CDNS earnings call summary

Prepared Highlights

• Q3 2025 total revenue $1.339B (prepared); full-year 2025 revenue guidance raised to ~14% growth (CEO)
• Non-GAAP operating margin Q3: 47.6%; GAAP operating margin Q3: 31.8% (CFO)
• Non-GAAP EPS Q3: $1.93; GAAP EPS Q3: $1.05 (CFO)
• Backlog grew to over $7B (CEO); bookings exceeded expectations (CEO)
• IP business strong momentum driven by AI/HPC/automotive; won HBM4 and DDR5 IP engagement at a marquee memory company (CEO)
• Core EDA: Cadence Cerebrus AI Studio helped Samsung achieve 4x productivity on SF2; Samsung used Cadence tools for 22% power reduction on multibillion instance AI design (CEO)
• Hardware verification had record Q3, especially with AI/HPC customers; OpenAI expanded commitment to Palladium (CEO)
• System design/analysis: three major memory providers increased Clarity use, displacing competitive solutions; Infineon standardized PCB design on Allegro X (CEO)
• Signed definitive agreement to acquire Hexagon’s D&E/MSC software business (CEO)
• Q3 operating cash flow $311M; cash balance $2.753B; debt $2.5B (CFO)
• DSOs 55 days; used $200M to repurchase shares in Q3 (CFO)

Official Guidance

• Q4 2025: Revenue $1.405B–$1.435B; GAAP operating margin 32.5%–33.5%; non-GAAP operating margin 44.5%–45.5%; GAAP EPS $1.17–$1.23; non-GAAP EPS $1.88–$1.94 (CFO)
• Full-year 2025: Revenue $5.262B–$5.292B; GAAP operating margin 27.9%–28.9%; non-GAAP operating margin 43.9%–44.9%; GAAP EPS $3.80–$3.86; non-GAAP EPS $7.02–$7.08; operating cash flow $1.65B–$1.75B; expects to use at least 50% of annual FCF to repurchase shares (CFO)

Mgmt Quotes

• “Cadence delivered excellent results for the third quarter of 2025, with strong operational and financial performance across all product categories and geographies.” (CEO)
• “The accelerating AI megatrend is fueling an unprecedented wave of design activity across industries.” (CEO)
• “With the updated outlook and at the midpoint, we now expect our 2025 revenue to grow approximately 14% year-over-year.” (CFO)
• “We deepened our long-standing partnership with a leading semiconductor company… and are closely collaborating on next-generation agentic AI EDA solutions.” (CEO)
• “Bookings were strong, resulting in a backlog of $7 billion.” (CFO)

Prepared Metrics

Q&A Batch (1-5 of 14)

Q1 — Vivek Arya

• Topic: IP business growth drivers and sustainability
• Key points:
  • IP business tracking to over 20% growth for the second year; Q3 was exceptional.
  • Growth driven by design IP focused on AI, HPC, and chiplet-based architectures (SerDes, PCIe, HBM4 IPs).
  • Multiple foundries at leading nodes (TSMC, Samsung, Intel, Rapidus) and improved PPA driving customer demand.
• Mgmt stance: Bullish — optimistic about IP business; expects it to grow better than Cadence average next year, given profitability profile.

Q2 — Jason Celino

• Topic: Q3 backlog strength and Q4 renewal outlook
• Key points:
  • Q3 performance better than expected, driven by accelerating AI infrastructure build-out across all geographies.
  • Core EDA and IP backlog weighted toward multiyear recurring arrangements, supporting durable double-digit growth.
  • Strong demand continuing from Mag 7, internal chip design, and merchant silicon companies (NVIDIA, Broadcom, AMD).
• Mgmt stance: Bullish — sees strong demand continuing; mix healthy across EDA, IP, hardware, and SDA.

Q3 — Joseph Vruwink

• Topic: 2026 expectations based on current visibility
• Key points:
  • All 5 lines of business expected to grow double digits in FY 2025; performance strong across all geographies.
  • Exiting FY 2025 with record backlog and broad-based momentum from strategic partnerships.
  • Framework remains disciplined: double-digit top line ambition, continued operating leverage, balanced capital allocation.
• Mgmt stance: Neutral — prudent on FY 2026 outlook; will provide update in January/February; positioned better than last several years.

Q4 — Lee Simpson

• Topic: China revenue strength and drivers
• Key points:
  • China up ~53% year-on-year, now 18% of mix; Q3 only slightly better than expected.
  • Region returned to business as usual after export regulation changes for EDA lifted in early July; Q2 disruption made up with hardware deliveries in Q3.
  • Design activity strong across China; no unusual pull-in from future quarters observed.
• Mgmt stance: Neutral — China back to normal; broad-based strength not tied to any one geography; difficult to predict future.

Q5 — Sitikantha Panigrahi

• Topic: System design strategy (BETA CAE, MSC software acquisitions)
• Key points:
  • SD&A business run rate expected to cross $1 billion in 2026 if Hexagon acquisition closes.
  • Two pillars: (1) 3D-IC/chiplets (Allegro, Sigrity, Clarity, Celsius) driven by HPC; (2) physical AI (structural analysis, multibody dynamics, CFD) for cars, drones, robots.
  • Investment split: ~70-80% in horizon 1 (infrastructure AI), ~20% in horizon 2 (physical AI), few percent in horizon 3 (sciences AI).
• Mgmt stance: Bullish — optimistic about positioning for physical AI; simulation essential for training world models in robotics.

Q&A Batch (6-10 of 14)

Q6 — James Schneider

• Topic: AI功能对EDA产品的生产率提升与货币化路径
• Key points:
  • SimAI实现5至10倍逻辑仿真效率与覆盖率提升（客户包括Samsung、Qualcomm、NVIDIA）
  • Cerebrus AI Studio为Samsung带来4倍生产率提升与22% PPA改进（对标5→3nm节点迁移通常仅10-20% PPA改善）
  • AI工具已成为行业事实标准，货币化通常需要2个合同周期
  • 芯片设计工作负载呈指数增长（5年后芯片规模5-10x，复杂度20-30x），AI生产率是必需
• Mgmt stance: 强烈看多（认为AI是下一10倍自动化动力，客户主动采用，长期价值显著）

Q7 — Harlan Sur

• Topic: 第三代硬件升级周期（Palladium/Protium）需求曲线与2026年展望
• Key points:
  • 硬件业务表现强劲，2026年收入预计优于2025年，但仅保持6个月可见性，指引偏审慎
  • 自研全掩模芯片（TSMC），144液冷芯片通过InfiniBand与光互联，16机架可仿真1万亿晶体管设计
  • 需求主要来自AI、HPC、汽车市场；公司正扩大制造产能并改善交期，Q3已主动增建库存
  • 硬件毛利率已变得更健康
• Mgmt stance: 中性偏乐观（技术趋势与竞争地位极强，但因可见性限制而采取谨慎指引）

Q8 — Jay Vleeschhouwer

• Topic: NVIDIA与Intel合作对EDA需求及投资的影响
• Key points:
  • CPU+GPU组合是EDA优化方向（公司已提前开始移植，如Millennium平台）
  • 该合作带来x86+GPU新平台，同时推进与Intel的对话（进展顺利）
  • 具体设计活动涉及NVLink-based IP，公司正与两家企业合作
  • 工具在CPU+GPU上优化后，客户采纳度将提升
• Mgmt stance: 看多（认为行业向CPU+GPU趋势有利，合作带来具体设计需求与工具性能提升）

Q9 — Gianmarco Conti

• Topic: 中国区可持续增长与出口管制风险
• Key points:
  • 中国设计活动已恢复正常，预计2025年全年增长，但具体幅度难预测
  • Q3表现强劲源于中国区贡献，但整体基础广泛（AI基建、物理AI等），非单一国别驱动
  • Q4及全年指引假定当前出口管制制度基本不变；公司纳入审慎波动因素
  • 两国高层对话带来稳定性，有利于业务
• Mgmt stance: 谨慎乐观（中国反弹明显，但地缘政治不确定性仍存在，公司以保守假设做规划）

Q10 — Joseph Quatrochi

• Topic: OpEx动态（Q3优于预期、Q4弱于预期）
• Key points:
  • Q3优于预期部分来自小规模重组收益及硬件毛利率健康
  • Q4弱于预期主要受新收购（如Artisan）带来额外费用，以及部分硬件交付从Q3移至Q4
  • 全年整体略高于预期：核心EDA软件、硬件、SDA、IP均表现强劲
• Mgmt stance: 中性（波动源于时间性因素与收购整合，整体业务执行稳健）

Q&A Batch (11-14 of 14)

Q11 — Yu Shi

• Topic: Hardware and China growth momentum into next year; hardware refresh cycle risk
• Key points:
  • 公司过去3年整体增长率维持在40%左右；今年预收款（upfront revenue）增速可能接近50%，中国增速预计高于公司平均。
  • 硬件需求强劲，未来6个月管道非常强，正在为未来几个季度的大订单增加库存；Q3 backlog达70亿美元（创纪录），预计Q4 backlog将再创新高。
  • 当前Z3/X3硬件支持1万亿晶体管设计（目标2030年），但下一代硬件将在2030年前推出；客户购买行为已变为几乎每年购买（类似订阅），而非4-5年前的模式。
• Mgmt stance: 乐观——管理层认为硬件需求无系统性下降风险，2026年硬件收入将强于2025年，且公司持续将35%收入投入研发（65%费用在研发，25%在应用工程）。

Q12 — Gary Mobley

• Topic: 中国出口管制解除对backlog的影响
• Key points:
  • backlog从Q2的64亿美元增至Q3的70亿美元，增长6亿美元。
  • 其中约25%（约1.5亿美元）来自Q2至Q3的中国 backlog 补回（catch-up），其余增长来自业务整体强劲。
• Mgmt stance: 中性——管理层确认数字，未对后续中国业务趋势做额外判断。

Q13 — Clarke Jeffries

• Topic: AI定制芯片（custom silicon）对IP业务盈利能力和投资计划的影响
• Key points:
  • 公司45%业务来自系统公司，55%来自半导体公司；AI加速了系统公司定制芯片趋势（如Google、Meta、Microsoft、Amazon、Tesla等）。
  • 这些大客户预计AI计算需求未来几年每年翻倍，因此愿意投资EDA、IP和硬件；公司保持定价纪律，盈利能力与以往类似。
  • 定制芯片对推理（inference）的财务和定制化收益显著，推动客户内部芯片设计投入增加。
• Mgmt stance: 乐观——管理层认为定制芯片趋势健康且加速，公司受益于客户内部设计投入增长，盈利能力不受影响。

Q14 — Ruben Roy

• Topic: 代理型AI（agentic AI）解决方案的合作与商业化
• Key points:
  • 公司有5大AI平台，代理型AI用于自动化芯片设计中尚未自动化的部分（如RTL编码，占工作流约10%）。
  • 通过JedAI平台部署，部分标准化（数据库、模型、工具接口），部分客户定制化（数据本地化，支持本地和云端部署）。
  • 已有多项合作，未来将推出更多产品，重点自动化RTL开发和验证计划生成等重复性任务。
• Mgmt stance: 乐观——管理层认为代理型AI能显著提升自动化水平，JedAI平台架构支持通用和定制化部署，前景乐观。