益華電腦 (CDNS) 2022 Q4 法說會逐字稿

Good afternoon. My name is Julianne, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence Fourth Quarter and Fiscal Year 2022 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.

下午好。我叫 Julianne，今天我將擔任你們的會議運營商。此時，我想歡迎大家參加 Cadence 第四季度和 2022 財年收益電話會議。 （操作員說明）我現在將電話轉給 Cadence 投資者關係副總裁 Richard Gu。請繼續。

Thank you, operator. I would like to welcome everyone to our Fourth Quarter of Fiscal Year 2022 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.

謝謝你，運營商。歡迎大家參加我們的 2022 財年第四季度收益電話會議。今天，總裁兼首席執行官 Anirudh Devgan 加入了我的行列；高級副總裁兼首席財務官 John Wall。本次電話會議的網絡直播和今天準備好的評論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. 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, 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 them. In addition, we will present certain non-GAAP measures, which 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.

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業績的展望。由於存在風險和不確定性，實際結果可能與今天討論中預測或暗示的結果存在重大差異。有關可能導致實際結果不同的因素的信息，請參閱我們向美國證券交易委員會提交的文件，包括我們最近的 10-K 和 10-Q 表格，以及今天的收益發布。本次電話會議期間的所有前瞻性陳述均基於截至今天我們可獲得的估計和信息，我們不承擔更新這些陳述的任何義務。此外，我們將介紹某些非 GAAP 指標，不應將其與 GAAP 結果分開考慮或作為其替代品。 GAAP 與非 GAAP 措施的對賬包含在今天的收益發布中。

Today's earnings release for the fourth quarter of fiscal 2022, related financial tables and CFO commentary are also available on our website. (Operator Instructions) Now I'll turn the call over to Anirudh.

今天發布的 2022 財年第四季度收益報告、相關財務表格和 CFO 評論也可在我們的網站上找到。 （操作員說明）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 record results for 2022 as we exceeded our guidance yet again, achieving 19% revenue growth and over 40% non-GAAP operating margin. Cadence's innovative solutions are essential and especially relevant in the current environment, enabling customers to achieve their increasingly challenging design goals.

謝謝你，理查德。大家下午好，感謝您今天加入我們。我很高興地報告，Cadence 在 2022 年取得了創紀錄的業績，我們再次超出了我們的指導，實現了 19% 的收入增長和超過 40% 的非 GAAP 營業利潤率。 Cadence 的創新解決方案是必不可少的，在當前環境中尤為重要，可幫助客戶實現其日益具有挑戰性的設計目標。

Secular mega trends, such as 5G, hyperscale computing and AI/ML that are driving sustained, long-term semiconductor and system growth, remain unchanged. Amid ongoing macroeconomic uncertainty, companies continue making significant investment in their next-generation products, resulting in robust design activity. We expect our pioneering solutions to continue fueling broad-based business momentum in 2023, driving strong revenue growth and profitability. John will provide more details in a moment.

5G、超大規模計算和 AI/ML 等推動半導體和系統持續長期增長的長期大趨勢保持不變。在持續存在的宏觀經濟不確定性中，公司繼續對其下一代產品進行大量投資，從而產生了強勁的設計活動。我們預計我們的開創性解決方案將在 2023 年繼續推動廣泛的業務發展勢頭，推動強勁的收入增長和盈利能力。約翰稍後會提供更多細節。

Our Intelligent System Design strategy greatly broadens our total available market, and leading end-to-end EDA, IP, hardware and expanding system analysis portfolio uniquely position us to capture a wide range of market opportunities. During the year, we introduced 9 significant innovative products across all of our business groups, and we expect these to be key drivers of our future growth.

我們的智能係統設計戰略極大地拓寬了我們的總體可用市場，領先的端到端 EDA、IP、硬件和不斷擴展的系統分析產品組合使我們能夠抓住廣泛的市場機會。在這一年裡，我們在所有業務部門推出了 9 項重要的創新產品，我們預計這些將成為我們未來增長的主要驅動力。

The age of AI is upon us, and Cadence provides several groundbreaking, computational software-driven generative AI technologies at both the chip and system level unified by JedAI, our differentiated big data analytics platform. Our customers are seeing dramatic results with these solutions delivering highly optimized design and unprecedented efficiency gains. Additionally, by automating repetitive tasks and producing new ideas, our generative AI frees up engineers to focus on more advanced, high-value activities, opening up more opportunities for innovation.

AI 時代已經來臨，Cadence 在芯片和系統層面提供了多項突破性的、計算軟件驅動的生成 AI 技術，由我們差異化的大數據分析平台 JedAI 統一。我們的客戶通過這些提供高度優化的設計和前所未有的效率提升的解決方案看到了顯著的結果。此外，通過自動執行重複性任務和產生新想法，我們的生成式 AI 可以讓工程師騰出時間專注於更高級、高價值的活動，從而為創新開闢更多機會。

During the year, we also materially expanded our core EDA, IP and system solutions footprint at market-shaping customers. In Q2, we extended our collaboration with AMD to a far-reaching commitment to our innovative core EDA, hardware, design IP and system software solutions. In Q3, we deepened our partnership with BAE Systems across our core EDA and systems portfolio, including proliferation of our digital full flow and analog products, and a broad expansion of our PCB and multiphysics system analysis solutions.

在這一年中，我們還大幅擴展了核心 EDA、IP 和系統解決方案在市場塑造客戶中的足跡。在第二季度，我們將與 AMD 的合作擴展到對我們創新的核心 EDA、硬件、設計 IP 和系統軟件解決方案的深遠承諾。在第三季度，我們在我們的核心 EDA 和系統產品組合中加深了與 BAE Systems 的合作夥伴關係，包括我們數字全流程和模擬產品的擴散，以及我們 PCB 和多物理系統分析解決方案的廣泛擴展。

And in Q4, we broadened our relationship with a global leader in memory and storage solutions through an extensive proliferation of our custom, digital and system solutions. We also expanded our strategic partnership with a global leader in networking and telecommunications through their renewed commitment to our core EDA, IP and system solutions. In addition, we furthered our partnership with leading foundry, IP and cloud service providers and won 6 Open Innovation Platform Partner of the Year awards from TSMC.

在第四季度，我們通過廣泛擴展我們的定制、數字和系統解決方案，擴大了我們與內存和存儲解決方案全球領導者的關係。我們還通過重新承諾對我們的核心 EDA、IP 和系統解決方案的承諾，擴大了與網絡和電信領域全球領導者的戰略合作夥伴關係。此外，我們進一步加強了與領先晶圓代工、IP 和雲服務提供商的合作夥伴關係，並獲得了台積電頒發的 6 項年度開放創新平台合作夥伴獎。

Now let's talk about some of the product highlights for both Q4 and 2022. Our digital IC business finished another strong year with 17% revenue growth. Deployment of our digital full flow, delivering industry-leading quality of results at the most advanced nodes continue to accelerate with nearly 50 additional customers adopting it during the year. Our digital software is now deployed in all top 20 semiconductor companies. We are pleased with the accelerating growth of our front-end Genus and Joules tools and signoff products, such as Tempus and Quantus, complementing the broad proliferation of Innovus.

現在讓我們談談第四季度和 2022 年的一些產品亮點。我們的數字 IC 業務以 17% 的收入增長結束了又一個強勁的一年。我們數字全流程的部署，在最先進的節點上提供行業領先的結果質量，在這一年中繼續加速，近 50 家客戶採用它。我們的數字軟件現已部署在所有排名前 20 位的半導體公司中。我們很高興我們的前端 Genus 和 Joules 工具以及簽核產品（例如 Tempus 和 Quantus）的加速增長，補充了 Innovus 的廣泛擴散。

Our transformative Cadence Cerebrus AI-driven solution continued to deliver impressive PPA and productivity gains across a wide range of designs, resulting in broader adoption and accelerating proliferation. Among others, it is now deployed at 10 of the top 20 semiconductor companies, including 7 of the top 10 semis and at several major hyperscalers. In Q4, GUC successfully delivered an advanced HPC design and a CPU design using our digital full flow and Cadence Cerebrus on TSMC N5 process technology, delivering 8% reduced power and a 9% area improvement while significantly improving engineering productivity.

我們的變革性 Cadence Cerebrus AI 驅動解決方案繼續在各種設計中提供令人印象深刻的 PPA 和生產力提升，從而獲得更廣泛的採用並加速擴散。其中，它現在部署在前 20 大半導體公司中的 10 家，包括前 10 大半導體公司中的 7 家和幾家主要的超大規模公司。在第四季度，GUC 在台積電 N5 工藝技術上使用我們的數字全流程和 Cadence Cerebrus 成功交付了先進的 HPC 設計和 CPU 設計，功耗降低了 8%，面積增加了 9%，同時顯著提高了工程生產力。

In 2022, several market-shaping customers, including Intel, NVIDIA, Broadcom, Samsung and Renesas, shared their remarkable successes with Cadence Cerebrus at our CadenceLIVE user conferences. Escalating system and software bring-up complexities combined with relentless first-pass silicon requirements continued to drive strong demand for our essential functional verification solutions. Our verification business grew 28% year-over-year, fueled by secular trend for hardware, which had another record year.

2022 年，包括英特爾、英偉達、博通、三星和瑞薩在內的幾家塑造市場的客戶在我們的 CadenceLIVE 用戶大會上分享了他們使用 Cadence Cerebrus 取得的非凡成就。不斷升級的系統和軟件啟動複雜性與無情的首次通過矽要求相結合，繼續推動對我們基本功能驗證解決方案的強勁需求。在硬件長期趨勢的推動下，我們的驗證業務同比增長 28%，硬件再次創下歷史新高。

Our dynamic duo of Palladium Z2 and Protium X2 platforms providing best-in-class system verification and software bring-up solutions saw accelerated growth and strong momentum across mobile, hyperscale, HPC and increasingly, auto EV segments. Our hardware family added 30 new customers and over 160 repeat orders during the year. Due to the compelling value offered by common front-end compiler, demand for the pair greatly exceeded our expectations with more than 2/3 of the orders in the year, including both platforms.

我們動態的 Palladium Z2 和 Protium X2 平台組合提供一流的系統驗證和軟件啟動解決方案，在移動、超大規模、HPC 和越來越多的汽車 EV 領域實現了加速增長和強勁勢頭。我們的硬件系列在這一年中增加了 30 個新客戶和 160 多個重複訂單。由於通用前端編譯器提供的極具吸引力的價值，對這兩款產品的需求大大超出了我們的預期，全年超過 2/3 的訂單，包括這兩個平台。

Our new Cadence Verisium AI verification platform enables dramatic improvements in debug productivity. And in early production usage at several market-shaping customers, Verisium delivered up to a 30x improvement in efficient root/cause analysis. Our custom IC, Virtuoso and Spectre, franchise solutions tackle the toughest challenges in analog, mixed signal, RF design and circuit simulation. And as electrification and digital transformation trends gain momentum, they are becoming increasingly crucial to our customers.

我們新的 Cadence Verisium AI 驗證平台可顯著提高調試效率。在幾個市場塑造客戶的早期生產使用中，Verisium 在有效的根本/原因分析方面提供了高達 30 倍的改進。我們的定制 IC、Virtuoso 和 Spectre 特許解決方案解決了模擬、混合信號、RF 設計和電路仿真中最嚴峻的挑戰。隨著電氣化和數字化轉型趨勢的發展，它們對我們的客戶變得越來越重要。

Building on our market leadership, our custom IC revenue grew 13% year-over-year in 2022 with Virtuoso growth spurred by demand in advanced nodes, heterogeneous integration and the emerging silicon photonics segment. We added 200 new Virtuoso logos and more than 150 logos for Spectre with Spectre FX making strong headway in FastSPICE memory applications. Increasing usage of preconfigured IP blocks to reduce risk and time to market, coupled with our STAR IP portfolio, led to a strong year for our IP business, which grew 12% year-over-year in 2022.

憑藉我們的市場領導地位，我們的定制 IC 收入在 2022 年同比增長 13%，而 Virtuoso 的增長受到高級節點、異構集成和新興矽光子學領域需求的推動。我們為 Spectre FX 添加了 200 個新的 Virtuoso 徽標和 150 多個徽標，在 FastSPICE 存儲器應用中取得了長足的進步。增加預配置 IP 塊的使用以降低風險和縮短上市時間，再加上我們的 STAR IP 產品組合，使我們的 IP 業務在 2022 年同比增長 12%。

Demand was particularly strong in HPC, 5G and automotive segments with our silicon proven high-performance PCIe Gen 5, LPDDR5 and Ethernet interfaces helping secure key wins in advanced node designs. Our Tensilica DSP portfolio continued expanding its footprint in smart speakers and true wireless stereo headsets, imaging and machine learning applications. Rising system complexity and challenges stemming from the growing hyperconvergence of the electrical, mechanical and physical worlds are driving the strong need for a seamless platform solution across design, packaging, simulation and analysis.

HPC、5G 和汽車領域的需求特別強勁，我們經過矽驗證的高性能 PCIe Gen 5、LPDDR5 和以太網接口有助於確保在高級節點設計中取得關鍵勝利。我們的 Tensilica DSP 產品組合繼續擴大其在智能揚聲器和真正無線立體聲耳機、成像和機器學習應用方面的足跡。電氣、機械和物理世界日益超融合所帶來的系統複雜性和挑戰不斷增加，推動了對跨設計、封裝、仿真和分析的無縫平台解決方案的強烈需求。

Our System Design & Analysis business that is expanding our TAM beyond EDA continued its strong momentum, delivering 27% year-over-year growth. Industry interest in advanced packaging solutions notably spiked in 2022 with customers embracing our revolutionary Integrity 3D-IC solution, the industry's only comprehensive platform providing tightly integrated system planning, implementation and analysis technologies.

我們的系統設計和分析業務將我們的 TAM 擴展到 EDA 之外，繼續保持強勁勢頭，同比增長 27%。行業對先進封裝解決方案的興趣在 2022 年顯著飆升，客戶採用了我們革命性的 Integrity 3D-IC 解決方案，這是業界唯一提供緊密集成的系統規劃、實施和分析技術的綜合平台。

Our multiphysics portfolio comprising of leading electromagnetic, electrothermal, signal and power integrity and CFD solutions continued ramping strongly across multiple end markets. Our in-design analysis solutions had several significant wins with HPC and hyperscaler customers, while our CFD Fidelity platform proliferated with market-shaping aerospace and defense customers. During the year, Fidelity CFD's software meshing capabilities were chosen by Toyota Motor Europe to be their standard workflow for CFD preprocessing. And in numerous customer engagements, Optimality Explorer, the industry's first AI-driven, multidisciplinary system analysis and optimization solution, has demonstrated up to a 10x efficiency improvement in design space exploration, leading to faster time to results.

我們的多物理場產品組合包括領先的電磁、電熱、信號和電源完整性以及 CFD 解決方案，在多個終端市場繼續強勁增長。我們的設計分析解決方案在 HPC 和超大規模客戶中取得了多項重大勝利，而我們的 CFD Fidelity 平台則在塑造市場的航空航天和國防客戶中激增。在這一年中，Fidelity CFD 的軟件網格劃分功能被豐田汽車歐洲公司選為其 CFD 預處理的標準工作流程。在眾多客戶參與中，Optimality Explorer 是業界首個人工智能驅動的多學科系統分析和優化解決方案，它在設計空間探索方面的效率提高了 10 倍，從而縮短了獲得結果的時間。

Lastly, in keeping with our transition plan, Lip-Bu Tan has notified the Cadence Board that he will not seek reelection at our upcoming 2023 Annual Stockholders' Meeting in May. He will continue to serve as an advisor to me. We will return to an independent Chair structure with ML Krakauer becoming our next Board Chair following that meeting. The rest of the Board and I look forward to working closely with ML in her new role.

最後，根據我們的過渡計劃，Lip-Bu Tan 已通知 Cadence 董事會，他將不會在我們即將於 5 月舉行的 2023 年年度股東大會上尋求連任。他將繼續擔任我的顧問。在那次會議之後，我們將恢復獨立主席結構，ML Krakauer 將成為我們的下一任董事會主席。董事會其他成員和我期待在 ML 擔任新職務後與她密切合作。

In closing, we are pleased with our strong execution in 2022 and are thrilled by the business momentum and market opportunities ahead of us in 2023. Now I will turn it over to John to provide more details on the Q4 results and our 2023 outlook.

最後，我們對 2022 年的強勁執行感到高興，並對 2023 年我們面前的商業勢頭和市場機會感到興奮。現在我將把它交給約翰，提供有關第四季度業績和我們 2023 年展望的更多細節。

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that we exceeded all of our key financial and operating metrics for the fourth quarter and 2022. Robust customer design activity and demand for our strong technology portfolio continued to drive growth across all of our businesses. Cadence had an excellent 2022, and we began 2023 with a lot of confidence and strong momentum on the back of the stability and resilience you'd expect from a predominantly recurring revenue model.

謝謝，Anirudh，大家下午好。我很高興地報告，我們超過了第四季度和 2022 年的所有關鍵財務和運營指標。強勁的客戶設計活動和對我們強大技術組合的需求繼續推動我們所有業務的增長。 Cadence 在 2022 年表現出色，在以經常性收入為主的收入模式所帶來的穩定性和彈性的支持下，我們在 2023 年開始時充滿信心和強勁勢頭。

Here are some of the financial highlights from the fourth quarter and the year, starting with the P&L. Total revenue was $900 million for the quarter and $3.562 billion for the year. GAAP operating margin was 23.5% for the quarter and 30.1% for the year. Non-GAAP operating margin was 35.6% for the quarter and 40.3% for the year. GAAP EPS was $0.88 for the quarter and $3.09 for the year, and non-GAAP EPS was $0.96 for the quarter and $4.27 for the year.

以下是第四季度和今年的一些財務亮點，首先是損益表。本季度總收入為 9 億美元，全年總收入為 35.62 億美元。本季度 GAAP 營業利潤率為 23.5%，全年為 30.1%。本季度非美國通用會計準則營業利潤率為 35.6%，全年為 40.3%。本季度 GAAP 每股收益為 0.88 美元，全年為 3.09 美元，非 GAAP 每股收益本季度為 0.96 美元，全年為 4.27 美元。

Next, turning to the balance sheet and cash flow. Our cash balance was $882 million at year-end, while the principal value of debt outstanding was $750 million. Operating cash flow in the fourth quarter was $264 million and $1.24 billion for the full year. DSOs were 49 days, and we repurchased $1.05 billion worth of Cadence shares during the year.

接下來，轉向資產負債表和現金流量。年末我們的現金餘額為 8.82 億美元，而未償債務的本金價值為 7.5 億美元。第四季度運營現金流為 2.64 億美元，全年運營現金流為 12.4 億美元。 DSO 為 49 天，我們在這一年中回購了價值 10.5 億美元的 Cadence 股票。

Before I provide our outlook for Q1 and 2023, I'd like to share a few comments. Our most recent fiscal year ended on December 31, 2022. This fiscal year will also end on December 31 as we've now moved our fiscal year to a calendar year. Approximately 15% of our annual revenue for fiscal 2022 was upfront, with 85% recurring. At the midpoint of our 2023 revenue outlook, we're expecting a similar revenue mix for the year. And our outlook for 2023 assumes export control regulations remain substantially similar for the remainder of the year.

在我提供我們對第一季度和 2023 年的展望之前，我想分享一些評論。我們最近的財政年度於 2022 年 12 月 31 日結束。本財政年度也將於 12 月 31 日結束，因為我們現在已將財政年度移至日曆年。我們 2022 財年年收入的大約 15% 是預付的，85% 是經常性的。在我們 2023 年收入展望的中點，我們預計今年的收入組合會類似。我們對 2023 年的展望假設出口管制法規在今年剩餘時間內保持基本相似。

In our outlook for 2023, we expect revenue in the range of $4 billion to $4.06 billion, GAAP operating margin of 30.5% to 32%, non-GAAP operating margin of 40.5% to 42%, GAAP EPS in the range of $3.24 to $3.34, non-GAAP EPS in the range of $4.90 to $5, operating cash flow in the range of $1.3 billion to $1.4 billion and we expect to use approximately 50% of our free cash flow to repurchase Cadence shares in 2023. For Q1, we expect revenue in the range of $1 billion to $1.02 billion, GAAP operating margin in the range of 31% to 32%, non-GAAP operating margin of 41% to 42%, GAAP EPS in the range of $0.84 to $0.88 and non-GAAP EPS in the range of $1.23 to $1.27. And as usual, we published the 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.

在我們對 2023 年的展望中，我們預計收入在 40 億美元至 40.6 億美元之間，GAAP 營業利潤率為 30.5% 至 32%，非 GAAP 營業利潤率為 40.5% 至 42%，GAAP 每股收益在 3.24 美元至 3.34 美元之間，非 GAAP 每股收益在 4.90 美元到 5 美元之間，運營現金流在 13 億美元到 14 億美元之間，我們預計將在 2023 年使用大約 50% 的自由現金流來回購 Cadence 股票。對於第一季度，我們預計收入在 10 億美元到 10.2 億美元之間，GAAP 營業利潤率在 31% 到 32% 之間，非 GAAP 營業利潤率在 41% 到 42% 之間，GAAP 每股收益在 0.84 美元到 0.88 美元之間，非 GAAP 每股收益在 1.23 美元到 1.27 美元之間。和往常一樣，我們在投資者關係網站上發布了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分析和 GAAP 與非 GAAP 對賬。

In conclusion, I am pleased that we achieved double-digit revenue growth across all of our businesses, increased 3-year revenue CAGR into the mid-teens and I'm especially pleased that we continue to expand annual operating margins. As always, I'd like to thank our customers, partners and our employees for their continued support. And with that, operator, we will now take questions.

總之，我很高興我們所有業務都實現了兩位數的收入增長，將 3 年的收入複合年增長率提高到十幾歲左右，我特別高興的是我們繼續擴大年度營業利潤率。一如既往，我要感謝我們的客戶、合作夥伴和我們的員工一直以來的支持。那麼，接線員，我們現在開始提問。

(Operator Instructions) Our first question comes from Charles Shi from Needham & Company.

（操作員說明）我們的第一個問題來自 Needham & Company 的 Charles Shi。

Maybe the first one -- it's kind of like a 2-part question. Maybe this is to John. John, your fiscal year and Q1 guidance seem to imply a relatively flat revenue profile through the year. And I'm sure you heard through a peer, they seems to be seeing a slightly upward trending profile in their fiscal year. Although I know the fiscal year, yours and theirs, end at slightly different months. Is the difference just a matter of more conservatism on your side? Or is it a matter of a different end market mix or product mix? Maybe let me ask the second part. I think they are kind of related. I'll ask all at once. I think one quarter ago, you were cautious your hardware sales into fiscal '23 because 2022, like you just said, it's a record year for the hardware sales. So what is your assumption in the full year guidance for fiscal '23 on your hardware revenue profile in the year? Is it flat? Or is it -- are you assuming there is some recovery in the second half of the year?

也許是第一個問題——它有點像一個由兩部分組成的問題。也許這是給約翰的。約翰，你的財政年度和第一季度的指導似乎意味著全年收入狀況相對平穩。而且我敢肯定，您通過同行聽說過，他們似乎在財政年度看到了略微上升的趨勢。儘管我知道財政年度，但您和他們的財政年度結束的月份略有不同。區別只是你這邊更保守的問題嗎？或者是不同的終端市場組合或產品組合的問題？也許讓我問第二部分。我認為它們有點相關。我一口氣問完。我認為一個季度前，您對 23 財年的硬件銷售持謹慎態度，因為正如您剛才所說，2022 年是硬件銷售創紀錄的一年。那麼，您對 23 財年全年硬件收入狀況的全年指導假設是什麼？它是平的嗎？或者是 - 您是否假設今年下半年會有一些復甦？

Charles, thanks for the questions -- those excellent questions. Yes, we're expecting another strong start to the year as we continue to see strength in our hardware business. You did pick up some caution ahead, I think, the last quarter because hardware was proving not to be discretionary at all for our customers, and hardware demand was outpacing our ability to produce the hardware. And that was the case for the entirety of last year. We were not able to keep up with demand. Demand is really, really strong as we head into the new year for 2023. But we've increased our production capacity, and we feel more confident in our ability to meet that demand for 2023. And as you can see from the outlook, we're expecting that last year, the 85-15 split in recurring revenue to upfront revenue mix, we expect that to continue.

查爾斯，謝謝你提出的問題——那些很好的問題。是的，隨著我們繼續看到硬件業務的強勁勢頭，我們預計今年會有另一個強勁的開端。我認為，您確實在上個季度提前採取了一些謹慎態度，因為事實證明硬件對我們的客戶來說根本不是可自由支配的，而且硬件需求超過了我們生產硬件的能力。去年全年都是如此。我們跟不上需求。隨著我們進入 2023 年的新一年，需求真的非常強勁。但是我們提高了產能，我們對滿足 2023 年需求的能力更有信心。正如您從展望中看到的那樣，我們預計去年，85-15 的經常性收入與預付收入組合的分配，我們預計這種情況將繼續下去。

Now the second half of the year is harder to predict for hardware. So typically, I wait until the middle of the year. And if we see continued strength into the second half of the year, we can increase the second half later. And then I think that impacts the quarterly profile, because you started your question with why the year felt a little bit flat, and that's because we expect to deliver a lot of hardware in the first half of the year. And in the second half of the year, we've kind of derisked the second half of the year for hardware. If we see continued demand at this pace, we'll have to take the second half up.

現在下半年的硬件更難預測。所以通常情況下，我會等到年中。如果我們看到下半年持續強勁，我們可以在下半年增加。然後我認為這會影響季度概況，因為你開始問為什麼今年感覺有點平淡，那是因為我們預計在今年上半年交付大量硬件。而在今年下半年，我們對硬件下半年的風險有所降低。如果我們看到持續的需求以這種速度增長，我們將不得不增加下半年。

That's excellent. Maybe my second question, maybe this is for Anirudh. I think some investors are worried about hyperscaler spending on EDA, I mean, both on near-term and long-term sense. I mean in the near term, I mean, obviously, the layoff in tech, well, some people think it could mean reduce the chip design projects and lost jobs for design engineers. And in the long term, I mean, there's a smaller portion of the investors are kind of worried about the sustainability of the hyperscalers' spending strength. Although I think chatters around AI recently seems to suggest that, that strength is probably not going to -- going down but probably going to accelerate. So can you share your observation or any insights where demand coming from hyperscaler side? And what's your thought -- what's your current outlook going into like a 2-, 3-year horizon?

那太好了。也許我的第二個問題，也許這是給 Anirudh 的。我認為一些投資者擔心在 EDA 上的超大規模支出，我的意思是，無論是短期還是長期。我的意思是在短期內，我的意思是，很明顯，技術部門的裁員，嗯，有些人認為這可能意味著減少芯片設計項目和設計工程師的工作崗位。從長遠來看，我的意思是，只有一小部分投資者擔心超大規模企業支出實力的可持續性。儘管我認為最近圍繞 AI 的討論似乎表明，這種力量可能不會下降，但可能會加速。那麼，您能否分享您的觀察結果或對來自超大規模數據中心的需求的任何見解？你有什麼想法——你目前對 2 年、3 年的前景有何看法？

Yes, Charles, thanks for the question. Our demand is broad-based, right, across both our semiconductor customers and our system customers. So as we mentioned before, right now, about 45% of our business is coming from system companies. That includes hyperscaler and other kind of system companies. And I'm very cautiously optimistic that this trend is continued for a long time, okay, because I think some of these trends are irreversible of system companies doing silicon design. They're already having a lot of success. If you look at whether they're social media companies or phone companies or data center companies or car companies, they have multiple design projects in different stages of development. And overall, we see strong design activity on the systems side and on the semi side.

是的，查爾斯，謝謝你的提問。我們的需求是廣泛的，正確的，跨越我們的半導體客戶和我們的系統客戶。所以正如我們之前提到的，現在，我們大約 45% 的業務來自系統公司。這包括超大規模和其他類型的系統公司。而且我非常謹慎樂觀地認為這種趨勢會持續很長時間，好吧，因為我認為其中一些趨勢對於進行矽設計的系統公司來說是不可逆轉的。他們已經取得了很大的成功。如果你看看他們是社交媒體公司、電話公司、數據中心公司還是汽車公司，你會發現他們有多個處於不同開發階段的設計項目。總的來說，我們在系統方面和半方面看到了強大的設計活動。

Now there's always some reports here and there. This is not a straight line, right? Some customers may do more. Some customers may do less. But overall, like there will be more and more designs done by system companies. And as you know, the other part of our interaction with system companies is also expanding our portfolio to include system design and analysis products, our SDA segment. So we are working with system companies not just on the silicon side but on the system side, whether it's 3D-IC or thermal simulation or CFD. And you can see that part of our business is also growing very strongly. So last year, we reported about 27% growth in the system business, which is beyond our traditional EDA business. So overall, I'm pretty pleased, and I think design activity remains very strong driven by 3-nanometer and other things and the expansion of our portfolio through system design and analysis.

現在到處都有一些報導。這不是一條直線，對吧？有些客戶可能會做得更多。有些客戶可能做得更少。但總的來說，像系統公司做的設計會越來越多。如您所知，我們與系統公司互動的另一部分也在擴大我們的產品組合，以包括系統設計和分析產品，即我們的 SDA 部門。因此，我們不僅在矽方面而且在系統方面與系統公司合作，無論是 3D-IC 還是熱模擬或 CFD。你可以看到我們的部分業務也增長非常強勁。所以去年，我們報告系統業務增長了約 27%，這超出了我們傳統的 EDA 業務。所以總的來說，我很高興，我認為設計活動仍然非常強勁，受 3 納米和其他事物的推動，以及通過系統設計和分析擴展我們的產品組合。

Our next question comes from Jason Celino from KeyBanc Capital Markets.

我們的下一個問題來自 KeyBanc Capital Markets 的 Jason Celino。

Great quarter. When I look at the guidance, 12% to 14% growth at the start of the year, best growth guidance I've seen and you invest last year's on a much tougher comp. John, any change to philosophy in terms of how you set guidance?

偉大的季度。當我查看指導時，年初增長了 12% 到 14%，這是我見過的最好的增長指導，而且你在去年的指導上投資了一個更艱難的組合。約翰，在你如何設定指導方面對哲學有什麼改變嗎？

That's a great question, Jason. No, we've approached guidance the same way we normally do. Last year, if you recall, we wanted to wait until we had increased visibility into the hardware pipeline. So we were a bit more conservative about the second half for hardware. We're approaching this year very similarly. But we had -- we just finished with really, really strong momentum through the year. And the guide, I mean, there's so much of it coming out of backlog. We feel very confident in the year.

這是一個很好的問題，傑森。不，我們已經像往常一樣接近指導。去年，如果你還記得的話，我們想等到我們提高了對硬件管道的可見性。因此，我們對硬件的下半年更為保守。我們今年非常相似。但是我們 - 我們剛剛結束了這一年非常非常強勁的勢頭。我的意思是，指南中有很多來自積壓。我們對這一年充滿信心。

Okay. And then I think you mentioned on the hardware side, you increased capacity. How hard is it to toggle that up and down? I'm just trying to understand what this means in terms of confidence level and pipeline.

好的。然後我想你提到了硬件方面，你增加了容量。上下切換有多難？我只是想了解這在置信度和管道方面意味著什麼。

Yes, Jason, it's -- we've got access to more production capacity now. We have added additional lines to build the hardware. For the last year, we couldn't build it fast enough. We did ramp up production capacity, I think, 40% last year, but we sold more than that. So we didn't keep up with demand. So we ramped up production capacity again for 2023, and we feel very confident that we have access to the inventory and the components we need to meet that production demand. But also, to the extent that we can produce extra systems, we have a number of underserved or unserved customers that want access to our hardware in the cloud. So any excess capacity we can generate or any excess production we can generate, we can put into the cloud for that offering.

是的，傑森，它是——我們現在可以獲得更多的生產能力。我們添加了額外的線路來構建硬件。在過去的一年裡，我們構建它的速度不夠快。我們確實提高了產能，我認為，去年提高了 40%，但我們的銷量遠不止於此。所以我們沒有跟上需求。因此，我們在 2023 年再次提高了產能，我們非常有信心能夠獲得滿足生產需求所需的庫存和組件。而且，就我們可以生產額外系統而言，我們有許多服務不足或未得到服務的客戶想要訪問我們在雲中的硬件。因此，我們可以產生的任何過剩產能或我們可以產生的任何過剩產量，我們都可以將其放入雲中以提供該產品。

Our next question comes from Gary Mobley from Wells Fargo Securities.

我們的下一個問題來自 Wells Fargo Securities 的 Gary Mobley。

I wanted to ask about JedAI and all the related machine learning and AI-enabled tools. We've been getting a lot of questions from investors in terms of how to think about how that becomes accretive to your growth rate and how it becomes accretive to your average deal size. Maybe if you can just share with us where you're at in this price discovery phase and how you plan to, I guess, mass market price it? And if I'm not mistaken, this will all be included in digital IC, which according to the finish to the year was dilutive to your overall revenue growth. So how should we read into that? Is AI/machine learning simply just not impacting that line item yet?

我想詢問 JedAI 以及所有相關的機器學習和支持 AI 的工具。我們從投資者那裡收到了很多關於如何考慮它如何增加你的增長率以及它如何增加你的平均交易規模的問題。也許您可以與我們分享您在這個價格發現階段所處的位置，以及您打算如何為大眾市場定價？如果我沒記錯的話，這將全部包含在數字 IC 中，根據今年年底，這會稀釋您的整體收入增長。那麼我們應該如何解讀呢？人工智能/機器學習是否只是還沒有影響該訂單項？

Gary, great question. So first of all, I'm very optimistic about AI, and we always have talked about applying AI for optimization. I mean in EDA or in chip design or system design, it's more about automating the design process and producing better results. So even if you look at -- the way I look at it, even if you get it right now, some of these chips have 100 billion transistors, right, on 1 inch by 1 inch. And if you look at by 2030, they will have 1 trillion transistors, okay? So just in terms of size, it will be 10x more. And then the chips are more complicated, and then you add software on top of it. So the design complexity that our customers need to do will go up by at least 20, 30x in the next 5 to 7 years. So the only way to meet that is by more automation. That's the history of our industry. And the best way to do more automation right now is using AI, okay?

加里，好問題。所以首先，我非常看好人工智能，我們一直在談論應用人工智能進行優化。我的意思是在 EDA 或芯片設計或系統設計中，更多的是關於自動化設計過程並產生更好的結果。所以即使你看——我看它的方式，即使你現在明白了，其中一些芯片有 1000 億個晶體管，對，在 1 英寸乘 1 英寸的面積上。如果你看看到 2030 年，它們將擁有 1 萬億個晶體管，好嗎？因此，就尺寸而言，它將增加 10 倍。然後芯片更複雜，然後你在它上面添加軟件。因此，在未來 5 到 7 年內，我們客戶需要完成的設計複雜度將至少增加 20 倍、30 倍。因此，實現這一目標的唯一方法是提高自動化程度。這就是我們行業的歷史。現在實現更多自動化的最佳方法是使用 AI，好嗎？

And of course, we have done other ways to do automation in our industry. We started by doing more higher-level design, moving from transistor level to gate level. Over the last 5, 10 years, we have done a lot of massive [viralism], running things on more CPUs using cloud. But going forward, one of the biggest ways to improve productivity is using AI. And you see that across our product portfolio. And the real benefit is that a lot of the mundane tasks can be done by -- the repetitive tasks and mundane tasks can be done by AI, so the designer can move to more higher-value tasks, right?

當然，我們已經採用其他方式在我們的行業中實現自動化。我們從做更多更高層次的設計開始，從晶體管層次轉移到門層次。在過去的 5 年、10 年裡，我們做了很多大規模的[病毒式傳播]，使用雲在更多 CPU 上運行。但展望未來，提高生產力的最大方法之一是使用人工智能。您會在我們的產品組合中看到這一點。真正的好處是很多平凡的任務可以通過人工智能來完成——重複的任務和平凡的任務可以由 AI 完成，所以設計師可以轉向更高價值的任務，對吧？

And so the way we approach it through is a very comprehensive -- we build this JedAI data analytics because data is critical, too, as you know, to a data analytics and AI platform. And then we have multiple applications on top of it, okay? So Cerebrus is a key one for implementation. We also launched a few months ago Verisium for verification, which is another very difficult and all-consuming problem for our customers. And then we are not only focused on the chip side but also on the system side. So we have Optimality, which is having great success on the system side. And typically, the system customers are not used to optimization or this level of automation that the chip industry has seen. But we are getting dramatic improvements with the Optimality as well.

因此，我們採用的方法非常全面——我們構建了這個 JedAI 數據分析，因為數據對於數據分析和 AI 平台也很重要。然後我們在它上面有多個應用程序，好嗎？所以 Cerebrus 是實現的關鍵。幾個月前我們還推出了 Verisium 進行驗證，這對我們的客戶來說是另一個非常困難和耗費精力的問題。然後我們不僅專注於芯片端，還專注於系統端。所以我們有 Optimality，它在系統方面取得了巨大的成功。通常，系統客戶不習慣優化或芯片行業所見的這種自動化水平。但是我們也通過 Optimality 獲得了顯著的改進。

So taken together, I believe that we have the most comprehensive AI portfolio. And we have always focused AI on optimization, which, of course, is now called generative AI. And we have been working on it for 5 years. I think the products introduced were about 2 years ago. And even for Cerebrus, I mentioned in my prepared remarks several leading companies like Intel, NVIDIA, Samsung, Renesas, they all talked about the results they're seeing. We have more than 160 designs that we are tracking on Cerebrus. And if you'll include Verisium and Optimality, we have hundreds of designs being done by AI and -- because I believe in the next few years, almost all designs will have some AI component. And that is driving the growth that you're seeing in our business and the outlook.

因此，我相信我們擁有最全面的 AI 產品組合。我們一直將 AI 的重點放在優化上，當然，現在稱為生成 AI。我們已經為此努力了 5 年。我認為推出的產品大約是 2 年前的。甚至對於 Cerebrus，我在準備好的發言中提到了幾家領先的公司，如英特爾、英偉達、三星、瑞薩，他們都談到了他們看到的結果。我們在 Cerebrus 上跟踪了 160 多種設計。如果你將 Verisium 和 Optimality 包括在內，我們有數百種設計是由人工智能完成的——因為我相信在接下來的幾年裡，幾乎所有的設計都會有一些人工智能組件。這正在推動您在我們的業務和前景中看到的增長。

And a quick follow-up for John. Backlog, up again. Nice job on that, up 32% in the year. Next 12-month backlog, up 26% on the year. Was there a large deal or renewal that came into the fray in the fourth quarter? Or maybe you could just speak in terms of diversity of that growth?

以及約翰的快速跟進。積壓，再次上升。這方面做得很好，今年增長了 32%。接下來 12 個月的積壓訂單，同比增長 26%。第四季度是否有大宗交易或續約？或者，也許您可以就這種增長的多樣性發表意見？

Yes, Gary, great question. I mean we had a really strong finish to the year. And as you know, I mean, contract timing typically impacts cRPO in any one quarter. But if you look over a typical contract cycle, you'll see that we're particularly pleased with the 3-year CAGR on cRPO. It's tracking to mid-teens growth now, and that's very consistent to the mid-teens growth we achieved over the last 3 years to 2022 and the mid-teens growth that's implied at the midpoint of our guidance for 2023.

是的，加里，好問題。我的意思是我們今年的成績非常出色。如您所知，我的意思是，合同時間通常會影響任何一個季度的 cRPO。但是，如果您查看一個典型的合同周期，您會發現我們對 cRPO 的 3 年 CAGR 特別滿意。它現在正在追踪十幾歲中期的增長，這與我們在過去 3 年到 2022 年實現的十幾歲中期增長以及我們 2023 年指導的中點所暗示的十幾歲中期增長非常一致。

Our next question comes from Jay Vleeschhouwer from Griffin Securities.

我們的下一個問題來自 Griffin Securities 的 Jay Vleeschhouwer。

Anirudh, a wise man once said that silicon companies are becoming increasingly like systems companies, and systems companies are becoming increasingly like semiconductor companies. You alluded earlier to some of the additional opportunities that systems companies represent for you, for example, in CFD and so forth. In what other ways would you say that these 2 classes of customers do still remain different or different enough for you, even though they are becoming more alike, and in ways that perhaps influence your -- either your R&D or your go-to-market? And then the second question is if you could talk about where your R&D priorities go from here. The last number of years, you and, for that matter, Synopsys have significantly ramped up, for example, your investments in synthesis, verification, AI, you mentioned, of course. And from here, how are you thinking about the R&D priorities in areas like custom and CFD and of course, in AI?

Anirudh，一位智者曾經說過，矽公司越來越像系統公司，而係統公司越來越像半導體公司。您之前提到了系統公司為您提供的一些額外機會，例如 CFD 等。您還會以其他什麼方式說這兩類客戶對您來說仍然不同或足夠不同，即使他們變得越來越相似，並且可能影響您的方式 - 無論是您的研發還是您的上市？然後第二個問題是你是否可以談談你的研發重點從這裡開始。在過去的幾年裡，你和 Synopsys 已經顯著增加了，例如，你在綜合、驗證、人工智能方面的投資，你當然提到了。從這裡開始，您如何考慮定制和 CFD 等領域的研發優先事項，當然還有人工智能？

Yes, thanks, Jay. Very valid questions. And of course, great point that system companies are becoming semi companies and semi companies are becoming system companies. And like I said before, this is irreversible trend, okay? This is irreversible. And for us, we invest heavily in R&D, as you know. I mean about 35% of our revenue is invested in R&D. That's one of the highest percentages of any S&P 500 company. Just for your reference, you may know this already, we have 10,000 people in the company. 9,000 are engineers or computer scientists. Either they are in customer support roles or in R&D. And we always make sure that the core is good, okay, because the core has to be best-in-class, so whether that is synthesis, like you mentioned, place and route, circuit simulation, analog, verification. So that's a given, making sure that the core is best in class.

是的，謝謝，傑伊。非常有效的問題。當然，重要的一點是系統公司正在成為半公司，而半公司正在成為系統公司。就像我之前說的，這是不可逆轉的趨勢，好嗎？這是不可逆轉的。如您所知，對於我們來說，我們在研發方面投入了大量資金。我的意思是我們大約 35% 的收入投資於研發。這是標準普爾 500 強公司中比例最高的公司之一。僅供參考，您可能已經知道了，我們公司有 10,000 人。 9,000 名工程師或計算機科學家。他們要么擔任客戶支持角色，要么從事研發工作。我們始終確保核心是好的，好吧，因為核心必須是一流的，所以無論是合成，就像你提到的，佈局佈線，電路仿真，模擬，驗證。所以這是給定的，確保核心是一流的。

And then on top of that, there are certain thematic things that we are investing in. And then the 3 that I want to highlight, which I think will be thematic for years to come, one is, of course, AI. AI will have a big effect on, like I mentioned, the design work. And that's true for both semi and system companies. Second is 3D-IC, the emergence of chiplets and heterogeneous integration, and Cadence is the best positioned to take advantage of this. And then the third area is this move to systems, system design and analysis. And our engagement with system companies are similar, to a lot of extent, with semi companies. Everybody wants to do more with less now. So the benefits of AI and productivity and better results are there in both set of customers.

然後最重要的是，我們正在投資某些主題。然後是我想強調的 3 個，我認為這將是未來幾年的主題，其中之一當然是 AI。正如我提到的，人工智能將對設計工作產生重大影響。對於半成品和系統公司來說都是如此。其次是 3D-IC，小芯片和異構集成的出現，而 Cadence 最適合利用這一點。然後第三個領域是轉向系統、系統設計和分析。我們與系統公司的合作在很大程度上與半導體公司相似。現在每個人都想事半功倍。因此，人工智能和生產力的好處以及更好的結果在兩組客戶中都有。

But of course, our emerging portfolio and system design and analysis provides unique value to our system companies because we are no longer talking to the chip designers. We're talking to the system designers, the architect. The coupling of the mechanical and the electrical designs uniquely positions us. This convergence is going to happen between system and semi, between electrical and mechanical. And there is no other company better positioned than Cadence to take advantage of this, okay? So overall, to answer your question, these are the 3 big themes on top of the base. See, the base is always important, the best-in-class of the basic algorithms. But AI, 3D-IC and systems will be with us for years to come.

但當然，我們新興的產品組合和系統設計與分析為我們的系統公司提供了獨特的價值，因為我們不再與芯片設計師交談。我們正在與系統設計師、架構師交談。機械和電氣設計的結合使我們獨樹一幟。這種融合將發生在系統和半之間，電氣和機械之間。沒有其他公司比 Cadence 更能利用這一點，好嗎？所以總的來說，為了回答你的問題，這些是基礎之上的 3 個大主題。看，基礎總是很重要的，最好的基礎算法。但人工智能、3D-IC 和系統將在未來幾年與我們同在。

Our next question comes from Harlan Sur from JPMorgan.

我們的下一個問題來自摩根大通的 Harlan Sur。

So on hardware verification, you guys have, it sounds like, very good visibility to the first half of the year. I know the team still has some concerns on maybe more discretionary-type spending pullback there. But we've also talked about how these hardware platforms are becoming a need-to-have, right, not a nice-to-have but a need-to-have as it relates to these very complex digital SoC platforms. We're well into the semiconductor industry downturn. I would have assumed that you would have already seen some cancellations in orders or pushouts in hardware shipments if customers were concerned. So has the team seen any of this type of activity? I'm just trying to figure out what's driving the conservatism here on hardware.

因此，在硬件驗證方面，你們對今年上半年的了解非常好。我知道團隊仍然擔心那裡可能會出現更多可自由支配的支出回調。但我們也討論了這些硬件平台如何成為一種必需品，對，不是一種很好的，而是一種必需品，因為它與這些非常複雜的數字 SoC 平台相關。我們正處於半導體行業的低迷時期。如果客戶擔心的話，我本以為您已經看到一些訂單被取消或硬件發貨被推遲。那麼團隊是否看到過任何此類活動？我只是想弄清楚是什麼在推動硬件上的保守主義。

Well, thanks for the question. This is Anirudh. I think -- you said it. I think what we see is that hardware is no longer -- I know it could be part of CapEx, but it's no longer discretionary spend for our customers. I mean any chips that are designed today, any complex chips, you have to use these hardware platforms to verify them. And if you don't verify them properly, then you spend all this money, the chips comes back and it doesn't work. I mean that's a big no-no, right? It delays the whole product and the expense. So hardware platforms, both Palladium and Protium, are no longer like nice-to-have. You need to have them, okay? And then we have also strengthened our portfolio by not just chip verification but software bring-up. Chip verification has been traditional sense of Palladium. And now with software bring-up with Protium, we provide a unique -- so, so far, the demand is strong. And we had a record year last year.

好吧，謝謝你的提問。這是阿尼魯德。我想——你說了。我認為我們看到的是硬件不再——我知道它可能是資本支出的一部分，但它不再是我們客戶的可自由支配支出。我的意思是今天設計的任何芯片，任何復雜的芯片，你都必須使用這些硬件平台來驗證它們。如果你沒有正確地驗證它們，那麼你就會花掉所有這些錢，芯片又回來了，它就不起作用了。我的意思是這是一個很大的禁忌，對吧？它耽誤了整個產品和費用。因此，硬件平台，包括 Palladium 和 Protium，不再是錦上添花。你需要擁有它們，好嗎？然後我們不僅通過芯片驗證而且通過軟件啟動來加強我們的產品組合。芯片驗證一直是傳統意義上的鈀金。現在，通過 Protium 的軟件開發，我們提供了一個獨特的——所以，到目前為止，需求很強勁。去年我們創下了紀錄。

These days, we have -- I think we had a record year, year after that, too. But we see continued growth in hardware, and we are very pleased with our competitive position in hardware. We are very pleased with the demand as it's no longer nice-to-have. It's no longer discretionary. And also in multiple end markets, so what happened is like we talked about automotive, right? So automotive also has much higher complexity chips now. So the other thing with hardware is it is also expanding to other end markets. Whereas traditionally, the big chips used to do it, like data center chips. But now even automotive chips are almost -- networking, almost all kinds of chips will require the use of these hardware systems. And we are very well positioned to take advantage of that.

這些天，我們 - 我認為我們有創紀錄的一年，之後的一年。但我們看到硬件的持續增長，我們對我們在硬件方面的競爭地位感到非常滿意。我們對需求感到非常滿意，因為它不再是可有可無的了。它不再是自由裁量權。而且在多個終端市場，發生的事情就像我們談論汽車一樣，對吧？所以汽車現在也有更複雜的芯片。所以硬件的另一件事是它也在擴展到其他終端市場。而傳統上，大芯片過去常常這樣做，比如數據中心芯片。但是現在連汽車芯片都差不多——聯網了，幾乎所有的芯片都會需要用到這些硬件系統。我們完全有能力利用這一點。

Perfect. And then my follow-up. The team has done a great job on integrating machine learning-based methods as a part of your customer's digital SoC design and verification tools, although different, but nevertheless still needing many iterations around many variables. Is your custom cell and IT design and analog simulation and verification, it seems like the team could take advantage of your ML frameworks and apply it to your custom and analog franchise. Is the team working on integrating ML into Virtuoso and other parts of the analog and custom portfolio?

完美的。然後是我的後續行動。該團隊在將基於機器學習的方法集成為客戶數字 SoC 設計和驗證工具的一部分方面做得很好，儘管有所不同，但仍然需要圍繞許多變量進行多次迭代。是您的自定義單元和 IT 設計以及模擬仿真和驗證，團隊似乎可以利用您的 ML 框架並將其應用到您的自定義和模擬特許經營中。該團隊是否致力於將 ML 集成到 Virtuoso 以及模擬和定制產品組合的其他部分？

That's a very good point. So I mean, like I mentioned, we have more than 30 projects in all aspects of AI. AI, inherently, is computational software, right? It's computer science plus math, which is what we're very good at. So you can assume that we are working through other design flow. And then we announce products in a more conservative way now. We want to make sure they're working with several customers with different end markets, and then we announce them. So last year -- the year before, we talked about digital implementation. Last year, we talked about JedAI and verification. And I think this year, you will hear more from us in other parts of the business. The analog custom business is ripe for more automation and packaging PCB. So please stay tuned. But you can be rest assured, we are applying AI wherever it is possible, and it is possible to apply it everywhere, okay?

這是一個很好的觀點。所以我的意思是，就像我提到的那樣，我們在 AI 的各個方面都有 30 多個項目。人工智能本質上是計算軟件，對吧？它是計算機科學加數學，這是我們非常擅長的。所以你可以假設我們正在處理其他設計流程。然後我們現在以更保守的方式發布產品。我們想確保他們正在與具有不同終端市場的多個客戶合作，然後我們會宣布他們。所以去年 - 前一年，我們談到了數字實施。去年，我們談到了 JedAI 和驗證。我認為今年，您會在業務的其他部分聽到更多來自我們的消息。模擬定制業務對於更多自動化和封裝 PCB 來說已經成熟。所以請繼續關注。但是你放心，我們正在把人工智能應用到任何可能的地方，而且是可以應用到任何地方的，好嗎？

Our next question comes from Joe Vruwink from Baird.

我們的下一個問題來自 Baird 的 Joe Vruwink。

I wanted to go back quickly to the current RPO topic. So fourth quarter was very good, finished north of 25% growth. And then even if I appreciate that probably has some hardware in it, John, as you said, that the trend growth is closer to a mid-teens-type number now. Can you just help reconcile that mid-teens growth with the implied outlook for recurring revenue? I think it would be more like 12% or 14% growth. It does seem like starting visibility to come from backlog is higher than usual. And so is there an implicit bookings assumption in 2023 that's factoring into this? Or those conservatism, like it applies to hardware, also maybe apply to software bookings entering the year?

我想快速回到當前的 RPO 主題。所以第四季度非常好，完成了超過 25% 的增長。然後，即使我很欣賞其中可能有一些硬件，約翰，正如你所說，現在趨勢增長更接近十幾歲左右的數字。您能否幫助協調十幾歲中期的增長與隱含的經常性收入前景？我認為這更像是 12% 或 14% 的增長。看起來來自積壓的開始可見度確實比平時高。那麼，2023 年是否有一個隱含的預訂假設正在考慮這個因素？或者那些保守主義，就像它適用於硬件一樣，也可能適用於進入這一年的軟件預訂？

Great question, Joe. I mean, as Gary mentioned earlier, that he was asking about the number of customers, was there any big customer, there wasn't one big customer in Q4. We just had a very strong finish to the year, very strong finish across all lines of business. And particularly, hardware was very strong as we closed out the year in terms of bookings. I think our hardware systems are just not discretionary spend. As Anirudh said earlier, they've become an indispensable part of our customer spend on the R&D side, and people wanted to get their orders in before the end of the year. So we have a lot of visibility into the next year. And as you know, cRPO, from a quarter-to-quarter basis, can be lumpy in nature. And I think like if you want to reconcile between where the growth in cRPO, say, for the last year is probably up closer to 20%, but if you look over 3 years, it's mid-teens. And if you look over 3 years for our revenue CAGR, that's mid-teens to '22, and we're guiding to mid-teens for '23. So I think you can take out some of the noise by looking over a 3-year kind of average period.

好問題，喬。我的意思是，正如加里之前提到的，他在詢問客戶數量，是否有任何大客戶，第四季度沒有一個大客戶。我們剛剛在這一年取得了非常出色的成績，所有業務領域都取得了非常出色的成績。特別是，硬件非常強大，因為我們在預訂方面結束了這一年。我認為我們的硬件系統不是可自由支配的開支。正如 Anirudh 早些時候所說，它們已經成為我們客戶在研發方面支出中不可或缺的一部分，人們希望在年底前收到訂單。因此，我們對明年有很多了解。如您所知，從季度到季度來看，cRPO 本質上可能是不穩定的。而且我認為，如果你想在 cRPO 的增長之間進行調和，比如說，去年可能接近 20%，但如果你看 3 年以上，它是十幾歲左右。如果你看一下 3 年以上的收入複合年增長率，那就是 22 歲左右，而我們正在指導 23 歲左右。所以我認為你可以通過查看 3 年的平均週期來消除一些噪音。

Okay. That's helpful. And then I guess I'll stay on the topic of AI. And obviously, Cadence is already using a lot of reinforcement learning in design flow and verification. I guess any thoughts on kind of the recent attention around large language models. You've started to see, I think, some tinkering with like RTL code generation or maybe within certain aspects of verification. Does this maybe supplement what you're already doing? Or does it have the potential to bring about entirely new products?

好的。這很有幫助。然後我想我會繼續討論人工智能的話題。顯然，Cadence 已經在設計流程和驗證中大量使用強化學習。我猜對最近圍繞大型語言模型的關注有什麼想法。我認為，您已經開始看到一些修補，例如 RTL 代碼生成或驗證的某些方面。這是否可以補充您已經在做的事情？或者它是否有可能帶來全新的產品？

Yes, good point. I mean one thing to mention is we always look at all the kind of AI technologies. We have a pretty talented team. And like I mentioned, for us, the biggest application is simulation plus optimization because you can generate much better results for the customers. So we are using reinforcement learning and all kinds of other kind of ML techniques for a while now. And to give you an example, like some of the results we are seeing, the PPA can be improved by 10% to 15%, okay? That's a lot of improvement. And typically, you see this improvement from -- going from one node to the next, okay, spending billions of dollars, okay? You can get that from better kind of AI algorithms. Now some of the language models, there could be other applications. Especially with the interface of our tools for customer support, there could be applications. But the main customer-facing applications are when we give better results, better productivity, which we are already doing in multiple ways, as I mentioned, and we will continue to look at all possible ways to improve that.

是的，好點。我的意思是要提到的一件事是我們總是關注所有類型的人工智能技術。我們有一個非常有才華的團隊。就像我提到的，對我們來說，最大的應用是模擬加優化，因為你可以為客戶產生更好的結果。所以我們現在使用強化學習和各種其他類型的 ML 技術已有一段時間了。舉個例子，就像我們看到的一些結果一樣，PPA 可以提高 10% 到 15%，好嗎？這是很大的改進。通常，你會看到這種改進——從一個節點到下一個節點，好吧，花費數十億美元，好嗎？你可以從更好的人工智能算法中得到它。現在一些語言模型，可能會有其他的應用。特別是對於我們用於客戶支持的工具的界面，可能會有應用程序。但主要的面向客戶的應用程序是當我們提供更好的結果、更高的生產力時，正如我提到的，我們已經以多種方式在做這件事，我們將繼續尋找所有可能的方法來改進它。

Our next question comes from Vivek Arya from Bank of America Securities.

我們的下一個問題來自美國銀行證券公司的 Vivek Arya。

Anirudh, when I look at the relative growth in your different segments, the System Design & Analysis seems to be the fastest-growing segment, obviously, off a lower base over the last 5 years. What do you think has driven that outperformance? And as you get bigger in that market, does the competitive landscape change? Can you continue to outgrow the market? Or just how do you think about that particular aspect of your growth drivers?

Anirudh，當我查看您不同細分市場的相對增長時，系統設計與分析似乎是增長最快的細分市場，顯然，過去 5 年的基數較低。您認為是什麼推動了這種出色表現？隨著您在該市場中變得越來越大，競爭格局是否會發生變化？你能繼續超越市場嗎？或者您如何看待增長動力的特定方面？

Yes, Vivek, I just want to point out first that all businesses are doing great. If you look at even our analog business, that's grew 13%, okay? That was used to be a stable business a few years ago, okay? Digital business, 17%, okay? These are very remarkable numbers, okay? Verification business, 28%, okay? So I mean, just to put that in context to what was happening in EDA and chip design like 5 years ago to now, I mean, this is a remarkable growth in our business and at the same time, very good profitability. We have more than 40% operating margin. So I think we focus on growth and profitability.

是的，Vivek，我只想首先指出所有企業都做得很好。如果你看看我們的模擬業務，它增長了 13%，好嗎？幾年前那是個穩定的生意好嗎？數字業務，17%，好嗎？這些都是非常了不起的數字，好嗎？驗證業務，28%，好嗎？所以我的意思是，只是把它放在 5 年前到現在 EDA 和芯片設計中發生的事情的背景下，我的意思是，這是我們業務的顯著增長，同時也是非常好的盈利能力。我們的營業利潤率超過 40%。所以我認為我們專注於增長和盈利能力。

And to come to your question on systems, I mean, that had a very strong year with 27% growth, and that's because our products are just better. And the customers want this integration, like I mentioned. Like Jay mentioned earlier, system companies and semi companies. So I think this is, again, an irreversible trend. There is a need for power analysis, thermal analysis. When you look at 3D-IC, one of the biggest things I've mentioned before is thermal simulation. Cadence is in the best position to provide that. Electromagnetic simulations, our products will run like 10, 20x faster and higher performance because of our computational software [trend].

關於系統問題，我的意思是，那一年增長非常強勁，增長了 27%，那是因為我們的產品更好。就像我提到的那樣，客戶想要這種集成。就像傑伊前面提到的，系統公司和半公司。所以我認為這又是一個不可逆轉的趨勢。需要進行功率分析、熱分析。當您查看 3D-IC 時，我之前提到的最重要的事情之一就是熱模擬。 Cadence 最適合提供這一點。電磁模擬，由於我們的計算軟件[趨勢]，我們的產品運行速度將提高 10 倍、20 倍，性能也會提高。

So I don't see anything in the near term that will change that. I mean this is going to continue, right? So we are pretty confident where we are. And as you know, we are expanding to other areas using the expertise, whether it's computational fluid dynamics or biosimulation. And I think this is another irreversible trend, the need of simulation. And the other thing that we are investing heavily in that space, which I mentioned in my prepared remarks, is AI and optimization. That area is still -- they could barely simulate things in that -- you could barely simulate a [wing] or a car properly, right? Forget about optimization.

所以我認為短期內不會有任何改變。我的意思是這會繼續下去，對吧？所以我們對自己的位置非常有信心。如您所知，我們正在利用這些專業知識擴展到其他領域，無論是計算流體動力學還是生物模擬。而且我認為這是另一個不可逆轉的趨勢，模擬的需要。我在準備好的發言中提到的另一件事是我們在該領域進行了大量投資，那就是人工智能和優化。那個區域仍然 - 他們幾乎無法模擬其中的東西 - 你幾乎無法正確模擬[機翼]或汽車，對吧？忘記優化。

But with Optimality, not only we can simulate, we can put that in the inner loop of a reinforcement-based AI engine to give results automatically that, that space has not seen. So I'm actually very optimistic about AI-driven optimization in the system space because it's entirely new to that set of customers, along with like regular simulation performance and capacity improvement. So I think we are still a small part of the market but growing rapidly. I think we crossed $400 million in revenue in that important segment, and I think there's a lot of room to grow there.

但是有了 Optimality，我們不僅可以模擬，還可以將其置於基於強化的 AI 引擎的內部循環中，以自動給出那個空間還沒有看到的結果。所以我實際上對系統空間中人工智能驅動的優化非常樂觀，因為它對這組客戶來說是全新的，以及常規模擬性能和容量改進。所以我認為我們仍然是市場的一小部分，但增長迅速。我認為我們在這個重要領域的收入超過了 4 億美元，而且我認為那裡還有很大的增長空間。

Understood. And for my follow-up, John, not complaining at all, but when I look at the implied incremental EBIT margin for this year, it seems somewhat lower than the average incremental margins that you've managed to achieve over the last few years. Just wanted to make sure we are not missing anything from a cost perspective that could restrain incremental EBIT leverage this year.

明白了。對於我的後續行動，約翰，一點也不抱怨，但當我查看今年隱含的遞增息稅前利潤率時，它似乎比過去幾年你設法實現的平均遞增利潤率要低一些。只是想確保我們不會從成本角度遺漏任何可能限制今年 EBIT 槓桿增量的因素。

That's a great question, Vivek, I mean, and thanks for pointing out incremental margins. As you know, we focus very carefully on those. Over the last -- I mean, we're very pleased that -- I think 2022, I think, was the sixth year in a row we achieved over 50% incremental margins. And the range that we've achieved over that period of time have ranged from 52%, I think, to 58% incremental margin over the last 6 years. So very, very pleased with that. And we're starting this year with slightly less in the guide. I think it's 48.5% is what is implied in the guide. But that is the largest and strongest -- the largest initial guide and the strongest start to any of the 7 years -- of the last 7 years. So very, very pleased to be in this position starting off the year. And as you know, throughout the year, we try to find profitable and sustainable revenue growth to improve that through the year. And we've managed to achieve that 6 years running, feel confident about doing that for a seventh year.

這是一個很好的問題，Vivek，我的意思是，感謝您指出增量利潤。如您所知，我們非常關注這些。在過去——我的意思是，我們很高興——我認為 2022 年是我們連續第六年實現超過 50% 的增量利潤率。我認為，在過去 6 年中，我們在那段時間取得的利潤率從 52% 到 58% 不等。對此非常非常滿意。我們從今年開始，指南中的內容略有減少。我認為指南中暗示的是 48.5%。但這是過去 7 年中最大和最強的 - 最大的初始指南和 7 年中任何一年的最強勁開端。非常非常高興從今年開始擔任這個職位。如您所知，我們全年都在努力尋找盈利和可持續的收入增長，以改善這一年。我們已經連續 6 年實現了這一目標，對連續第七年做到這一點充滿信心。

Our next question comes from Blair Abernethy from Rosenblatt Securities.

我們的下一個問題來自 Rosenblatt Securities 的 Blair Abernethy。

Great quarter, guys. Just wanted to see if there's anything to update on the Future Facilities acquisition last summer. You've had it for a couple of quarters now. How is that trending? And are you seeing opportunities to cross-sell your other simulation software into that space?

偉大的季度，伙計們。只是想看看去年夏天 Future Facilities 的收購是否有任何更新。你現在已經有幾個季度了。趨勢如何？您是否看到了將您的其他仿真軟件交叉銷售到該領域的機會？

Yes, great point. Great point. I mean I'm very excited about Future Facilities because I think that the way to differentiate in the system space is also through differentiated vertical offerings. Because if you look at the simulation space, there are only few like kind of unknown algorithms. There's finite element, there is CFD and now we are doing molecular simulation. But there are a lot of end markets, okay? And the one way to even further differentiate CFD -- one way we differentiate CFD is by our compute power, right, doing things -- bigger things and faster. But the other way to differentiate it is building a vertical kind of end market. And in CFD, one of the biggest thing is this data centers and smart buildings.

是的，很重要。好點。我的意思是我對 Future Facilities 感到非常興奮，因為我認為在系統空間中實現差異化的方式也是通過差異化的垂直產品。因為如果你看看模擬空間，就會發現很少有類似的未知算法。有有限元，有 CFD，現在我們正在進行分子模擬。但是有很多終端市場，好嗎？進一步區分 CFD 的一種方法——我們區分 CFD 的一種方法是通過我們的計算能力，正確的，做事情——更大的事情和更快的事情。但另一種區分它的方法是建立一個垂直的終端市場。在 CFD 中，最重要的事情之一就是數據中心和智能建築。

So with Future Facilities, not only the thermal simulation is good, the core CFD engines and we can couple it with the rest of the CFD product. But they also have built like a lot of models for -- like if you go into a building or a data center, they have models for all the components, whether they're AC vents or AC systems and racks and all that. So the customers can get very, very accurate digital twin for the data center. And that technology can be actually applied to any building, not just a data center. Now we started with data center because that's a big market, and as you know, they consume a lot of power.

因此，有了 Future Facilities，不僅熱模擬很好，核心 CFD 引擎也很好，我們可以將它與 CFD 產品的其餘部分結合起來。但他們也建立了很多模型——比如如果你進入建築物或數據中心，他們有所有組件的模型，無論是交流通風口還是交流系統和機架等等。因此，客戶可以獲得非常非常準確的數據中心數字雙胞胎。而且該技術實際上可以應用於任何建築物，而不僅僅是數據中心。現在我們從數據中心開始，因為這是一個很大的市場，而且如你所知，它們消耗大量電力。

So right now, data centers consume as much energy as like the whole airline industry, okay? So that's a lot of energy that they're consuming. And we have a lot of engagements that have picked up since we -- because they are now in a bigger home with the whole Cadence reach. So we have a lot of engagements which are picked up with the hyperscalers and with other companies even in other industries because a lot of companies have big data centers, whether they're internal or on the cloud. So you will see more of that from us, and I'm pretty optimistic about -- that's why we made that acquisition of this verticalization of -- having these vertical solutions in the CFD space.

所以現在，數據中心消耗的能源和整個航空業一樣多，好嗎？所以他們消耗了大量的能量。自從我們以來，我們有很多參與——因為他們現在在一個更大的家裡，擁有整個 Cadence 的影響力。因此，我們與超大規模企業、甚至其他行業的其他公司有很多合作，因為很多公司都有大數據中心，無論是內部數據中心還是雲端數據中心。所以你會從我們這裡看到更多，我非常樂觀——這就是為什麼我們收購了這種垂直化——在 CFD 領域擁有這些垂直解決方案。

Our next question comes from Ruben Roy from Stifel Financial.

我們的下一個問題來自 Stifel Financial 的 Ruben Roy。

Anirudh, you got a bunch of questions on hardware, and I think Harlan's question hit on most of the things that I wanted to ask about. But I guess, one thing that stood out to me from your commentary was that 2/3 of your, I think, new orders in '22 were for both Palladium and Protium. And that was interesting to me because it seems like as we're moving to more SSD designs, more firmware and software on those designs, that, that sort of selling Palladium and Protium together is likely to move up. So I guess if you could just comment on that and how you're thinking about that as either TAM expansion or a revenue growth driver kind of longer term behind hardware would be interesting.

Anirudh，你有很多關於硬件的問題，我認為 Harlan 的問題涉及到我想問的大部分問題。但我想，從你的評論中讓我印象深刻的一件事是，我認為你在 22 年的新訂單中有 2/3 是鈀金和 Protium。這對我來說很有趣，因為似乎隨著我們轉向更多的 SSD 設計，這些設計上的更多固件和軟件，那種一起銷售鈀和 Protium 的情況可能會上升。所以我想如果你能對此發表評論，以及你如何將其視為 TAM 擴張或硬件背後的收入增長驅動力，那將很有趣。

Yes, absolutely. Great point. I mean the thing with hardware, first of all, it's in a secular growth trend like we mentioned, but it's always good to have multiple products. And that's why the addition of Palladium and Protium that are complementary but address different parts, which both of them are growing, it also provides more predictable growth for our business. So like even though we had a record year last year, we still see growth this year and in the future because of this product mix because software bring-up -- I don't need to tell you, most of the designs now are software-defined hardware design. Software is what is driving the requirements for the hardware design. So initially, they actually start with the software model, and that is run on Protium, to figure out what kind of architecture to do. And then once you do the chip design, then Palladium comes in. And then at the end, Protium comes in again.

是的，一點沒錯。好點。我的意思是硬件方面，首先，它正處於我們提到的長期增長趨勢中，但擁有多種產品總是好的。這就是為什麼添加 Palladium 和 Protium，它們是互補的，但針對不同的部分，它們都在增長，這也為我們的業務提供了更可預測的增長。因此，儘管我們去年創下了紀錄，但由於這種產品組合，我們今年和未來仍然會看到增長，因為軟件啟動——我不需要告訴你，現在大部分設計都是軟件- 定義的硬件設計。軟件是推動硬件設計需求的因素。所以最初，他們實際上是從軟件模型開始的，它在 Protium 上運行，以弄清楚要做什麼類型的架構。然後一旦你進行了芯片設計，鈀就會出現。最後，Protium 會再次出現。

So there is a lot of back and forth between software bring-up and chip design. And therefore, having a common compiler gives a unique value that you can move it seamlessly between the 2 platforms. But I also believe that just like our customers have different silicon platforms, we need to do that, too, okay? So Palladium, the reason it is well differentiated is we use our own Cadence silicon. We design it ourselves, built with one of our foundry partners, and that gives advantage in terms of compile time that is unmatched, right? But for software bring-up, FPGA is good. So we use FPGA platforms, and then we have built a differentiated offering with Protium.

所以在軟件啟動和芯片設計之間有很多來回。因此，擁有一個通用編譯器具有獨特的價值，您可以在兩個平台之間無縫移動它。但我也相信，就像我們的客戶擁有不同的矽平台一樣，我們也需要這樣做，好嗎？因此，Palladium 之所以與眾不同，是因為我們使用了自己的 Cadence 矽。我們自己設計，與我們的代工廠合作夥伴之一一起構建，這在編譯時間方面具有無與倫比的優勢，對吧？但是對於軟件啟動，FPGA 是好的。所以我們使用 FPGA 平台，然後我們用 Protium 構建了一個差異化的產品。

And same thing on like our regular verification systems like formal with Jasper and Xcelium logic simulation, we also offer multiple hardware platforms. Of course, x86, that has been traditional. But we have also ported all our software platforms, especially in verification, to ARM-based systems because they can offer price performance. So not only we have multiple products with Palladium and Protium, we are always looking at the right hardware to run them on. So in case of Palladium, it's a custom silicon. In case of Protium, it's FPGA. And then in logic simulation and formal, we look at both x86 and ARM, so to give a full variety of options to our customers.

和我們的常規驗證系統一樣，比如正式的 Jasper 和 Xcelium 邏輯模擬，我們還提供多個硬件平台。當然，x86，那是傳統的。但我們也將我們所有的軟件平台，特別是在驗證方面，移植到基於 ARM 的系統，因為它們可以提供性價比。因此，我們不僅擁有多種包含 Palladium 和 Protium 的產品，而且我們一直在尋找合適的硬件來運行它們。因此，對於鈀，它是一種定制矽。對於 Protium，它是 FPGA。然後在邏輯模擬和形式化方面，我們同時看x86和ARM，所以給我們的客戶提供全方位的選擇。

If I could just ask a quick follow-up for John. I don't think you were asked on the pricing environment, John. And just watching the backlog move up again and kind of all the mega trends that you're talking about, it's understandable that things are going well. But I'm just wondering, in light of the macro, et cetera, if -- how you're thinking about pricing as -- in the context of the guidance that you've given for '23?

如果我能問一下約翰的快速跟進。我不認為你被問及定價環境，約翰。只是看著積壓再次上升，以及你正在談論的所有大趨勢，事情進展順利是可以理解的。但我只是想知道，從宏觀等方面來看，如果——你是如何考慮定價的——在你為 23 年給出的指導的背景下？

Thanks. Yes, I mean, we're very focused and disciplined on driving value for Cadence and for our customers. Like I said, on our hardware side, we don't believe that, and very much on the software side, too, that a lot of our customer spend with us is not really discretionary. It's quite indispensable tools that they need from us. And everybody these days wants to focus on improving productivity, and all the tools we provide help our customers to drive that. So I think we're in a sweet spot at the moment. So we're disciplined on pricing. But the pricing that we're extracting for our tools come from the increased value that our customers are getting from the use of our tools.

謝謝。是的，我的意思是，我們非常專注於為 Cadence 和我們的客戶創造價值。就像我說的，在我們的硬件方面，我們不相信，在軟件方面也非常相信，我們的很多客戶在我們這裡的消費並不是真正的自由裁量權。這是他們需要我們提供的不可或缺的工具。現在每個人都希望專注於提高生產力，而我們提供的所有工具都可以幫助我們的客戶推動這一目標。所以我認為我們目前處於最佳狀態。所以我們在定價方面很嚴格。但是我們為我們的工具提取的定價來自我們的客戶從使用我們的工具中獲得的增加的價值。

Yes, one thing to add to what John said, I think you know this already, but one thing to emphasize, like typically, when people move from one node to another, right -- so we are at 5-nanometer. A lot of the design is at 3, then 2, 1.4, 1. So we have like 10 years of node migration ahead of us. So any time you go from one node to the next node, the number of transistors effectively doubles in the chip. So number of blocks effectively doubles and complexity doubles. So for the same chip being done, it has more things in it. So that normally requires more use of our software.

是的，要補充約翰所說的一件事，我想你已經知道這一點，但要強調一件事，就像通常情況下，當人們從一個節點移動到另一個節點時，對 - 所以我們處於 5 納米。很多設計都在 3，然後是 2、1.4、1。所以我們前面有大約 10 年的節點遷移。因此，無論何時從一個節點轉到下一個節點，芯片中的晶體管數量實際上都會翻倍。所以塊的數量有效地加倍，複雜性加倍。所以同一個芯片在做，裡面的東西就多了。所以這通常需要更多地使用我們的軟件。

So that's one thing just to emphasize to our investors is that whenever there's a node transition -- and this is happening in the past, but this will continue to happen for next 10 years. So that also requires more hardware capacity. That requires more simulators. That requires more place and route runs to be done. And of course, with AI, we can make that even more productive. So -- and Moore's Law is not slowing down for at least 10 years, okay? And on top of that, you add 3D-IC. That adds another 10 years or so. So we have a lot of sustained growth that the customers will build amazing products, and they need more and more of our software and hardware to do that.

因此，要向我們的投資者強調的一件事是，每當有節點轉換時——這在過去發生過，但在未來 10 年將繼續發生。因此，這也需要更多的硬件容量。這需要更多的模擬器。這需要完成更多的佈局和佈線。當然，借助 AI，我們可以提高工作效率。所以——摩爾定律至少在 10 年內不會放緩，好嗎？最重要的是，您添加了 3D-IC。這又增加了 10 年左右。所以我們有很多持續的增長，客戶將製造出令人驚嘆的產品，他們需要越來越多的我們的軟件和硬件來做到這一點。

Yes, I guess that's where I was going with the question, I mean, obviously, with Cerebrus going -- running on top of Innovus. And I'm just kind of trying to think about how -- it seems, and you guys have talked about this, a greater portion of R&D budgets from both semiconductor and systems companies are going into kind of the suite of tools that you offer. So anyway, congrats on a great quarter.

是的，我想這就是我提出問題的地方，我的意思是，很明顯，Cerebrus 在 Innovus 之上運行。我只是想想想——看起來，你們已經談到了這一點，半導體和系統公司的大部分研發預算都用於你們提供的工具套件。所以無論如何，祝賀一個偉大的季度。

Our final question will come from Andrew DeGasperi from Berenberg.

我們的最後一個問題將來自 Berenberg 的 Andrew DeGasperi。

I guess first in terms of the analytics products that you mentioned, I was just wondering, in terms of your customers, how steep of a learning curve is it for them to adopt this? And what is the kind of time line that it typically takes from when you introduce it to them to when they're kind of using it in terms of their day-to-day processes?

我想首先就您提到的分析產品而言，我只是想知道，就您的客戶而言，他們採用該產品的學習曲線有多陡峭？從您將它介紹給他們到他們在日常流程中使用它通常需要什麼樣的時間線？

Yes, that's a very good point. So I think it's something like Cerebrus AI, our use model is very similar to what they are doing right now, okay? So we don't change any of the interfaces. And if they're used to running Innovus, then Cerebrus will run on top of it almost like a cockpit. But it will do -- those experiments that they're doing manually, they will do it in an automatic way. But the interfaces are similar. The commands they're used to running are similar. But instead of manually running and trying different things, Cerebrus will do that automatically.

是的，這是一個很好的觀點。所以我認為它類似於 Cerebrus AI，我們的使用模型與他們現在正在做的非常相似，好嗎？所以我們不更改任何接口。如果他們習慣於運行 Innovus，那麼 Cerebrus 將像駕駛艙一樣在它上面運行。但它會做——那些他們手動進行的實驗，他們將以自動方式進行。但是接口是相似的。他們習慣運行的命令是相似的。但是 Cerebrus 不會手動運行和嘗試不同的事情，而是會自動執行這些操作。

So typically, we have seen -- that's why you have seen a lot of kind of uptick to Cerebrus and these AI-based tools is because they don't fundamentally require a new working model. It just takes away some of the mundane tasks that the customers were doing. And similar things are true for Optimality, similar things are true for Verisium. There is always some learning gap, but it is -- our customers are very smart users anyway, right? They are designing all these complex chips. So this is nothing that they can't pick up in a -- pretty quickly.

所以通常情況下，我們已經看到——這就是為什麼你看到 Cerebrus 有很多提升，而這些基於 AI 的工具是因為它們從根本上不需要新的工作模型。它只是消除了客戶正在執行的一些平凡任務。 Optimality 也是如此，Verisium 也是如此。總是存在一些學習差距，但它是 - 我們的客戶無論如何都是非常聰明的用戶，對吧？他們正在設計所有這些複雜的芯片。所以這不是他們無法很快接受的。

That's helpful. And then one question for John. In terms of the share repurchases, I think you bought $1 billion or so last year. This year, you're guiding for a bit below that. I was just wondering, what was the thinking behind that? Is it a function of investing in the R&D in terms of the different -- the 3 different themes you mentioned earlier? Or is it potentially for additional M&A that you're planning for the year?

這很有幫助。然後有一個問題要問 John。就股票回購而言，我認為你去年購買了 10 億美元左右。今年，您的指導價略低於此。我只是想知道，這背後的想法是什麼？它是根據您之前提到的不同的 3 個不同主題投資研發的功能嗎？或者您是否有可能在今年計劃進行額外的併購？

Yes, Andrew, our approach for capital allocation hasn't changed and share repurchases that, effectively, for this year, again, we're expecting to use 50% of our free cash flow to repurchase shares. That's the way we started last year, too. We have a number of repurchase programs. We have our baseline repurchase program. We had an accelerated share repurchase last year to offset dilution from the general stock refresh in the middle of year because we do merit in the middle of the year. But also, we have an opportunistic repurchase program that kicks in when certain price levels are met. And last year, that kicked in, in Q1, Q2 and Q4. So we bought back shares when the prices were lower. We typically don't guide to that right now. But we assume, using at least 50% of free cash flow, to repurchase shares this year. If our opportunistic repurchase program kicks in, we'll buy back more.

是的，安德魯，我們的資本配置和股票回購方法沒有改變，今年，我們預計將再次使用 50% 的自由現金流來回購股票。這也是我們去年開始的方式。我們有許多回購計劃。我們有我們的基準回購計劃。去年我們加速了股票回購，以抵消年中一般股票更新帶來的稀釋，因為我們在年中表現出色。而且，我們還有一個機會主義的回購計劃，當達到特定價格水平時，該計劃就會啟動。去年，這在第一季度、第二季度和第四季度開始出現。所以我們在價格較低的時候回購了股票。我們現在通常不會對此進行指導。但我們假設，今年至少使用 50% 的自由現金流回購股票。如果我們的機會主義回購計劃啟動，我們將回購更多。

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

我現在將把電話轉回給 Anirudh Devgan 作結束語。

Thank you, all, for joining us this afternoon. It's an exciting time for Cadence as we enter 2023 with strong business momentum and robust design activity offering tremendous market opportunities. Our exceptional execution of the Intelligent System Design strategy, customer-first mindset and a high-performance and inclusive culture are driving accelerating growth as we grow our core EDA business while expanding our portfolio with new innovative solutions. Fostering sustainable innovation is a top priority, and we are thrilled to have been included in the newly released 2023 top-rated ESG Company list, ranking #18 out of over 1,000 companies in the software and services group. And on behalf of our employees and our Board of Directors, we thank our customers, partners and investors for their continued trust and confidence in Cadence.

謝謝大家今天下午加入我們。隨著我們進入 2023 年，強勁的業務勢頭和強勁的設計活動提供了巨大的市場機會，這對 Cadence 來說是一個激動人心的時刻。隨著我們發展核心 EDA 業務，同時通過新的創新解決方案擴展我們的產品組合，我們對智能係統設計戰略的卓越執行、客戶至上的心態以及高性能和包容性文化正在推動加速增長。促進可持續創新是重中之重，我們很高興被列入新發布的 2023 年 ESG 頂級公司名單，在軟件和服務組的 1,000 多家公司中排名第 18 位。我們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 in today's Cadence Fourth Quarter and Fiscal Year 2022 Earnings Conference Call. This concludes today's call. You may now disconnect.

感謝您參加今天的 Cadence 第四季度和 2022 財年收益電話會議。今天的電話會議到此結束。您現在可以斷開連接。