益華電腦 (CDNS) 2025 Q1 法說會逐字稿

Good afternoon. My name is Tina, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence first quarter 2025 earnings conference call. (Operator Instructions). I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

午安.我叫蒂娜，今天我將擔任您的會議接線生。現在，我歡迎大家參加 Cadence 2025 年第一季財報電話會議。（操作員指令）。現在我將電話轉交給 Cadence 投資人關係副總裁 Richard Gu。請繼續。

Thank you, operator. I would like to welcome everyone to our first quarter of 2025 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer, and John Wall, Senior Vice President, and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com.

謝謝您，接線生。歡迎大家參加我們 2025 年第一季財報電話會議。今天與我一起出席的還有總裁兼執行長 Anirudh Devgan 和資深副總裁兼財務長 John Wall。本次電話會議的網路直播和今天準備好的發言稿副本將在我們的網站 cadence.com 上提供。

Today's discussion will contain forward-looking statements, including our outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion.

今天的討論將包含前瞻性陳述，包括我們對未來業務和經營績效的展望。由於風險和不確定性，實際結果可能與今天討論中預測或暗示的結果有重大差異。

For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent Forms 10-K and 10-Q, CFO commentary and today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update that.

有關可能導致實際結果不同的因素的信息，請參閱我們的 SEC 文件，包括我們最新的 10-K 和 10-Q 表格、首席財務官評論和今天的收益報告。本次電話會議中的所有前瞻性陳述均基於我們截至今天掌握的估計和信息，我們不承擔更新這些估計和信息的義務。

In addition, all financial measures discussed on this call are non-GAAP unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. For the Q&A session today, we would ask that you observe a limit of one question only. If time permits, you can requeue with additional questions.

此外，除非另有說明，本次電話會議中討論的所有財務指標均為非公認會計準則。非 GAAP 指標不應與 GAAP 結果分開考慮或替代 GAAP 結果。今天的收益報告包含了 GAAP 與非 GAAP 指標的對帳。對於今天的問答環節，我們希望您只限提問一個問題。如果時間允許，您可以重新排隊詢問其他問題。

Now I'll turn the call over to Anirudh.

現在我將把電話轉給 Anirudh。

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm pleased to report that Cadence delivered excellent results for the first quarter of 2025, with robust ongoing customer demand for innovative technologies.

謝謝你，理查。大家下午好，感謝大家今天的參與。我很高興地報告，Cadence 在 2025 年第一季取得了優異的業績，客戶對創新技術的需求持續強勁。

We exceeded our guidance on all key financial metrics, achieving 23% year-over-year revenue growth and 34% increase in non-GAAP EPS. Given this outperformance and the continued strength of our business, we are raising our financial outlook for the year. John will provide more details in a moment.

我們在所有關鍵財務指標上都超出了預期，實現了同比增長 23% 的收入和非公認會計準則每股收益增長 34%。鑑於這一優異表現和我們業務的持續強勁，我們上調了今年的財務前景。約翰稍後將提供更多細節。

Before diving in, I would like to share my perspective on the prevailing macroeconomic uncertainty. Semiconductors remain foundational to realizing transformative technologies such as hyperscale computing, 5G and autonomous systems, all fueled by the AI megatrend. We haven't seen any shifts in customer behavior at this time, as they continue investing in their next-generation designs, recognizing that today's R&D efforts are critical to deliver groundbreaking products of tomorrow.

在深入探討之前，我想先分享我對當前宏觀經濟不確定性的看法。半導體仍然是實現超大規模運算、5G 和自主系統等變革性技術的基礎，而所有這些技術都受到人工智慧大趨勢的推動。目前我們還沒有看到客戶行為發生任何變化，因為他們繼續投資於下一代設計，並認識到今天的研發工作對於提供未來的突破性產品至關重要。

Additionally, our ratable software business model, strong Q1 exit backlog and a predominantly recurring revenue mix, provide resilience and excellent visibility. Customers are increasingly relying on our product. As we execute our intelligent system design strategy and expand our differentiated end-to-end portfolio to serve a growing and diversified customer base.

此外，我們的可評級軟體業務模式、強勁的第一季退出積壓訂單以及以經常性收入為主的收入組合，提供了彈性和出色的可見性。客戶越來越依賴我們的產品。我們執行智慧型系統設計策略並擴展差異化的端到端產品組合，以服務不斷增長和多樣化的客戶群。

As the fast-evolving AI landscape expands the market opportunities and reshapes the entire chip and system development process, our Cadence AI portfolio delivers unparalleled PPA, productivity and time-to-market benefits. At GTC, we announced an expanded partnership with NVIDIA on their latest Grace Blackwell architecture.

隨著快速發展的人工智慧領域擴大市場機會並重塑整個晶片和系統開發流程，我們的 Cadence AI 產品組合可提供無與倫比的 PPA、生產力和上市時間優勢。在 GTC 上，我們宣布與 NVIDIA 就其最新的 Grace Blackwell 架構擴大合作。

In addition to enabling up to an ADX acceleration of Cadence solvers, we are collaborating on developing a full stack agentic AI solution for engineering and science using the new Llama Nemotron Reasoning Model . We are also one of the first adopters of NVIDIA Omniverse blueprint for AI factory digital twins, advancing data center design and operational efficiency.

除了支援高達 ADX 加速的 Cadence 求解器之外，我們還在合作開發用於工程和科學的全端代理 AI 解決方案，使用新的 Llama Nemotron 推理模型。我們也是首批採用 NVIDIA Omniverse 藍圖實現 AI 工廠數位孿生的企業之一，致力於提升資料中心設計和營運效率。

In Q1, we expanded our footprint at several top-tier customers and further our relationship with key ecosystem partners. Last quarter, we announced a collaboration with Rapidus on two-nanometer IP development. And this quarter, we are pleased to share that Rapidus has made a wide-ranging commitment to our core EDA software portfolio across digital, custom analog and verification solutions.

在第一季度，我們擴大了與幾家頂級客戶的業務範圍，並進一步加強了與主要生態系統合作夥伴的關係。上個季度，我們宣布與 Rapidus 合作進行兩奈米 IP 開發。本季度，我們很高興地宣布，Rapidus 已對我們的核心 EDA 軟體產品組合做出了廣泛的承諾，涵蓋數位、客製化類比和驗證解決方案。

In Q1, we deepened our partnership with Socionext through a broad expansion of our EDA software, particularly AI-driven digital solutions, along with our system software. We furthered our partnership with a marquee hyperscaler through a broad proliferation of our digital and verification software, particularly our AI-driven Cadence Cerebrus and Verisium solutions.

在第一季度，我們透過廣泛擴展我們的 EDA 軟體（特別是人工智慧驅動的數位解決方案）以及系統軟體，深化了與 Socionext 的合作夥伴關係。我們透過廣泛推廣我們的數位和驗證軟體，特別是我們的人工智慧驅動的 Cadence Cerebrus 和 Verisium 解決方案，進一步加強了與大型超大規模企業的合作。

We expanded our collaboration with Intel foundry by officially joining the Intel Foundry Accelerator Design Services alliance. From systems on chip to advanced IP for AI and HPC applications, Cadence's inclusion in the Alliance helps Intel Foundry customer remain at the forefront of innovation.

我們正式加入英特爾代工廠加速器設計服務聯盟，擴大了與英特爾代工廠的合作。從系統單晶片到用於 AI 和 HPC 應用的高級 IP，Cadence 加入聯盟有助於英特爾代工廠客戶始終處於創新前沿。

Now let's talk about key product highlights for Q1. Our IP business grew 40% year-over-year in Q1 as we continue to benefit from the strong market opportunities offered by AI, chiplet-based architectures, and the foundry ecosystem build-out. We secured a major expansion at a global marquee system company for our AI, HPC design IP and deepened our partnership with a major foundry through their commitment to our memory and interface IP.

現在讓我們來談談第一季的主要產品亮點。由於我們繼續受益於人工智慧、基於小晶片的架構和代工生態系統建設所提供的強大市場機遇，我們的 IP 業務在第一季同比增長了 40%。我們在一家全球知名系統公司實現了 AI、HPC 設計 IP 的大規模擴展，並透過他們對我們的記憶體和介面 IP 的承諾深化了與一家大型代工廠的合作夥伴關係。

Following our pending acquisition of Secure-IC last quarter, we continue to build out our design IP portfolio. And earlier this month, we entered into a definite agreement to acquire Arm's Artisan Foundation IP business. Our Tensilica DSP is the defacto choice for automotive ADAS and infotainment systems and continued being widely integrated into vision, radar, LiDAR, and audio systems for the autonomous driving as well as emerging robotics use cases.

繼上個季度我們即將收購 Secure-IC 之後，我們繼續擴展我們的設計 IP 產品組合。本月初，我們達成了收購 Arm 的 Artisan Foundation IP 業務的最終協議。我們的 Tensilica DSP 是汽車 ADAS 和資訊娛樂系統的實際選擇，並繼續廣泛整合到自動駕駛和新興機器人用例的視覺、雷達、LiDAR 和音訊系統中。

Our core EDA revenue grew 16% and year-over-year in Q1 with further proliferation of our digital full flow at the most advanced nodes. Cadence Cerberus AI solution continued its strong momentum with nearly 50 new logos in Q1 and well over 1,000 tape-outs till date.

隨著我們最先進節點的數位全流程進一步擴散，我們的核心 EDA 收入在第一季同比增長了 16%。Cadence Cerberus AI 解決方案持續保持強勁勢頭，第一季新增近 50 個標識，迄今已完成超過 1,000 次流片。

Engineering change orders or ECOs are a critical part of any design process. AI is particularly suitable to dramatically improve ECOs. Using the new Cadence Conformal AI ECO flows, MediaTek saw early positive results, generating 83% smaller ECO patches in nearly half the run time.

工程變更單或 ECO 是任何設計流程的關鍵部分。人工智慧特別適合大幅改善ECO。使用新的 Cadence Conformal AI ECO 流程，聯發科看到了早期的積極成果，在近一半的運行時間內產生了小 83% 的 ECO 修補程式。

MediaTek also improved run time and memory by 100x through its deployment of Conformal AI low power. Our collection of new smart ECO technologies has better premises automotive functional ECO run times by more than 50% while improving quality. Our flagship Virtuoso platform, the industry's gold standard for advanced node custom analog and mixed signal design, continue to expand into new areas, such as photonics and circuit and yield optimization.

聯發科還透過部署 Conformal AI 低功耗技術，將運行時間和記憶體提高了 100 倍。我們收集的全新智慧 ECO 技術使汽車功能 ECO 運行時間延長了 50% 以上，同時提高了品質。我們的旗艦 Virtuoso 平台是業界先進節點客製化類比和混合訊號設計的黃金標準，並持續擴展到光子學、電路和產量最佳化等新領域。

Ever-increasing complexities in system verification and software bring up continue to drive secular demand for our market-leading Palladium Z3 and Protium X3 platforms. Demand for hardware was broad-based with particular strength driven by AI, HPC and hyperscaler customers. Our hardware products continue to proliferate at existing customers, especially top hyperscalers, while gaining notable competitive wins, including a market shaping semiconductor company.

系統驗證和軟體開發的日益複雜繼續推動我們市場領先的 Palladium Z3 和 Protium X3 平台的長期需求。硬體需求廣泛，尤其受到人工智慧、高效能運算和超大規模客戶的推動。我們的硬體產品在現有客戶（尤其是頂級超大規模客戶）中繼續激增，同時獲得了顯著的競爭勝利，包括一家塑造市場的半導體公司。

Our Verification software suite that includes Verisium, Xcelium and Jasper leverages big data and AI to optimize verification workloads and saw continued expansion across aerospace and defense, electronics and automotive segments.

我們的驗證軟體套件包括 Verisium、Xcelium 和 Jasper，利用大數據和人工智慧來優化驗證工作負載，並在航空航太和國防、電子和汽車領域持續擴展。

Our System Design & Analysis business delivered more than 50% year-over-year revenue growth in Q1, as our AI-driven optimization solutions integrated with our physics-based simulation platforms continue to deliver superior results across multiple end markets.

我們的系統設計與分析業務在第一季實現了超過 50% 的同比收入成長，因為我們的人工智慧驅動的最佳化解決方案與基於實體的模擬平台相結合，繼續在多個終端市場提供卓越的成果。

Our digital twin reality data center product gained momentum, signing multiple deals with large hyperscaler and cloud service providers in Q1. BETA CAE delivered a strong quarter and is opening up tremendous opportunities for us in the automotive vertical.

我們的數位孿生現實資料中心產品發展勢頭強勁，在第一季與大型超大規模企業和雲端服務供應商簽署了多項協議。BETA CAE 本季表現強勁，為我們在汽車垂直領域開闢了巨大的機會。

Allegro X's Omniverse integration was highlighted at GTC with photorealistic 3D views of a full multi-board system designed in Allegro X. Our AI-driven substitute router, the industry's premier solution for full package routing of complex 3D-IC designs saw strong customer interest and engagement during the early adopter program.

Allegro X 的 Omniverse 整合在 GTC 上得到了重點展示，展示了在 Allegro X 中設計的完整多板系統的逼真的 3D 視圖。我們的 AI 驅動替代佈線器是業界首屈一指的複雜 3D-IC 設計全封裝佈線解決方案，在早期採用者計畫期間引起了客戶的強烈興趣和參與。

In summary, I'm pleased with our Q1 results and the continued momentum of our business. The growing complexity of chip and system design, coupled with the transformative potential of AI-driven automation creates significant opportunities for our products to enable and empower our customers. In addition to our strong business results, I'm proud of our high-performance inclusive culture.

總而言之，我對我們的第一季業績和業務的持續發展勢頭感到滿意。晶片和系統設計日益複雜，加上人工智慧驅動自動化的變革潛力，為我們的產品創造了重大機遇，能夠支持和賦能我們的客戶。除了我們強勁的業務業績外，我還為我們高績效的包容性文化感到自豪。

I'm thrilled that Cadence was recently named by Fortune and Great Place to Work, yet again as one of the 100 best companies to work for. ranking number 11. Now I will turn it over to John to provide more details on the Q1 results and our updated 2025 outlook.

我很高興 Cadence 最近再次被《財星》雜誌和《最佳工作場所》評選為 100 家最適合工作的公司之一。排名第11名。現在，我將把時間交給約翰，讓他提供有關第一季業績和我們更新的 2025 年展望的更多詳細資訊。

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered excellent results for the first quarter of 2025 with broad-based strength across all of our businesses. Robust design activity and customer demand drove 23% year-over-year revenue growth and 42% non-GAAP operating margin for Q1. And we are tracking ahead of our original forecast for 2025. We repurchased more Cadence shares than initially planned in Q1, which reduced our share count.

謝謝，Anirudh，大家下午好。我很高興地報告，Cadence 在 2025 年第一季取得了優異的業績，我們所有業務都表現出色。強勁的設計活動和客戶需求推動第一季營收年增 23% 和非 GAAP 營業利潤率成長 42%。我們正在提前實現 2025 年的最初預測。我們在第一季回購了比最初計劃更多的 Cadence 股票，這減少了我們的股票數量。

Here are some of the financial highlights from the first quarter, starting with the P&L. Total revenue was $1.242 billion. GAAP operating margin was 29.1%, and non-GAAP operating margin was 41.7%, the GAAP EPS was $1, with non-GAAP EPS $1.57.

以下是第一季的一些財務亮點，首先是損益表。總收入為12.42億美元。公認會計準則營業利益率為29.1%，非GAAP營業利益率為41.7%，GAAP每股盈餘為1美元，非GAAP每股盈餘為1.57美元。

Next turning to the balance sheet and cash flow. Cash balance at quarter end was $2.778 billion, while the principal value of debt outstanding was $2 billion and $500 million. Operating cash flow was $487 million. DSOs were 44 days, and we used $350 million to repurchase Cadence shares.

接下來轉向資產負債表和現金流量。季末現金餘額為 27.78 億美元，未償還債務本金價值為 20.5 億美元。經營現金流為4.87億美元。DSO 為 44 天，我們用 3.5 億美元回購了 Cadence 股票。

Before I provide our updated outlook, I'd like to share the assumption that is embedded. It contains the usual assumption that export control regulations that exist today remain substantially similar for the remainder of the year.

在我提供我們最新的展望之前，我想先分享其中蘊含的假設。它包含通常的假設，即現有的出口管制法規在今年剩餘時間內基本保持不變。

Our updated outlook for 2025 is revenue in the range of $5.15 billion to $5.23 billion. GAAP operating margin in the range of 30.25% to 31.25%, non-GAAP operating margin in the range of 43.25% to 44.25%, GAAP EPS in the range of $4.21 to $4.31, non-GAAP EPS in the range of $6.73 to $6.83, operating cash flow in the range of $1.6 billion to $1.7 billion and we expect to use at least 50% of our annual free cash flow to repurchase Cadence shares.

我們對 2025 年的最新預測是營收將在 51.5 億至 52.3 億美元之間。GAAP 營業利潤率在 30.25% 至 31.25% 之間，非 GAAP 營業利潤率在 43.25% 至 44.25% 之間，GAAP EPS 在 4.21 美元至 4.31 美元之間，非 GAAP EPS 在 6.73 美元至 6.83 美元至 4.31 美元之間，非 GAAP EPS 在 6.73 美元至 6.83 美元之間。 50% 的年度自由現金流回購 Cadence 股票。

With that in mind, for Q2, we expect revenue in the range of [$1.250 billion to $1.27 billion], GAAP operating margin in the range of 27.5% to 28.5%, non-GAAP operating margin in the range of 41.5% to 42.5%, GAAP EPS in the range of $0.89 to $0.95 and non-GAAP EPS in the range of $1.55 to $1.61. And as usual, we published a CFO commentary document on our Investor Relations website, which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations.

考慮到這一點，對於第二季度，我們預計收入在 [12.50 億美元至 12.7 億美元] 範圍內，GAAP 營業利潤率在 27.5% 至 28.5% 範圍內，非 GAAP 營業利潤率在 41.5% 至 42.5% 範圍內，GAAP EPS 在 0.EPS8 1.61 美元範圍內。像往常一樣，我們在投資者關係網站上發布了一份首席財務官評論文件，其中包括我們對其他項目的展望以及進一步的分析和 GAAP 與非 GAAP 的對帳。

In conclusion, Cadence is off to a strong start to the year. We are raising our 2025 revenue and EPS outlook. Our technology platform is essential to customers' R&D investment and our resilient software model positions us well in navigating today's dynamic macro environment.

總而言之，Cadence 今年開局強勁。我們正在提高 2025 年營收和每股盈餘預期。我們的技術平台對於客戶的研發投資至關重要，而我們具有彈性的軟體模型使我們能夠很好地應對當今動態的宏觀環境。

I'd like to close by thanking our customers, partners, and our employees for their continued support. And with that operator, we will now take questions.

最後，我要感謝我們的客戶、合作夥伴和員工的持續支持。現在，我們將透過這位接線生來回答問題。

(Operator Instructions) Harlan Sur, JPMorgan.

（操作員指示）Harlan Sur，摩根大通。

Good to see the strong mid-teens percentage year-over-year growth in the core EDA business. On your China business in particular, the team did continue to see year-over-year growth acceleration, right, to plus 13% in the March quarter versus 10% in -- back in the December quarter. On top of this, tariff trade, regulatory dynamics are, I think, driving more focus on domestic China chip design programs, especially in AI, like we've heard, for example, ByteDance engaging on new custom AI chip programs.

很高興看到核心 EDA 業務的年增長率達到 15% 左右。特別是在您的中國業務方面，團隊確實繼續看到同比增長加速，3 月份季度增長了 13%，而 12 月份季度增長了 10%。除此之外，我認為關稅貿易和監管動態正在推動人們更加關注中國國內的晶片設計項目，特別是在人工智慧領域，例如我們聽說位元組跳動正在開發新的客製化人工智慧晶片項目。

We've heard Alibaba as well in automotive, there continues to be more domestic China design programs firing both analog, power management and digital. So is this domestic focus and design activity, a potential tailwind for your China business? Are you getting more optimistic on a growth profile for your China business this year versus the team's prior view of flattish?

我們也聽說阿里巴巴在汽車領域也採取了類似的舉措，中國國內將繼續推出更多設計項目，包括類比、電源管理和數位。那麼，這種國內重點和設計活動是否會對您的中國業務產生潛在的推動作用？與團隊先前認為的持平相比，您是否對今年中國業務的成長前景更加樂觀？

Harlan, great question. So first, I would like to say I'm pretty pleased by our performance in Q1. And the strength is coming in all the various product groups and also all the various geographies, like you mentioned. And the reason is even in this kind of dynamic macro environment, the customers are investing for the future and these R&D investment takes months to years. And as you know, we are tied to the design activity, which is strong.

哈蘭，這個問題問得好。首先，我想說我對我們第一季的表現非常滿意。正如您所說，實力來自各個不同的產品組以及各個不同的地區。原因是，即使在這種動態的宏觀環境下，客戶也在為未來投資，而這些研發投資需要數月甚至數年的時間。如您所知，我們與設計活動的連結非常緊密。

Now in China or actually in multiple countries, the AI development is taking multiple cycles. I mean not just -- as you know, not just the development of data centers, but I've said for a while now that also the development of physical AI systems, which is autos and robots and drones. And China is particularly strong in that as well.

現在在中國，或實際上在多個國家，人工智慧的發展正在經歷多個週期。我的意思不僅僅是——正如你所知，不僅僅是資料中心的發展，而且我已經說過一段時間了，還有物理人工智慧系統的發展，即汽車、機器人和無人機。中國在這方面也特別強大。

So overall, I think we are pleased with the silicon and system development driven for AI. I'm pleased with the performance of the China business and the start of the year, but we're still prudent assuming China flat for the whole year. John, do you want to comment on that?

所以總的來說，我認為我們對人工智慧驅動的矽片和系統開發感到滿意。我對中國業務的表現和今年的開局感到滿意，但我們仍然謹慎地認為中國全年的業務將保持穩定。約翰，你想對此發表評論嗎？

No, I think that's the important point that design activity remains strong in China, but we're continuing to be prudent with the guide for the year and continue to assume that China revenue is flat year-over-year at the midpoint of that guidance.

不，我認為重點是中國的設計活動依然強勁，但我們對今年的指引仍持謹慎態度，並繼續假設中國的收入在該指引的中點與去年同期持平。

Lee Simpson, Morgan Stanley.

摩根士丹利的李辛普森。

Great. Great quarter to everyone. I mean, as we all know, your main offering is a software and services product which may not have the same risk from reciprocal tariffing regimes that other parts of semiconductor ecosystem does. And I think if you look at your hardware chip, it's an ASIC fab TSMC. So I guess one assumes this is only a limited exposure to tariffing into China.

偉大的。祝大家本季一切順利。我的意思是，眾所周知，你們的主要產品是軟體和服務產品，它可能不會像半導體生態系統的其他部分那樣面臨互惠關稅制度帶來的風險。我認為如果你看一下你的硬體晶片，它就是 ASIC 晶圓廠台積電 (TSMC)。因此，我想人們會認為這只是對中國徵收關稅的有限影響。

I guess what I'm asking more generally is, can you confirm if at all, at this point, where you think you do have any exposure to the tariffing regime announced by the US government, at least as you currently understand things?

我想我更籠統地問的是，您能否確認，就目前而言，您是否認為您確實受到美國政府宣布的關稅制度的影響，至少就您目前了解的情況而言？

Yes. So with respect to tariffs, I mean, there are two parts to it, like, one is the customer behavior, which I mentioned, we don't see any change at this time. The second is in our own products, and like you said correctly, software and services are not subject to tariffs.

是的。因此，關於關稅，我的意思是，它有兩個部分，例如，一個是客戶行為，正如我所提到的，我們目前沒有看到任何變化。第二是我們自己的產品，正如您所說，軟體和服務不受關稅影響。

Now in terms of our hardware business, just to remind everyone, we do have multiple manufacturing lines in US and outside US and this was something we did as part of COVID a few years ago. So that's paying off well. So at this point, we don't believe that given our diversified supply chain, the tariffs will have effect on our hardware business as well, okay?

現在就我們的硬體業務而言，需要提醒大家的是，我們在美國和美國以外確實有多條生產線，這是我們幾年前在 COVID 期間所做的事情。所以這是有回報的。所以目前，考慮到我們多樣化的供應鏈，我們認為關稅不會對我們的硬體業務產生影響，好嗎？

Now on the other hand, we continue to monitor the situation. It's a dynamic situation. But I feel that we are more resilient for multiple reasons. One is -- one is that we are tied to the design cycle, we don't see that much change. Second is we are very diversified, both in terms of products and geographies.

另一方面，我們將繼續監測局勢。這是一個動態的情況。但我覺得我們比較有韌性，原因很多。一是──一是我們受設計週期的束縛，我們沒有看到太大的改變。第二，我們的產品和地理都非常多樣化。

And third, we have our business model, it's a ratable business model. So even in these uncertain times, we feel that we have enough confidence and visibility to raise our outlook for 2025.

第三，我們有我們的商業模式，這是可評估的商業模式。因此，即使在這些不確定的時期，我們也認為我們有足夠的信心和預見性來提升我們對 2025 年的展望。

Gary Mobley, Loop Capital.

加里·莫布利（Gary Mobley），Loop Capital。

Yes. Sorry, guys. This question might be more for Anirudh. And the question is more and more tools that you sell like simulation, synthesis, place, and route, et cetera, run on generation-based compute and running GPU-based servers. How does this impact your licensing model? And more specifically, how does it potentially affect your annual contract value?

是的。抱歉，各位。這個問題可能更適合 Anirudh。問題是，您銷售的越來越多的工具（如模擬、綜合、佈局和佈線等）都在基於生成的計算和基於 GPU 的伺服器上運行。這對您的授權模式有何影響？更具體地說，它如何影響您的年度合約價值？

Okay. That's a good question. I think you mean that as we enrich our hardware -- the hardware that our software runs on, how does it affect our business model? So actually, first of all, I've talked about this three-layer cake for a while, right, which is the bottom layer of the cake being accelerate compute.

好的。這是個好問題。我想你的意思是，當我們豐富我們的硬體——我們的軟體運行的硬體時，它會如何影響我們的商業模式？所以實際上，首先，我已經談論過這個三層蛋糕一段時間了，對吧，蛋糕的底層是加速計算。

So in the old days, we used to run on CPUs, mostly x86 CPUs. But now we run on all cancer CPUs, including ARM, and then GPUs, especially with our collaboration very strong collaboration with NVIDIA. And then for hardware, we have our own kind of Boolean processor because GPUs are excellent for numerical calculation, like circuit simulation or CFD or electromagnetics, but for zero-one kind of Boolean calculation, we make a special chip in palladium, right, which is a bullion supercomputer.

所以在過去，我們常常在 CPU 上運行，主要是 x86 CPU。但現在我們在所有癌症 CPU 上運行，包括 ARM，然後是 GPU，特別是我們與 NVIDIA 的密切合作。然後對於硬件，我們有自己的布林處理器，因為 GPU 非常適合數值計算，例如電路模擬或 CFD 或電磁學，但對於零一類布林計算，我們用鈀製作了一個特殊的晶片，對吧，它是金條超級電腦。

So as a reason we have a very rich kind of set of offerings on the hardware side, -- now on the -- how we go to market, we maintain our previous model in which you can subscribe to our -- like the traditional licensing model. which is typically early license over a three-year term.

因此，我們在硬體方面擁有非常豐富的產品系列，現在，關於我們如何進入市場，我們保留了以前的模式，您可以訂閱我們的模式，就像傳統的授權模式一樣。這通常是三年期限的早期許可。

But also, we are offering more and more cloud solutions, okay? And there is an uptick in cloud offering. Now sometimes the cloud could be on public clouds. So then access to CPUs, just like AWS has ARM CPUs and other cloud vendors and also GPUs. But we also have our own cloud offering through outsourced data centers, which are actually seeing good availability.

但是，我們也提供越來越多的雲端解決方案，好嗎？雲端運算服務也呈現上升趨勢。現在，有時雲端可能位於公有雲上。這樣就可以存取 CPU，就像 AWS 有 ARM CPU，其他雲端供應商也有 GPU 一樣。但我們也透過外包資料中心提供自己的雲端服務，這些服務實際上具有良好的可用性。

So in our own Cadence, all the cloud, we also are packaging hardware and software together. A good example is Millennium. Just like we do with Palladium for our own chips. So I think this is a new kind of a business model to see how the customers will react. So we have the existing business model of software only, and then there's cloud SaaS model, especially as the hardware gets more rich in its offering.

因此，在我們自己的 Cadence 中，整個雲端中，我們也將硬體和軟體打包在一起。一個很好的例子就是千禧年。就像我們用鈀來製作自己的晶片一樣。所以我認為這是一種新的商業模式，看看客戶會如何反應。因此，我們現有的商業模式只有軟體，然後還有雲端 SaaS 模式，特別是隨著硬體產品越來越豐富。

Vivek Arya, Bank of America.

美國銀行的維韋克·艾瑞亞（Vivek Arya）。

(technical difficulty) tools. And Anirudh I wanted to check what is Cadence's share gain opportunity because when I look at your main peer, they have double-digit sales exposure at Intel. And I'm curious, what is Cadence's exposure ? How soon can we start to see any potential share gains? Because these kind of relationships tend to be sticky. Just how large is the opportunity? And how soon can it start to show up tangibly for cadence?

（技術難度）工具。Anirudh，我想檢查一下 Cadence 的份額成長機會是什麼，因為當我查看您的主要同行時，發現他們在英特爾的銷售額達到了兩位數。我很好奇，Cadence 的曝光度如何？我們多久才能開始看到潛在的份額成長？因為這種關係往往很棘手。這個機會到底有多大？它多久才能開始明顯地顯示出節奏？

Great question. I think your question is on Intel because I missed the beginning part of the question. But first of all, I'm very pleased by Lip-Bu's appointment as CEO of Intel. And for Cadence, you know for a while, I think our Intel has been a, I would say, a weak spot, relatively speaking, for Cadence. And this goes back 10, 15 years, it's a new issue.

好問題。我認為您的問題是關於英特爾的，因為我錯過了問題的開頭部分。但首先，我很高興立武被任命為英特爾執行長。對於 Cadence 來說，有一段時間，我認為我們的英特爾一直是 Cadence 的一個弱點，相對而言。這已經是 10 到 15 年前的事了，這是一個新問題。

And then the other thing has been our investment in IP. So in general, I'm pleased with the -- so both of these two areas has been the main kind of areas to improve for Cadence IP and Intel. So IP, I think I'm pleased to see improvement last year and also in Q1 and I feel good about our IP business in general.

另一件事是我們對智慧財產權的投資。所以總的來說，我對此感到滿意——這兩個領域一直是 Cadence IP 和英特爾需要改進的主要領域。因此，我很高興看到 IP 去年和第一季都有所改善，而且我對我們的 IP 業務總體感覺良好。

And on Intel, I mean, this is a great opportunity as they have to, of course, reformulate their strategy and how they work with their ecosystem partners and we look forward to engaging with Intel in a lot more detail. And those discussions have already started and we'll keep you updated as they progress.

對於英特爾來說，這是一個很好的機會，因為他們當然必須重新制定他們的策略以及他們與生態系統合作夥伴的合作方式，我們期待與英特爾進行更詳細的合作。這些討論已經開始，我們將及時向大家通報進度。

Gianmarco Conti, Deutsche Bank.

吉安馬科·孔蒂，德意志銀行。

Yes. Congratulations on another great quarter. I guess, could you perhaps shed some detail into your hardware deliveries this quarter and whether you're seeing the demand for the third generation in line with what was expected, especially in customers except in the higher pricing in this environment.

是的。恭喜您又一個季度取得優異成績。我想，您能否詳細介紹本季的硬體交付情況，以及您是否認為第三代產品的需求符合預期，尤其是在客戶方面，除了這種環境下價格較高之外。

And also, I know that you've briefly mentioned about it not having any impact from tariffs. But could you share a few words on your Mexico plans where you assemble your -- your hardware and whether that could be a potential issue from a capacity production standpoint were the tariffs being enacted on that front?

而且，我知道您曾簡要地提到它不會受到關稅的影響。但是，您能否談談您在墨西哥組裝硬體的計劃，以及如果在這方面實施關稅，從產能生產的角度來看這是否會成為一個潛在的問題？

Gianmarco, great questions. The -- generally, our hardware revenue is limited by our production capacity because demand continues to outstrip our ability to supply the -- the hardware products. And as Anirudh mentioned earlier, we have a very resilient and agile supply chain capability, and that lessens the direct impact of tariffs. We -- we produce and build hardware for the US market in the US and for the international market outside the US.

Gianmarco，很棒的問題。一般來說，我們的硬體收入受到生產能力的限制，因為需求持續超過我們供應硬體產品的能力。正如阿尼魯德之前提到的，我們擁有非常有彈性和敏捷的供應鏈能力，這減輕了關稅的直接影響。我們在美國為美國市場以及美國以外的國際市場生產和製造硬體。

So that gives us kind of an optimized setup for hardware. So we're not seeing any real direct impact of tariffs in our numbers right now.

這為我們提供了一種針對硬體的最佳化設定。因此，我們目前沒有從數據中看到關稅的任何實際直接影響。

Also just to emphasize, we have multiple lines, right, not just the one you mentioned, so.

另外要強調的是，我們有多條線路，對，不只是你提到的那一條，所以。

Jay Vleeschhouwer, Griffin Securities.

Jay Vleeschhouwer，格里芬證券。

Anirudh, what would you say are the two or three most important technical or product enhancement deliverables this year that may have already occurred or yet to occur this year, even if not with an immediate impact on revenue, but in terms of enhancing your overall technical capabilities or positioning?

Anirudh，您認為今年最重要的兩三個技術或產品增強交付成果是什麼？這些交付成果可能已經發生或尚未發生，即使不會對收入產生直接影響，但可以增強您的整體技術能力或定位嗎？

And then related to that, maybe you could talk about some of the investments you're making in some critical technology areas that may not get a lot of attention, one being, for example, physical verification -- your physical verification, maybe the next-generation silicon for hardware, Gen four ? And then lastly, maybe talk about some of the incremental investments you're making in your acquired simulation solver products.

然後與此相關，也許您可以談談您在一些關鍵技術領域進行的一些投資，這些領域可能不會引起太多關注，例如物理驗證——您的物理驗證，可能是用於硬體的下一代矽片，第四代？最後，也許可以談談您在所收購的模擬求解器產品中所做的一些增量投資。

Jay, that's a great question. And it's a multipart answer. But just to highlight a few key things. I mean, first of all, before I get into specifics, I mean you know this, but just to emphasize, we have a massive investment in R&D, right? I mean like 35% to 40% of our revenue is going to R&D.

傑伊，這個問題問得非常好。這是一個由多部分組成的答案。但只是想強調幾個關鍵的事情。我的意思是，首先，在我講具體細節之前，我的意思是你知道這一點，但只是為了強調，我們在研發方面投入了大量資金，對嗎？我的意思是我們的收入的 35% 到 40% 將用於研發。

So at a given point, we always have a lot of exciting projects that are ongoing. In terms of a few themes that are critical, so of course, AI is supercritical and applying AI to the chip and system design process, I think we are seeing a lot of very good results. Actually, one thing on something like Cerebrus or AI-driven design tools, at this point we expect that the majority of -- more than 50% of the designs are now AI-enabled. Our tools are AI-enabled in those designs. And the others, I'm sure, will catch up as well.

因此，在特定時間點，我們總是有很多令人興奮的項目正在進行中。就一些至關重要的主題而言，人工智慧當然是至關重要的，將人工智慧應用於晶片和系統設計過程，我認為我們看到了許多非常好的結果。實際上，關於 Cerebrus 或 AI 驅動的設計工具，我們目前預計大多數（超過 50％）的設計現在都支援 AI。我們的工具在這些設計中支援人工智慧。我相信其他人也會趕上。

So -- but AI still has a long ways to go with all these new agentic features, and you will see us talk more about it throughout the year. So that's the number one thing I would like to highlight, and we have these five major platforms and all kinds of new models can be used in those.

所以——但是人工智慧在所有這些新的代理功能方面還有很長的路要走，你會看到我們全年都對此進行更多討論。這是我想強調的第一件事，我們有這五個主要平台，各種新模型都可以在這些平台上使用。

The second thing, which is also a pretty broad team, which makes sure we always state head is 3D-IC, okay? I mean 3D-IC is a big theme, another theme for next 5, 10 years. And you can see that in Blackwell. But right now, the packages can handle like three times reticle size, but they could easily go to ten times reticle size in the next few years.

第二件事，這也是一個相當廣泛的團隊，確保我們始終聲明頭部是 3D-IC，好嗎？我的意思是 3D-IC 是一個大主題，也是未來 5 年、10 年的另一個主題。你可以在布萊克威爾身上看到這一點。但目前，這些封裝可以處理三倍的光罩尺寸，但在未來幾年內它們很容易達到十倍的光罩尺寸。

So there's massive investment in 3D-IC in terms of the tool flow, in terms of analysis of these complex system, thermal and electromagnetic warpage analysis. Of course, IP that is needed for high-speed SerDes and UCI and other things. So 3D IC remains as a critical focus for both our chip business, system business and IP business.

因此，在工具流程、複雜系統分析、熱和電磁翹曲分析方面，對 3D-IC 進行了大量的投資。當然，高速 SerDes 和 UCI 等都需要 IP。因此，3D IC 仍然是我們晶片業務、系統業務和 IP 業務的重點。

And last thing is like, I think -- I think Gary asked earlier, is this rich hardware set, this bottom layer of the cake with ARM CPUs, x86 CPUs, Palladium hardware and GPUs. And we'll have some more developments next week at Cadence Live about that part. This hardware software co-optimization, I think, is going to be critical for our industry. So those are the three things I would highlight as top critical R&D focus area.

最後一件事是，我認為——我認為 Gary 之前問過，這是豐富的硬體集，蛋糕的底層有 ARM CPU、x86 CPU、Palladium 硬體和 GPU。下週我們將在 Cadence Live 上發布有關該部分的更多進展。我認為，這種硬體軟體的共同優化對於我們的產業至關重要。因此，我要強調的是作為最關鍵研發重點領域的三個面向。

Jason Celino, KeyBanc Capital Markets.

KeyBanc 資本市場 Jason Celino。

Great. Keeping to this IP topic, I think you planned to acquire ARM Artisan IP asset. I know this hasn't closed yet. But when we think about the foundational IP market, just curious why hasn't this been a bigger focus in the past and what's changed to make this opportunity more attractive? And then just to clarify with John, this pending acquisition is not included in guidance.

偉大的。繼續這個 IP 主題，我認為您計劃收購 ARM Artisan IP 資產。我知道這還沒結束。但是，當我們考慮基礎 IP 市場時，我們很好奇為什麼過去沒有對它給予更多關注，以及發生了哪些變化使得這個機會更具吸引力？然後只是想向約翰澄清一下，這項即將進行的收購不包括在指導中。

Yes. So let me take the second part first, Jason, just to clarify that, yes, it's not in our current guide. It's not closed yet. So we haven't put that into the guidance.

是的。因此，傑森，讓我先討論第二部分，只是為了澄清一下，是的，它不在我們目前的指南中。還沒關門。所以我們還沒有將其納入指導。

And then, yes, I was actually pleased with this partnership and acquisition. Arm Foundation IP business has a very rich history, as you know, we started with Artisan a long time ago and has very good credibility in the market. And also as our IP position improves with our kind of performance over the last two years, we also have to expand our portfolio, and I'm pleased with the performance and PPA of the current portfolio, but we have to fill in the gaps in our portfolio and that's the reason we for example, acquired Rambus' IP business and then acquired Secure-IC a few months ago.

是的，我確實對這次合作和收購感到高興。Arm Foundation IP 業務有著非常豐富的歷史，正如您所知，我們很久以前就與 Artisan 合作起步，在市場上享有非常好的信譽。而且隨著我們的 IP 地位在過去兩年隨著我們的業績而改善，我們也必須擴大我們的產品組合，我對當前產品組合的業績和 PPA 感到滿意，但我們必須填補我們產品組合中的空白，這就是我們收購 Rambus 的 IP 業務並在幾個月前收購 Secure-IC 的原因。

And this is a critical piece. Foundation -- and it is becoming more and more critical versus five, seven years ago. Foundation IP became a lot more critical, especially in the interaction with the software, with place and route. Also, the there is a massive, as you know, foundry build-out, not just -- there are multiple foundries in multiple countries, and they all require Foundation IP. So that's the other reason I feel that Foundation IP is much more critical now.

這是至關重要的部分。基金會——與五、七年前相比，它變得越來越重要。基礎 IP 變得更加重要，特別是在與軟體、位置和路線的互動中。此外，如您所知，代工廠的建設規模龐大，不僅僅是——多個國家有多家代工廠，它們都需要 Foundation IP。所以這也是我認為 Foundation IP 現在更為重要的另一個原因。

So it's a combination of Foundation IP being more critical. I need to broaden our IP portfolio work with our software tools and this opportunity came along to acquire a leading product in that space because we didn't want to enter the space with not a good product, but ARM's Artisan business has a good history grade product. So we -- we hope to finish that acquisition later in the year and talk to our customers about that.

因此，基礎 IP 的組合更為重要。我需要利用我們的軟體工具來拓展我們的 IP 產品組合，而這次機會就是收購該領域的領先產品，因為我們不想帶著不好的產品進入該領域，但 ARM 的 Artisan 業務擁有良好的歷史級產品。因此，我們希望在今年稍後完成收購，並與我們的客戶討論此事。

Joe Quatrochi, Wells Fargo.

富國銀行的喬‧夸特羅奇 (Joe Quatrochi)。

You mentioned that you were developing a full stack Agentic AI solution. I was wondering if you could talk about just the areas of the design workflow you're targeting first with that?

您提到您正在開發全端 Agentic AI 解決方案。我想知道您是否可以先談談您所針對的設計工作流程領域？

Great question. I mean it could be applied to almost all the design process. But sums that are super critical in terms of need, of course, verification is always critical because verification of course, we have great products with both hardware and software, but you never know when you are done with verifications, it's like an NP-complete problem.

好問題。我的意思是它可以應用於幾乎所有的設計過程。但從需求角度來說，總結起來非常關鍵，當然，驗證始終是至關重要的，因為驗證當然，我們有硬體和軟體的優秀產品，但你永遠不知道何時完成驗證，這就像一個 NP 完全問題。

And so if you really look at now our SimAI offering, which is -- works with our logic simulator can improve performance by 5x to 20x, we have all these several customer endorsement -- public consumer endorsements for applying CMAI to verification.

因此，如果您真的看看我們現在提供的 SimAI 產品，它與我們的邏輯模擬器配合使用可以將效能提高 5 倍到 20 倍，我們獲得了所有這些客戶的認可——將 CMAI 應用於驗證的公共消費者認可。

So I think verification to me is a great area for Agentic AI and then implementation is always a good area like things like Cerebrus can be further enhanced and you will see more and more developments from us digital design and Agentic AI pretty soon. So I would say that's the second area.

因此我認為驗證對於 Agentic AI 來說是一個很好的領域，而實施始終是一個很好的領域，例如 Cerebrus 可以進一步增強，並且您很快就會看到我們數位設計和 Agentic AI 的越來越多的發展。所以我想說這是第二個領域。

And then one area that traditionally has not seen that much automation but is becoming more and more critical is package and PCB design, Allegro, an Allegro X and Allegro X AI. Actually I'm very optimistic that as pack design becomes super critical and package is the critical thing in a 3D-IC system, it's 3D-IC is another name for system in a package. So there is a more need for automation in packaging and automation in PCB design. So I'm actually pretty optimistic that AI can finally provide automation in packaging.

傳統上自動化程度不高但變得越來越重要的一個領域是封裝和 PCB 設計、Allegro、Allegro X 和 Allegro X AI。實際上，我非常樂觀地認為，隨著封裝設計變得非常關鍵，並且封裝是 3D-IC 系統中的關鍵，3D-IC 是封裝系統的另一個名稱。因此，封裝自動化和PCB設計自動化的需求越來越大。因此，我實際上非常樂觀地認為人工智慧最終可以實現包裝自動化。

And then I would say, analog migration, another big area and analogs like art plus science, but AI can also help in analog. So I think these are the three or four key areas that we will work on. But the scope of Agentic AI is throughout the -- our product spectrum.

然後我想說，模擬遷移是另一個大領域，類似於藝術加科學的模擬，但人工智慧也可以在模擬方面提供幫助。所以我認為這些是我們將要努力的三到四個關鍵領域。但 Agentic AI 的範圍遍佈我們的產品範圍。

Joe Vruwink, Baird.

喬·弗魯溫克，貝爾德。

This might be an overly simplistic question relative to those before me. But your commentary and guidance strikes me as having greater visibility into strength during the second half of this year than maybe what's apparent just in our seat when analyzing the size of the 1Q beat or even your quarter-end backlog value. Can you maybe just speak to some of the things that changed for the better in the last 60 to 90 days that make you feel better about the 2025 outlook?

與我之前的問題相比，這可能是一個過於簡單的問題。但是，您的評論和指導讓我對今年下半年的實力有了更大的了解，而不是僅僅在我們分析第一季業績超出預期的規模或甚至季度末積壓訂單價值時所看到的情況。您能否談談過去 60 到 90 天內發生的一些好轉，讓您對 2025 年的前景感到更加樂觀？

Yes, Joe. I'll take that one. The -- yes, I mean, we're tracking ahead of our original forecast for 2025 with solid Q1 results. We only gave you the guide a few weeks ago for the year. So not a huge amount has changed. I think we have had a bit more strength in our recurring revenue. And of course, as you know, recurring revenue -- any beat in recurring revenue feeds through the rest of the year. So a small beat in Q1 kind of follows through in Q2, Q3 and Q4.

是的，喬。我要那個。是的，我的意思是，我們憑藉穩健的第一季業績領先於我們對 2025 年的最初預測。我們僅在幾週前為您提供了今年的指南。因此變化並不大。我認為我們的經常性收入更加強勁了。當然，正如你所知，經常性收入——任何超出預期的經常性收入都會影響今年剩餘時間的收入。因此，第一季的小幅上漲將延續到第二季​​、第三季和第四季。

And we were, like I said, ahead of the original forecast, even with including building in an expectation that -- that will pick up some slightly higher expenses on the tariff side. But we're still resilient on the tariff side, we felt it was right to raise the price.

正如我所說的那樣，我們的業績超出了最初的預測，甚至包括了這樣一種預期——關稅方面的支出將略有增加。但我們在關稅方面仍然有韌性，我們認為提高價格是正確的。

Joshua Tilton, Wolfe Research.

約書亞‧蒂爾頓 (Joshua Tilton)，沃爾夫研究公司。

I always love it when the guy before you kind of steals your question. But I guess, just to put a finer point on things, I think Anirudh, you mentioned that the guidance still assumes that China is flat for the year. You guys are being prudent. I guess my question to that is, I understand there was an easier comp in 1Q, but year-to-date, are you guys seeing anything that suggests that China should actually grow flat for the year?

我總是喜歡你前面的人搶先提出你的問題。但我想，只是為了更詳細地說明事情，我想阿尼魯德，你提到該指導仍然假設中國今年的經濟將保持平穩。你們很謹慎。我的問題是，我知道第一季的同比數據比較容易，但是從今年迄今為止的情況來看，你們是否看到任何跡象表明中國今年的經濟增速應該持平？

Or I guess another way of saying it is the decision to keep China flat for the year and the guide prudence or more along the lines of your expectations of how things will shake out by the end of the year?

或者我想換句話說，這是決定今年保持中國經濟平穩運行，並採取謹慎的指導方針，或者更符合您對年底情況如何發展的預期？

Yes, Josh, I'll take that one as well. I think it's more prudent that we thought it's wise to remain prudent in the current macro environment. I think we're very resilient in this environment. I mean, a company with high gross margin and essential products should do well in this type of environment.

是的，喬希，我也要這個。我認為我們認為在當前宏觀環境下保持謹慎是更明智的。我認為我們在這種環境下具有很強的適應力。我的意思是，一家擁有高毛利率和必需產品的公司在這種環境下應該會表現良好。

But the bookings were really solid in Q1. We had stronger bookings than expected. Kind of fed into with more recurring revenue bookings as well. But -- and that fell into us tracking ahead of the original forecast for 2025. So we thought the right thing to do was raise the guides.

但第一季的預訂量確實很穩定。我們的預訂量超出了預期。這也與更多經常性收入預訂有關。但是——這已經超出了我們最初對 2025 年的預測。因此我們認為正確的做法是提高指南。

Siti Panigrahi, Mizuho.

瑞穗的 Siti Panigrahi。

Anirudh, if you see a lot of news going around, we saw DeepSeek early this year. Wondering how that's impacting the design velocity or even how you see the compute capacity will change. But at the same time also, we are seeing hyperscaler trying to push their data center build out. Are you seeing any kind of impact there? So would love to hear your views on this.

Anirudh，如果你看到很多新聞，我們在今年年初看到了 DeepSeek。想知道這會如何影響設計速度，甚至您認為運算能力會如何變化。但同時，我們也看到超大規模企業正努力推動其資料中心的建置。您看到那裡有什麼影響嗎？所以很想聽聽你對此的看法。

Yes, Siti, that's a great question. I mean this is a big question. Let me try to -- I mean it's a critical thing for the whole industry. I mean, first of all, I mean, when I talk to customers, DeepSeek is one advancement. I mean a lot of customers expect multiple advancements like this.

是的，Siti，這是一個很好的問題。我的意思是這是一個大問題。讓我嘗試一下——我的意思是這對整個行業來說都是一件至關重要的事情。我的意思是，首先，當我與客戶交談時，DeepSeek 是一項進步。我的意思是很多客戶都期待這樣的多項進步。

The AI will get more and more efficient as -- and this happened if you look at even EDA's history, right, when AI started off, it was more like dense competition. And if you look at EDA history in the '70s, we started off with dense competition, and then we have Sparse computing, we have partitioning, we have hierarchy and latency, they're all cancer computer science methods that will be applied to AI, okay?

人工智慧將變得越來越有效率——如果你回顧 EDA 的歷史，你會發現當人工智慧剛開始時，競爭非常激烈。如果你回顧 70 年代的 EDA 歷史，我們一開始是密集競爭，然後我們有了稀疏計算、分區、層次結構和延遲，它們都是將應用於人工智慧的癌症計算機科學方法，好嗎？

So I believe that there will be multiple deep seek moments, not just one DeepSeek moment. But at the same time, the AI will get more and more prevalent in its use. And this happened in CPUs. I expect same thing to happen with AI and GPUs. So in terms of when we talk to customers.

因此我相信將會有多個深度搜尋時刻，而不僅僅是一個DeepSeek時刻。但同時，人工智慧的應用將變得越來越普遍。這發生在 CPU 中。我預計 AI 和 GPU 也會發生同樣的事情。就我們與客戶交談而言。

Now first, I have to remind you that we are on the design side, not on the production side. But in terms of design, we only see acceleration of more and more designs to do AI even on the data center side. And of course, NVIDIA is a great partner of ours, and I think NVIDIA is doing exceptionally well, but also a lot of the other hyperscalers, all the four or five major US hyperscalers and then companies in other countries are also investing heavily in their own designs, and we are glad to work with them.

首先，我必須提醒您，我們處於設計方面，而不是生產方面。但在設計方面，我們只看到越來越多的設計在加速實現人工智慧，甚至在資料中心方面也是如此。當然，NVIDIA 是我們的重要合作夥伴，我認為 NVIDIA 做得非常好，但許多其他超大規模企業，所有四到五家美國主要超大規模企業以及其他國家的公司也在大力投資自己的設計，我們很高興與他們合作。

And then I mentioned before, the physical AI part of it, which is cars, planes and drones, is also there's a lot of design activity. Now even though the car market may not be as strong, but that's for current products, but they are designing for future products, which will be AI-enabled. And then the other thing, as AI gets more efficient, I think there will be more inference use. So there's a lot of design of insurance chips to support that.

我之前提到過，實體人工智慧部分，也就是汽車、飛機和無人機，也有很多設計活動。現在，儘管汽車市場可能不那麼強勁，但這是針對當前產品而言的，但他們正在為未來的產品進行設計，這些產品將支援人工智慧。另一件事是，隨著人工智慧變得越來越高效，我認為推理的使用將會越來越多。因此有很多保險晶片的設計來支持這一點。

So right now in terms of design activity, I feel pretty good. And like I was mentioning before, even on the data center side, the 3D-IC is going to cause a massive change in complexity. Right now, we have Blackwell with like two kind of big dies with HPM, but the packaging road map is very aggressive by all foundries.

所以就目前設計活動而言，我感覺非常好。正如我之前提到的，即使在資料中心方面，3D-IC 也會導致複雜性發生巨大變化。目前，我們擁有 Blackwell 兩種採用 HPM 的大型晶片，但所有代工廠的封裝路線圖都非常積極。

So I think that is also going to drive new and new AI designs with putting more kind of big chips in packages and that should also drive AI performance and efficiency. So right now, we don't see any big change in terms of the design activity for AI.

因此，我認為這也將推動新的 AI 設計，將更多類型的大晶片放入封裝中，這也將推動 AI 的性能和效率。因此目前，我們沒有看到人工智慧設計活動方面有任何重大變化。

Nay Soe Naing, Berenberg.

奈蘇奈恩，貝倫貝格。

Minus on your SG&A segment, please. a very impressive quarter in Q1. And I was just trying to understand if you could maybe share where the stat is coming on, especially when we put it against the product market and your peers in the last three years, you've been growing at 2x to 3x what your competitors have been able to do. So I was wondering if there are any competitive market wins that's driving your performance, if not, what are the reasons, please?

請減去您的銷售、一般及行政費用部分。第一季的表現非常令人印象深刻。我只是想知道您是否可以分享統計數據，特別是當我們將其與產品市場和過去三年的同行進行比較時，您的成長率是競爭對手的 2 倍到 3 倍。所以我想知道是否有任何競爭市場的勝利推動了你的業績，如果沒有，原因是什麼？

And maybe if I could squeeze in a part -- a second part of this question. The solid performance we've had in the past, does it -- any chance at all create any headwinds for you or tough comp for you going forward?

也許我可以擠出一點時間來回答這個問題的第二部分。我們過去的出色表現是否會為您帶來任何不利因素或為您未來的發展帶來困難？

Thanks for the question. And you're right to raise the comps, because I think one of the benefits for Q1 is that Q1 '24 was a pretty easy comp for us on the SG&A side. BETA performed well in Q1 '25, but we only acquired that in the middle of the year last year. So I think the first time BETA turned up in our results was Q3 '24.

謝謝你的提問。你提高可比較價格是對的，因為我認為第一季的一個好處是，2024 年第一季在銷售、一般及行政管理費用方面對我們來說是一個相當容易的可比價格。BETA 在 25 年第一季表現良好，但我們只是在去年年中才收購了它。所以我認為 BETA 第一次出現在我們的結果是在 24 年第三季。

And then BETA is providing a huge amount of pull-through business for Cadence. So Cadence products are being sold alongside BETA products, and that's really helping us for growth. We're expanding reach to the longer tail customers through our e-commerce model. And we've seen strength in automotive and aero and defense. And I don't know, Anirudh, do you want to add anything about BETA?

然後 BETA 為 Cadence 提供了大量的拉動業務。因此，Cadence 產品與 BETA 產品一起銷售，這確實有助於我們的成長。我們正在透過電子商務模式擴大對長尾客戶的覆蓋範圍。我們也看到了汽車、航空和國防領域的優勢。我不知道，Anirudh，您是否想補充有關 BETA 的任何資訊？

BETA is performing great and also it has a pull-through to other parts of our portfolio. One thing we are always confident and I think you've seen some benchmarks in the industry in the past that our R&D products are solid. They are rewritten. They are -- they have no accuracy loss. They have much better performance because of massive parallelism whether it is on CPUs or GPUs. So we are always confident on our product side, and we continue to innovate there.

BETA 表現優異，並且對我們投資組合的其他部分具有拉動作用。有一件事我們始終充滿信心，我想您過去也看到了業界的一些基準，那就是我們的研發產品是可靠的。它們被重寫了。是的——它們沒有準確性損失。無論是在 CPU 還是 GPU 上，它們都具有更佳的效能，因為具有大規模並行性。因此，我們始終對我們的產品充滿信心，並將繼續在那裡進行創新。

I think what BETA helps is to increase the reach of those products because they are working with all the auto companies. And so not only we can benefit from performance of beta but also can sell our other products to those customers. And then we also improved our cloud offering and go-to-market with indirect channel. So I think you're seeing some benefits of that, better go-to-market operation on the system side.

我認為 BETA 的幫助在於擴大這些產品的覆蓋面，因為他們與所有汽車公司都有合作。因此，我們不僅可以從測試版的性能中受益，還可以向這些客戶銷售我們的其他產品。然後，我們還改進了我們的雲端產品並透過間接管道進入市場。所以我認為你看到了其中的一些好處，即係統方面的更好的市場運作。

Ruben Roy, Stifel.

魯本·羅伊（Ruben Roy），Stifel。

Anirudh, I had a similar question to Jason's on IP. And you touched on this in your answer to that question. But just wanted to make sure I understood in terms of sort of these, I would say, accelerated pace of recent tuck-ins on the IP side. It sounds like they're opportunistic rather than customer-driven, wanted to make sure I got that right.

Anirudh，我對 IP 有與 Jason 類似的疑問。您在回答該問題時也提到了這一點。但只是想確保我理解這些方面，我想說，最近 IP 方面的擴張步伐加快了。聽起來他們是機會主義者而不是客戶驅動的，想確保我理解正確。

And then the second part of the question is just maybe if you can update sort of how you're thinking about IP in terms of longer-term growth from that segment given all that's going on with foundry and customers, et cetera.

然後問題的第二部分是，考慮到代工廠和客戶等方面的現狀，您是否可以更新一下您對 IP 的看法，從該領域的長期成長角度來看。

Yes. Great question. So I think in terms of IP, there are multiple factors. So one is our performance is better. The portfolio we had, especially the PPA, power performance in the area for TSMC, the most advanced foundry.

是的。好問題。所以我認為就 IP 而言，有多種因素。所以一是我們的表現比較好。我們擁有的產品組合，尤其是 PPA，為最先進的代工廠台積電提供了該領域的動力性能。

The IP is just much better quality and performance. So and that's driving customer demand. So the customers want us to want get IP from us and want us to have a bigger portfolio, okay. So as a result, we do, of course, invest organically, and we will do in acquisition if we see a good fit, okay? But it's both a combination of what is available, but the customers are definitely driving us to broaden our portfolio. So that's one.

IP 的品質和性能要好得多。這就是推動客戶需求的因素。所以客戶希望我們從我們這裡獲得 IP，並希望我們有更大的產品組合，好吧。因此，我們當然會進行有機投資，如果我們發現合適的話，我們也會進行收購，好嗎？但這都是現有資源的組合，但客戶肯定會推動我們擴大產品組合。這就是其中之一。

Second thing is there is more and more foundry. So we are glad with our partnership with TSMC, but we highlighted even this time like Rapidus making massive investments. And of course, Samsung and Intel, and we are glad to. So there are at least four major foundries at advanced node along with all the foundries like Global and UMC at more mainstream nodes. So I think the foundry ecosystem is rich and that provides more opportunities in IP.

第二件事是代工廠越來越多。因此，我們很高興與台積電合作，但我們強調，即使是這次，Rapidus 也進行了大規模投資。當然還有三星和英特爾，我們很高興。因此，在先進節點上至少有四個主要代工廠，此外還有 Global 和 UMC 等在更主流節點上的代工廠。所以我認為代工生態系統非常豐富，並且在 IP 方面提供了更多機會。

And then the third thing, of course, is AI and 3D-IC and all this die-to-die IT. So for all these three reasons, I'm optimistic. We are performing well. We are increasing our portfolio. The market wants more IP from us. So I expect IP to -- of course, we had a good year last year, we had a very good quarter, Q1. So I expect IP to grow better than Cadence average going forward.

第三件事當然是人工智慧 (AI)、3D-IC 以及所有這些 die-to-die IT。因此，基於這三個原因，我感到樂觀。我們表現良好。我們正在增加我們的投資組合。市場希望我們提供更多知識產權。因此我預計 IP 將會——當然，去年我們的業績表現不錯，第一季的業績也非常好。因此我預計 IP 未來的成長速度將高於 Cadence 的平均值。

Blair Abernethy, Rosenblatt Securities.

羅森布拉特證券公司的布萊爾‧阿伯內西 (Blair Abernethy)。

Nice quarter. Anirudh, just wondering if you could give us sort of high-level thoughts on the data center digital twins. You bought future facilities back in July '22, there seems to be getting some traction out there. I just want to see how you feel about the size of the TAM, the size of the opportunity there with that asset and sort of are there expansion opportunities beyond data centers?

不錯的季度。Anirudh，我只是想知道您是否可以給我們一些關於資料中心數位孿生的高層想法。您在 2022 年 7 月購買了未來設施，目前看來已經取得了一些進展。我只是想看看您對 TAM 的規模、該資產帶來的機會規模有何看法，以及除了資料中心之外是否存在擴展機會？

Absolutely. The two areas that I have been very excited about, especially on the system side, one is, which is very close to the package, like I mentioned, thermal, electromagnetics, warpage because there's a massive transformation right next to the chip. And we are in a very strong position with Allegro and all the analysis tools. The second area is all the way at the data center level.

絕對地。我對兩個領域感到非常興奮，特別是在系統方面，一個是與封裝非常接近的領域，就像我提到的，熱、電磁、翹曲，因為晶片旁邊有一個巨大的轉變。憑藉 Allegro 和所有分析工具，我們處於非常有利的地位。第二個領域是資料中心層面。

That's why we acquired future facilities. And actually, it had a very strong Q1 and we also have a great partnership with NVIDIA there. NVIDIA is designing their data centers with our Cadence Reality Digital Twin and also working with joint partners as we go to market with them together on Cadence Reality.

這就是我們收購未來設施的原因。實際上，它的第一季表現非常強勁，我們與 NVIDIA 也建立了良好的合作關係。NVIDIA 正在使用我們的 Cadence Reality Digital Twin 設計他們的資料中心，並且還與合作夥伴一起在 Cadence Reality 上與他們一起將產品推向市場。

And that opportunity is huge. I mean, even we apply it to our own IT group, for example, we have our own data centers, right? So the product is general enough to be used in any data center and we saw like a 10% improvement in power, which is huge. And because typically, that's the area that has not seen a lot of analysis and automation, because it's typically done in like a typical construction kind of work whereas chip design is very analytical mathematical. So that area needs more kind of optimization and simulation.

這個機會是巨大的。我的意思是，即使我們將它應用到我們自己的 IT 團隊，例如，我們有自己的資料中心，對嗎？因此，該產品足夠通用，可以在任何資料中心使用，我們看到功率提高了 10%，這是一個巨大的進步。而且通常情況下，這個領域還沒有進行大量的分析和自動化，因為它通常是在典型的建築工作中完成的，而晶片設計是非常分析性的數學。因此該領域需要更多種類的最佳化和模擬。

So I am optimistic. And then the way Cadence Reality works, there is a design side of it, in which we help customers design the data center. And then there's the operations side of it, in which we can analyze on the fly, use the Digital Twin to optimize their operation. So if you include both sides, I mean, the opportunity is huge. So -- but we are still kind of prudent in our assumptions like we do in everything, but I'm very optimistic about these two extremes.

所以我很樂觀。Cadence Reality 的工作方式是，它有一個設計方面，我們幫助客戶設計資料中心。然後是操作方面，我們可以動態分析，使用數位孿生來優化其操作。所以如果把雙方都考慮進來，那麼機會是巨大的。所以——但我們在假設方面仍然很謹慎，就像我們在所有事情中所做的那樣，但我對這兩個極端非常樂觀。

One is all the way to data center and second is all the way very close to the chip on the package. And that's where you see us investing yes.

一是一直到資料中心，二是一直到非常靠近封裝上的晶片。是的，這就是我們所投資的地方。

I will now turn the call back over to Anirudh Devgan for closing remarks.

現在我將把電話轉回給 Anirudh Devgan 做結束語。

Thank you all for joining us this afternoon. It's an exciting time for Cadence as our broad portfolio and product leadership ideally positions us to maximize the growing opportunities in the semiconductor and system industry. And on behalf of our employees and our Board of Directors, we thank our customers, partners and investors for your continued trust and confidence in Cadence.

感謝大家今天下午加入我們。對於 Cadence 來說，這是一個令人興奮的時刻，因為我們廣泛的產品組合和產品領導力使我們能夠最大限度地利用半導體和系統行業的成長機會。我謹代表我們的員工和董事會感謝我們的客戶、合作夥伴和投資者對 Cadence 的持續信任和信心。

Thank you for participating here today's Cadence's first quarter 2025 earnings conference call. This concludes today's call. You may now disconnect. Goodbye.

感謝您參加今天的 Cadence 2025 年第一季財報電話會議。今天的電話會議到此結束。您現在可以斷開連線。再見。