the day, though, when you look at the system architectures, all this stuff needs to fit together, and you saw one of the big accelerator guys talking about this at CES, like, hey, we need this NAND stuff showing up. That's happening clearly, so. Into '27, I think we're going to see another year where everything is growing. I'm not ready to rank order it quite yet, Jim, no.

Foundry-logic 今年發展迅速。今年DRAM成長迅猛。NAND 的成長速度略有放緩，但今年仍成長良好。但歸根結底，當你審視系統架構時，你會發現所有這些東西都需要相互契合。你在 CES 上也看到一位大型加速器廠商談到這一點，例如，嘿，我們需要 NAND 技術來應用。很明顯，這種情況正在發生。我認為，到了 2027 年，我們將會看到各行各業持續成長的另一年。吉姆，我還沒準備好要給它排名。

I think as we move through this year, though, we already, I would say, have better visibility into the following year than I think I can ever remember. And that's simply because customers know that they're building these fabs. They're announcing them. They're signaling to their customers they're going to have that capacity available. And so clearly, we're having discussions at this point on what tools are going to be needed, what technology nodes they're going to be running in those fabs. And they want to make sure that they can secure the capacity such that that fab can be started up and producing as quickly as possible.

不過，我認為隨著今年的推進，我們對明年的前景比以往任何時候都更加清晰，這大概是我記憶中最深刻的一次了。那是因為客戶知道他們正在建造這些晶圓廠。他們正在宣布這些消息。他們這是在向客戶表明他們將具備相應的產能。因此，很顯然，我們目前正在討論需要哪些工具，以及這些晶圓廠將採用哪些技術節點。他們希望確保產能，以便工廠能夠盡快啟動並投入生產。

And so those discussions on those fabs are clearly out into 2027. But I think in terms of exactly how those decisions get made through this year, once you have cleanroom space, in some cases, Ken, as we just talked about, they can trade off sometimes a little bit of a cleanroom space to be used for DRAM or for NAND or for what we're seeing in a few cases is for advanced packaging.

因此，關於這些晶圓廠的討論顯然會持續到 2027 年。但我認為，就今年這些決策的具體制定方式而言，一旦有了潔淨室空間，在某些情況下，肯，正如我們剛才討論的那樣，他們有時會犧牲一些潔淨室空間用於 DRAM 或 NAND，或者用於我們看到的一些先進封裝。

I talked about the tremendous growth in advanced packaging and the importance that it's -- the role that it's playing. And so we've been seeing that. So I guess we'd have to see the year continuing to evolve and how the markets -- kind of where the demand is the shortest, but we would anticipate, as we said, robust investment across all device segments, and I think that continues on into 2027 across all three segments.

我談到了先進包裝的巨大增長及其重要性——它所發揮的作用。所以我們已經看到了這種情況。所以我想我們得看看今年接下來的發展情況以及市場——也就是需求最短的市場——會如何變化，但正如我們所說，我們預計所有設備領域都會有強勁的投資，我認為這種情況會持續到 2027 年，涵蓋所有三個領域。

Krish Sankar, TD Cowen and Company.

Krish Sankar，TD Cowen and Company。

Congrats on the good results and guide. Doug, my first question is, I understand that you spoke about the global manufacturing footprint doubled over the last four years. Just wondering, as your customers ramp up more onshore manufacturing, would it lead to you increasing shipments from your US specialities in California and Oregon rather than Malaysia for some of your products? If so, what would be the margin implications?

恭喜取得好成績，也感謝你的指導。道格，我的第一個問題是，我了解到你談到了過去四年全球製造業規模翻了一番。我想知道，隨著您的客戶加大本土生產力度，您是否會將部分產品的發貨地從加利福尼亞州和俄勒岡州的美國特產工廠增加，而不是從馬來西亞增加？如果真是如此，利潤率會受到怎樣的影響？

Yeah, Chris. Listen, we have a global manufacturing footprint, right? We've got factories in Oregon, California, Ohio, Malaysia, Taiwan, Korea, Austria. I think I didn't miss anything there. We have some level of flexibility given enough time to move things around if we really need to do that. And as customers tell us what they need and where they need it, we may adjust things. Right now, I think we feel pretty good about how we've got things set up though.

是的，克里斯。聽著，我們擁有全球生產佈局，對吧？我們在俄勒岡州、加利福尼亞州、俄亥俄州、馬來西亞、台灣、韓國、奧地利都設有工廠。我覺得我沒錯過什麼。如果真的需要調整安排，我們有一定的彈性，並有足夠的時間進行調整。根據客戶的需求和需求地點，我們會進行相對應的調整。不過就目前而言，我覺得我們對目前的安排相當滿意。

And then Tim, I just had a follow-up for you, like a technical question. Last year, you had really good traction in ALD moly. Are customers moving away from single wafer ALD to batch ALD for moly? And if so, how would that impact Lam?

提姆，我還有一個後續問題想問你，是一個技術問題。去年，你們在 ALD 鉬領域取得了非常好的進展。客戶是否正在從單晶圓ALD製程轉向批量ALD製程來沉積鉬？如果真是這樣，會對蘭姆產生什麼影響？

No. I mean, well, at this point, if we look -- we had said previously that in kind of the order of adoption, NAND would be first to adopt moly, and we're seeing that followed by foundry-logic and then ultimately by DRAM. What we can say right now is that the customers that have committed to production of using moly in NAND have gone with Lam's tools. We have a very strong position there.

不。我的意思是，嗯，就目前而言，如果我們看一下——我們之前說過，按照採用順序，NAND 將首先採用鉬，我們看到接下來是晶圓代工邏輯，最後是 DRAM。我們現在可以確定的是，那些承諾在 NAND 中使用鉬進行生產的客戶都選擇了 Lam 的工具。我們在那裡擁有非常強大的地位。

And I think the value of that, as we talked in the past, is it means that throughout these first production ramps with ALD moly, we are building an installed base, we're maturing the tool, we're getting process learning. Competitors aren't going to give up. This is an incredibly important market and a big inflection that we've talked about. But we feel really good about our single -- we call it single wafer moly, but if you look at the tool itself, it has multiple stations inside of one chamber in order to give ourselves high productivity. So that's a production tool close today for the industry, and we intend to continue to keep it that way.

我認為，正如我們過去所討論的，它的價值在於，在 ALD 鉬的首次量產爬坡過程中，我們正在建立一個安裝基礎，我們正在使工具成熟，我們正在獲得工藝經驗。競爭對手不會放棄。這是一個極為重要的市場，也是我們之前討論過的重大轉折點。但我們對我們的單晶圓鉬加工設備非常滿意——我們稱之為單晶圓鉬加工設備，但如果你仔細觀察設備本身，你會發現它在一個腔室內有多個工位，從而實現了高生產率。所以，這是目前業界常用的生產工具，我們打算繼續保持這種狀態。

Harlan Sur, JPMorgan.

哈蘭‧蘇爾，摩根大通。

Great job on the quarterly execution. Just as many of your customers have been surprised by the sudden rise in compute and storage demand. And therefore, requirements for more GPUs, XPUs, CPUs, and the associated memory and storage they obviously got caught somewhat flat footed in terms of sort of near to midterm capacity to support that demand curve, right, as you guys outlined.

季度執行工作做得很好。就像你們的許多客戶一樣，運算和儲存需求的突然成長也讓大家感到驚訝。因此，對更多 GPU、XPU、CPU 以及相關記憶體和儲存的需求，顯然他們在中短期產能方面有些措手不及，無法滿足這種需求曲線，正如你們所概述的那樣。

Is the stronger velocity of demand having a similar impact to your manufacturing capability and ability to procure the necessary components and subsystems? And any bottlenecks that you have in your supply chain?

需求成長加快是否對貴公司的生產能力以及採購必要零件和子系統的能力產生了類似的影響？你們的供應鏈中是否有任何瓶頸？

Well, it isn't without a lot of hard work. But one good news is we did a lot of fact finding post the COVID pandemic and the supply shortages that occurred in our own systems at that time. And we made a lot of improvements. And Doug just talked about the global nature of our manufacturing facilities, standing from the US, in Europe, and all through Asia. And we looked at the same thing with respect to our supply chain.

當然，這需要付出大量的努力。但一個好消息是，在新冠疫情爆發後，以及當時我們自身系統出現供應短缺的情況下，我們進行了大量的事實調查。我們做了很多改進。道格剛才談到了我們製造工廠的全球性特點，從美國、歐洲到整個亞洲都有我們的工廠。我們也從供應鏈的角度檢視了同樣的問題。

And I would say today compared to when we had those shortages, we have built a much stronger, broader, deeper supply chain. And so I don't want to sell short the hard work of our supply chain guys today to meet all these expedited pull-in requests from customers. It's very hard work. But today, I would say we're not the big constraint for any of the devices compared to cleanroom space being a constraint to the industry.

而且我認為，與當時物資短缺的情況相比，如今我們已經建立了一個更強大、更廣泛、更深入的供應鏈。因此，我不想低估我們供應鏈人員今天為滿足客戶所有這些加急提貨請求所付出的辛勤努力。這是一項非常辛苦的工作。但就目前而言，我認為與無塵室空間對產業的限制相比，我們並不是任何設備的主要限制因素。

And so as the industry continues to go, we need to keep working to again, expand our capacity, as I said, make our own operations faster. That's why we've done things like automating our warehouses to make the rate at which we can feed those parts from the time they received from the supplier into the manufacturing that much quicker and more efficient. And so we're just continually working on what I would do is our operational velocity, so that we're not the constraint.

因此，隨著產業的不斷發展，我們需要繼續努力，再次擴大產能，正如我所說，加快我們自身的營運速度。這就是為什麼我們採取了諸如倉庫自動化之類的措施，以便更快、更有效率地將零件從供應商收到並輸送到生產環節。因此，我們一直在努力提高我們的營運速度，以免成為瓶頸。

Appreciate that. And then for my second question, one of the significant, obviously, in the incremental drivers of your business among many has been advanced packaging in HBM. You guys did about $1-plus billion in advanced packaging revenues. I think it was in calendar '24, you're anticipating strong 40%-plus growth this year, but can you guys quantify how much advanced packaging grew for the team in calendar '25. And then of that 40% growth this year, is that being more driven by 2.5D the -- 3.5D advanced packaging or HBM?

謝謝。那麼，我的第二個問題是，顯然，在眾多推動您業務成長的因素中，HBM 的先進包裝技術是一個重要的因素。你們在先進封裝領域獲得了超過10億美元的收入。我認為是在 2024 年日曆中，你們預計今年將實現 40% 以上的強勁增長，但是你們能否量化一下 2025 年團隊的高級包裝業務增長了多少？那麼，今年這 40% 的成長，更多是由 2.5D 或 3.5D 先進封裝或 HBM 驅動的呢？

Yeah, Harlan. We didn't quantify 2025 in packaging, except to tell you it grew nicely, and I think we're going to kind of leave it at that. Tim gave you the 40% this year. So we're super excited about what's going on there.

是的，哈蘭。我們沒有對 2025 年進行量化，只是告訴大家它發展得很好，我想我們就到此為止。提姆今年給了你40%的收益。所以我們對那裡正在發生的事情感到非常興奮。

And I'll let Tim talk about the technologies.

接下來就讓蒂姆來談談這些技術吧。

Yeah. We've lumped it together. I mean, it is strength in HBM. Clearly, there's strong demand there. But also I talked about more complex packaging schemes across advanced foundry-logic, and that's an important driver for us as well.

是的。我們已將它們歸為一類。我的意思是，這是 HBM 的優勢。顯然，那裡的需求很旺盛。但我還談到了先進代工邏輯中更複雜的封裝方案，這對我們來說也是一個重要的驅動因素。

The great thing about our advanced packaging capabilities is they're used in the advanced packaging of all device types. And so it's things like copper plating. It's things like etch, dielectric gap fills. And so they're really fundamental technologies to the success. So we see that as a really important business, and we've talked about the fact that we continue to invest in new technologies in that space.

我們先進封裝技術的優點在於，它可以應用於所有類型設備的先進封裝。所以就包括鍍銅之類的工藝。例如蝕刻、介電間隙填充之類的。因此，它們是成功的基礎技術。所以我們認為這是一項非常重要的業務，我們也討論過我們將繼續投資該領域的新技術。

Stacy Rasgon, Bernstein Research.

Stacy Rasgon，伯恩斯坦研究公司。

For my first one, Doug, you've clearly said it's a second-half-loaded year, which is fine. What does that imply for the first half of the calendar year? Like is March quarter the trough? Do you think things are kind of flattish at the March level until we get that second half inflection? Just how are you thinking about the shape of the year?

道格，對於我的第一個問題，你已經明確表示今年下半年工作量很大，這沒問題。這對今年上半年意味著什麼？例如，三月的季度是低谷期嗎？你認為在下半年出現轉機之前，3月的市場水準會比較平穩嗎？你覺得今年的發展趨勢如何？

Yeah. Stacy, it's a great question. Frankly, as I sit here right now, I think we're going to see growth every quarter from the previous quarter. I'm not going to give you a precise number. We feel good about that March quarter. I think June probably grows from that, September from that, and it ends up being a second-half-weighted year both from a WFE standpoint and from our revenue.

是的。史泰西，這是一個很好的問題。坦白說，就我目前的情況來看，我認為每季都會比上個季度有所成長。我不會給你一個確切的數字。我們對三月的季度業績感到滿意。我認為 6 月可能會因此成長，9 月也會因此成長，最終無論從 WFE 的角度來看還是從我們的收入來看，這都將是一個下半年加權的年份。

I guess to get there, would you need an inflection in that growth rate in the second half? I guess some of it compares which makes it easier. But just are you thinking there's an inflection? Do you think the growth is steady or --?

我想，要達到這個目標，下半年的成長率是否需要出現拐點？我想其中一些方面是可以比較的，這讓比較變得容易一些。但你認為這其中有轉捩點嗎？你認為這種成長是穩定的嗎？——？

I think it's reasonably steady. I mean, part of this is going to be modulated by, okay, how much cleanroom space there is available available at each customer. And I think that they're trying to figure out still and so are we, which is why I'm not giving you more specificity. It will be second half weighted. But like I said, I think you'll see growth quarter by quarter as we go through '26.

我認為它相當穩定。我的意思是，這其中一部分將取決於，好吧，每個客戶有多少潔淨室空間可用。我認為他們還在努力弄清楚，我們也是一樣，所以我才沒有給出更具體的解釋。下半年權重將有所不同。但就像我說的，我認為隨著我們進入 2026 年，你會看到每個季度都有成長。

Stacy, I guess the only thing I would add is -- I was just going to add that my comment about basically every customer is asking for pull-ins. And so there is some element of whether or not we can accelerate some small portion into the first half of the year. We would still see growth in the second because obviously, that probably been things start pulling in from the first half of next year as well. But we're in an accelerating environment of both demand and also timing requests.

Stacy，我想我唯一要補充的是——我本來想補充的是，我之前說過，基本上每個顧客都在要求拉入式。因此，能否將其中一小部分提前到今年上半年完成，就存在一定的考量因素。我們預計第二季度仍將保持成長，因為很顯然，明年上半年的一些因素也可能開始顯現。但我們正處於一個需求和時間要求都在加速變化的環境中。

And so I think that back to the question which was asked about constraints, I think we need to see through the year how those play out as to kind of how much we can do.

所以，回到之前提出的關於限制的問題，我認為我們需要觀察這一年中這些限制是如何發揮作用的，才能確定我們能做多少事情。

And for my second question, I just wanted to ask about China. So Doug, I think you said you expected China to be flattish year over year. Was that a market statement or was that a Lam revenue statement?

我的第二個問題，我想問的是關於中國的問題。道格，我記得你說過你預計中國經濟將逐年保持穩定。那是市場報告還是林氏集團的營收報告？

And the percentage should go down. I guess, it was 36% or something in calendar '25. You had talked previously about like a 30% threshold. Do you think it gets to that 30%? Or do you think it's just down but doesn't quite get there?

這個比例應該會下降。我猜，2025 年的日曆年裡，這個數字是 36% 左右。你之前提到過30%的門檻。你認為會達到30%嗎？還是你認為它只是下降了，但還沒有完全達到目標？

Yeah, Stacy. The comment, in fact, we think WFE in China is flat-ish, ‘25to ‘26, and everything else is going to grow. So as a percent of the total, it's going to be down. We didn't give a precise number. Whether it's in the low-30s or high-20s, that's plus or minus probably where it is. And part of it will be modulated by how much growth comes from outside of China. So numerator-denominator thing as well, obviously.

是的，斯黛西。事實上，我們認為中國的WFE（女性就業和收入）基本上持平，在25%到26%之間波動，而其他一切都將成長。所以佔總數的百分比將會下降。我們沒有給出確切的數字。可能是在30度出頭，也可能是20度以上，大概就是這個範圍。其中一部分將受到中國以外地區經濟成長比例的影響。所以，顯然也牽涉到分子分母的問題。

Blayne Curtis, Jefferies.

布萊恩·柯蒂斯，傑富瑞。

A couple of questions. Maybe just Doug I wanted to just understand the strength in Reliant with China down, is that multinational? I just was curious where you're seeing that demand. I know you said it was lumpy. I just was curious why it was up so much.

幾個問題。也許只是道格，我只是想了解在中國經濟低迷的情況下，Reliant 的優勢是什麼，它是一家跨國公司嗎？我只是好奇你在哪裡看到了這種需求。我知道你說過它表面凹凸不平。我只是好奇為什麼它出現得這麼頻繁。

It was multinational and it was China. It was a little bit of both of them, Blayne.

它是一家跨國公司，而且是一家中國公司。布萊恩，這其中既有他們的一些特質。

And then just on the NAND front, obviously, the demand is very strong. You talked about the upgrades happening earlier. Is that in the camp of also second half weighted? I mean, it's not waiting on clean room space? I'm just kind of curious the shape of NAND for the year.

而就 NAND 快閃記憶體方面，顯然需求非常強勁。你之前提到過正在進行的升級。這是否也屬於下半年權重分配的範疇？我的意思是，它不需要等待無塵室空間嗎？我只是有點好奇今年NAND快閃記憶體的走勢。

Yeah, it probably is a little bit second half weighted, Blayne.

是的，布萊恩，這可能有點偏重下半年的成績。

Melissa Weathers, Deutsche Bank.

梅麗莎‧韋瑟斯，德意志銀行。

I wanted to go back to the NAND side and touch on something that Tim mentioned in his prepared remarks on the expanding applications for NAND in the data center. And Doug, you kind of alluded to some of the CES announcements as well. So is the right interpretation that those applications in the data center have expanded versus what you guys had been thinking? Because you guys have been talking about NAND in the data center for several quarters now. So is that the right way to think about it?

我想回到 NAND 方面，談談 Tim 在他準備好的發言稿中提到的關於 NAND 在資料中心不斷擴展的應用。道格，你也稍微提到了一些CES展會上的發佈內容。所以，資料中心中的這些應用是否比你們之前想的還要多？這種解釋是否正確？因為你們已經連續好幾個季度都在討論資料中心的 NAND 快閃記憶體了。這種思考方式正確嗎？

And then what could this mean for, like your moly ALD, your 300-layer-type devices and the expanding share you could get there?

那麼，對於像鉬原子層沉積（Moly ALD）、300層裝置以及您可能獲得的不斷擴大的市場份額來說，這意味著什麼？

Yeah, sure. I think we characterize it as a new use case. So I don't think we saw this particular use case coming, which is related to the AI inference and kind of the expansion of KV cache and such. I think our previous estimates have been more -- kind of on more traditional storage for using enterprise SSDs. And so, yeah, this is an expansion and kind of presents a bit longer-term growth opportunity for NAND.

當然可以。我認為我們可以將其定義為一個新的應用案例。所以我覺得我們之前並沒有預料到這種特殊的用例，它與人工智慧推理以及鍵值快取的擴展等等有關。我認為我們之前的估計更多是基於使用企業級固態硬碟等傳統儲存方式。所以，是的，這是一種擴張，也為 NAND 帶來了一些更長期的成長機會。

And so therefore, it would be beyond the kind of projections that we would have given back at Investor Day a year ago for the outlook and long term.

因此，這將超出我們一年前在投資者日上對前景和長期發展所做的預測範圍。

And then a quick question on the inventory side of things. Doug, I just wanted to check in and see how you're thinking about parts availability and your ability to scale production in line with demand. Can you help us with a framework to think about how you're thinking about inventories on a days or dollars basis?

然後還有一個關於庫存方面的問題。道格，我只是想了解你對零件供應情況的看法，以及你根據需求擴大生產規模的能力。您能否幫助我們建立一個框架，以便我們思考您是如何以天數或金額為基礎來考慮庫存的？

Yeah, Melissa. No, it's a great question. Listen, if we're right about how things play out here, it's very likely that we're going to need to build some inventory in total dollar terms as we go through the year, right? When business grows, you got to have stuff ready for that growing business. So I think that's clearly going to happen.

是的，梅麗莎。不，這是一個很好的問題。聽著，如果我們對事情的發展走向判斷正確，那麼很可能我們需要在今年內增加一些庫存，以總美元計算，對吧？當業務成長時，你必須為業務成長做好準備。所以我認為這顯然會發生。

We will remain focused on asset utilization and efficiencies and hopefully be able to drive turns up a little bit from here. But we definitely are going to need to build some inventory in advance of a growing top line. So we'll be working on all of that, Melissa.

我們將繼續專注於資產利用率和效率，希望能夠在此基礎上略微提高營業額。但我們肯定需要提前儲備一些庫存，以應對不斷增長的營收。所以我們會著手處理所有這些事情，梅麗莎。

Joe Quatrochi, Wells Fargo.

喬·夸特羅奇，富國銀行。

Maybe just a follow-up on that. Is there any area of your supply chain where you're pursuing suppliers maybe that could be a potential area of shortage? Or do you feel like there's available capacity to continue to kind of support the growth you're talking about?

或許可以就此做個後續說明。你們供應鏈中是否有某個環節，正在尋找可能短缺的供應商？或者您認為目前還有足夠的能力繼續支持您所說的這種成長？

Well, I think that we don't have any line of sight to significant problems at this point. I made the comment a couple of times. It's really a lot of work given the accelerated nature of the demand and the the high levels of demand and customer requests for pull-ins well within our normal lead times. But at this point, we're working across our global supply chain to ensure that we can meet the demand.

嗯，我認為目前我們還看不到任何重大問題的跡象。我發表過幾次這樣的評論。鑑於需求的加速成長以及客戶對在正常交貨期內提貨的高需求和要求，這確實是一項繁重的工作。但目前，我們正在與全球供應鏈各環節合作，確保能夠滿足市場需求。

And maybe as a follow-up, China now expected to be flattish. Is that a reflection of just the affiliate rule impacts re-entering kind of WFE across the company base in terms of just your peers? Or is there a change in the underlying demand that you're seeing as well in China?

或許接下來，中國預計經濟將趨於穩定。這是否僅僅反映了關聯規則對公司內部重新進入WFE（員工福利執行）領域的影響，尤其是在你的同業中？或者，您是否也在中國觀察到潛在需求的變化？

I think, Joe, it's probably a little bit of affiliate rule, but frankly, there's a broad-based set of customer spending in China that have nothing to do with the affiliate rules. So it's the mosaic of everything that's going on there. It's very broad.

喬，我覺得這可能跟聯盟行銷規則有點關係，但坦白說，中國消費者的消費行為中有很多與聯盟行銷規則無關。所以，它就像一幅馬賽克畫，展現了那裡正在發生的一切。範圍非常廣泛。

Thanks, Joe. Operator, we will be one more question.

謝謝，喬。接線員，我們還有一個問題。

Vijay Rakesh, Mizuho.

Vijay Rakesh，瑞穗銀行。

Just a quick question on the Foundry side. I think China was down, I guess, you mentioned affiliate rule, but your Foundry is growing almost 100%-plus year on year. Just wondering as you look at '26, '27 with some of the leading-edge Foundries accelerating how you see that roadmap.

關於 Foundry 這邊，我有個小問題。我想中國市場可能下滑了，你提到了聯盟行銷規則，但你們的 Foundry 每年成長近 100% 以上。我只是好奇，當您展望 2026 年和 2027 年，以及一些領先的代工廠加速發展時，您是如何看待這一發展路線圖的。

Yeah. Well, I guess speaking to whatever it looks like from a roadmap perspective. Each technology node, we said the opportunity for Lam from an etch and intensity perspective and how our tools like the Akara and others fit into that, the opportunities get bigger. As you move forward, you start seeing things that, again, we would anticipate future nodes -- talk '27, '28 introduction of things like backside power again, more use of advanced packaging across more of the leading-edge Foundry space. All of those things are good for us from both the SAM and a share perspective. So from a product perspective, it's a very good picture for us.

是的。嗯，我想說的是，從路線圖的角度來看。我們說過，每個技術節點都為 Lam 提供了從蝕刻和強度角度來看的機會，以及我們的工具（如 Akara 等）如何融入其中，機會也隨之增加。隨著技術的進步，我們開始看到一些事情，這些事情我們預計會在未來的節點上出現——例如在 2027 年、2028 年再次引入背面供電等技術，以及在更多領先的代工廠領域更多地使用先進的封裝技術。從 SAM 和股份的角度來看，所有這些都對我們有利。所以從產品角度來看，這對我們來說是一個非常好的前景。

Got it. And then on the DRAM side, I know you mentioned briefly HBM4 with 16 layers. Obviously, that's a nice step-up from where HBM is now. Can you talk to what that does for your WFE, the content and the growth there on the DRAM HBM side? Thanks.

知道了。至於DRAM方面，我知道您曾經簡單提到16層的HBM4。顯然，這比 HBM 目前的狀態有了很大的提升。您能談談這對您的WFE、DRAM HBM的內容和成長有何影響嗎？謝謝。

Yeah. I mean, just in general terms, I mean, what happens is you end up going to next-generation HBM, dies become bigger, and that's generally what is creating the majority of the problem relative to when we talk about clean room space constraints. You need more clean room space and more tooling per fit that comes out of the fab. So therefore, that was what we're trying to communicate. Obviously, the performance improves, but the space required and the equipment required increases.

是的。我的意思是，總的來說，是說，最終你會採用下一代 HBM，晶片會變得更大，而這通常是造成無塵室空間限制問題的主要原因。你需要更大的無塵室空間，以及更多的模具來組裝每個零件。所以，這就是我們想要傳達的訊息。顯然，性能提高了，但所需的空間和設備也增加了。

Yeah. Appreciate everybody's questions today. That concludes our call for today. We look forward to seeing everybody as we do the conference circuit and get out on the road. So thank you for your time today.

是的。感謝大家今天提出的問題。今天的通話就到這裡。我們期待在參加各種會議和外出巡迴演出時見到大家。非常感謝您今天抽出時間。

Thank you, everyone. The conference has now concluded. You may now disconnect.

謝謝大家。會議已經結束。您現在可以斷開連線了。