CDNS

Cadence Design Systems, Inc.

AI '24: AI Software AI '25: Chip Back-End, Packaging & Design AI '25: Chip Mfg & Tooling AI '26: Mizuho Most AI-Resilient Software AI '26: Processor IP AI '26: Semi Yield Analytics & Software Humanoid '25: Compute & Edge AI Memory '26: CXL / Memory Expansion Regulatory Testing '26: Compliance Software

Loading…

Overview

Cadence provides software, hardware, and design blocks used to create advanced microchips and electronic systems. Their tools automate design and testing to ens

Cadence provides software, hardware, and design blocks used to create advanced microchips and electronic systems. Their tools automate design and testing to ensure chips work correctly. About 80% of revenue comes from recurring software subscriptions, while 20% is from hardware and design assets. Key customers include semiconductor leaders like NVIDIA and Broadcom, plus major global cloud and automotive providers.

What They Do (Plain English & Analogies): Cadence provides the 'digital architect' tools and 'virtual wind tunnels' used to design the world's most complex electronics. Think of them as providing the AutoCAD for microchips: because modern chips have billions of microscopic components, they are impossible to design by hand. Cadence provides the software (EDA) to draw and simulate these chips, the 'Lego blocks' (IP) of pre-designed circuitry to speed up the process, and massive supercomputers (Palladium/Protium) that act as a 'flight simulator' to test if a chip works before a company spends $100M+ to manufacture it. Their new 'Agentic AI' acts like a virtual senior engineer that can autonomously write code and debug designs, further accelerating the process.
Very Brief History: Cadence was formed in 1988 through the merger of SDA Systems and ECAD, Inc. It established itself as a leader in Electronic Design Automation (EDA) for the semiconductor industry. Over the last decade, under the 'Intelligent System Design' strategy, the company evolved beyond just chip software into system-level analysis (thermal, fluid dynamics) and hardware-software co-verification. In 2024-2025, it pivoted heavily into AI-driven design, launching the JedAI platform and acquiring BETA CAE to expand into physical simulation for automotive and aerospace.
"Street Stereotype": Cadence is viewed as a 'high-quality compounder' and a premier 'picks and shovels' play for the AI era. Investors see it as part of an entrenched EDA duopoly (alongside Synopsys) with massive barriers to entry, high switching costs, and a highly predictable recurring revenue model. It is often perceived as a 'safe' way to play semiconductor growth because it profits from the *activity* of designing new chips, regardless of whether those chips eventually sell well in the consumer market.
Subsidiaries On Linked In*: BETA CAE Systems, Tensilica, Sigrity, AWR Corporation, Jasper Design Automation, Inspectiv, and (pending acquisition) Hexagon's Design & Engineering business.
Customer Sectors & Example Clients: Primary sectors include Semiconductors, Hyperscale Computing, Automotive, Aerospace/Defense, and Industrial Healthcare. Specific marquee clients mentioned or inferred include NVIDIA, Broadcom, Samsung, TSMC, Intel, Qualcomm, Apple, Altera, Tenstorrent, and major hyperscalers like Amazon (AWS), Google, and Microsoft.
New Customers / Segments They'Re Targeting: Cadence is aggressively targeting 'System Companies' (Hyperscalers and Automotive OEMs) that are now designing their own custom silicon (Customer Owned Tooling or COT) rather than buying off-the-shelf chips. They are also moving into 'Physical AI,' targeting robotics and autonomous vehicle companies that need to simulate how chips, sensors, and mechanical systems interact in the real world (accelerated by the Hexagon D&E acquisition).
How Key Themes May Help/Hurt: The build-out of AI and Motion Control (Physical AI) is a massive tailwind. As robots and autonomous systems require more sensors and complex 'brains,' the design workload grows exponentially. Cadence benefits because its tools are required to manage this complexity. However, increased complexity also requires Cadence to spend more on its own R&D and compute power. Export controls on high-end chip design software to China remain a persistent regulatory risk that could cap growth in that region.

3 Main Long-Term Bull Details

AI Complexity Multiplier: As chips move toward 1-trillion transistors and multi-chiplet designs, the 'workload' for EDA tools grows exponentially, allowing Cadence to capture more value per design. 2) Agentic AI Monetization: The shift from selling 'seats' for human engineers to selling 'AI Agents' (like ChipStack) opens a new high-margin revenue stream where Cadence is paid for productivity gains. 3) System Company Expansion: The trend of non-semiconductor companies (Tesla, Meta, Amazon) building in-house chip teams creates a massive new customer base that is less cyclical than traditional chipmakers.

3 Main Long-Term Bear Details

Geopolitical/Export Risk: Tightening U.S. restrictions on EDA software and hardware sales to China could permanently impair a significant portion of Cadence's growth engine. 2) M&A Integration: The move into physical simulation (BETA CAE, Hexagon) pits Cadence against established giants like Dassault Systèmes and requires successful integration of very different software cultures. 3) Valuation Sensitivity: CDNS typically trades at a high P/E multiple; any slowdown in the AI infrastructure build-out could lead to significant multiple compression even if the business remains healthy.

Competitors And Differentiation: The primary competitor is Synopsys (SNPS), followed by Siemens EDA (Mentor Graphics) and Ansys (in simulation). Cadence differentiates through its '3-layer cake' framework: a superior hardware-software 'Dynamic Duo' (Palladium/Protium) for emulation, a lead in 'Agentic AI' with the ChipStack super agent, and a unified data platform (JedAI) that allows AI to learn across different design stages better than fragmented competitor tools.
Recent Performance & What The Market'S Focused On: Cadence finished FY 2025 with 14% revenue growth and a record $7.8 billion backlog. The market is currently laser-focused on the rollout of 'ChipStack AI,' the world's first Agentic AI for chip design, and the re-acceleration of the recurring software business to double-digit growth. Investors are also watching the hardware segment, which saw record demand from hyperscalers using Palladium systems to test next-gen AI chips before production.
Brands And Revenue Segments: Revenue is categorized into: Digital IC Design & Signoff (approx. 25-30%), Custom IC Design & Simulation (approx. 20-25%), Functional Verification (including Palladium/Protium hardware, approx. 20-25%), IP (Intellectual Property, approx. 12-15%), and System Design & Analysis (approx. 12-15%). Key brands include Palladium, Protium, Virtuoso, Allegro, Spectre, Verisium, and Cerebrus.

Bull / Bear Details

As of February 19, 2026, Cadence remains a top-tier play on AI silicon complexity and the shift toward custom hyperscaler silicon. The launch of ChipStack AI Su

Thesis

As of February 19, 2026, Cadence remains a top-tier play on AI silicon complexity and the shift toward custom hyperscaler silicon. The launch of ChipStack AI Super Agent and the expansion into Physical AI via the Hexagon acquisition broaden its TAM beyond traditional EDA. With a record $7.8 billion backlog and reaccelerating double-digit recurring revenue, Cadence's '3-layer cake' strategy effectively monetizes the exponential increase in design iterations, making the bull case for sustained double-digit growth highly compelling.

Bull case

Cadence is pioneering 'Agentic AI' with the launch of ChipStack AI, which automates previously manual RTL design and verification tasks, delivering up to 10x productivity gains. This 'force multiplier' increases the volume of design experiments, driving higher usage of underlying simulation and optimization tools. As AI models move toward 1 trillion transistors, Cadence's AI-driven flows become indispensable for managing exponential design complexity and time-to-market pressures.
The structural shift of hyperscalers toward Customer Owned Tooling (COT) and 3D-IC multichip architectures significantly expands Cadence's footprint. Major cloud providers are standardizing on Cadence's full flows to design proprietary AI training and inference chips. This transition, combined with a record $7.8 billion backlog and 25% growth in the IP business, provides high visibility into high-margin recurring revenue as custom silicon becomes a competitive necessity.
Cadence is aggressively expanding into 'Physical AI' and system analysis through the acquisition of Hexagon's D&E business and BETA CAE, targeting the autonomous driving and robotics markets. Meanwhile, the hardware business continues to hit record levels, with the 'Dynamic Duo' (Palladium/Protium) seeing massive repeat demand from AI leaders. This diversification into system-level simulation ensures Cadence captures value across the entire lifecycle of intelligent physical systems.

Bear case

While hardware reached record levels in 2025, it remains an upfront revenue source prone to cyclicality and digestion periods. Management's 'prudent' outlook for the second half of 2026 reflects limited visibility into the hardware pipeline beyond two quarters. If AI infrastructure build-outs normalize or if customers face budget constraints, the high-margin hardware segment could see a sharp deceleration, challenging the company's aggressive top-line growth targets.
Geopolitical tensions and evolving export controls remain a persistent threat, particularly regarding high-end EDA tools and IP sales in China. China accounted for 13% of 2025 revenue, and any tightening of regulations could disrupt this significant growth engine. Furthermore, the reliance on a few marquee hyperscalers and semiconductor giants creates concentration risk; any delays in their internal chip programs would directly impact Cadence's RPO and backlog consumption.
The pending acquisition of Hexagon's D&E business introduces integration risk and potential near-term margin dilution. While Cadence targets 'Physical AI,' competing with industrial incumbents requires heavy R&D. Furthermore, the shift toward usage-based pricing for AI agents could introduce revenue volatility compared to traditional multi-year subscriptions. If startups disrupt specific automation niches, Cadence may struggle to maintain its 45% operating margin while funding expansion into new, non-core verticals.

Bull / Bear Case

Bear Case: Cadence's valuation remains the primary risk, as the stock frequently trades at a Forward P/E above 40x, a significant premium to the broader software sector. While 2025 was a record year for hardware, management's 'prudent' guidance for the second half of 2026 suggests a looming digestion phase after a massive infrastructure build-out. Geopolitical headwinds remain a critical concern, with China accounting for 13% of revenue and facing ongoing export control uncertainty. Furthermore, the expansion into 'Physical AI' through the Hexagon acquisition introduces integration risks and pits Cadence against entrenched industrial software incumbents, potentially diluting margins. If the shift to usage-based pricing for AI agents introduces revenue volatility or if hyperscalers slow their internal chip programs due to cost pressures, Cadence's high-multiple valuation could face a severe de-rating.
Bull Case: Cadence is successfully pivoting from a traditional EDA provider to a 'Physical AI' and 'Agentic AI' leader. The launch of ChipStack AI Super Agent represents a structural shift in monetization, moving from selling tools to providing 'virtual engineers' that deliver 10x productivity gains. This 'force multiplier' increases design exploration, which in turn drives higher usage of Cadence's core simulation and hardware engines. The bull case is further bolstered by the massive shift of hyperscalers toward Customer Owned Tooling (COT) and 3D-IC architectures, where Cadence is gaining significant share. With a record $7.8 billion backlog and the IP business growing at 25%, Cadence is capturing the lion's share of the AI silicon complexity boom while expanding its TAM into robotics and autonomous systems via the Hexagon D&E acquisition.
More Compelling & Why: Bull. Despite a high Forward P/E of approximately 42x (based on the $8.10 mid-point EPS guidance), the investment is anchored by a record $7.8 billion backlog that provides 67% visibility into 2026 revenue. The strongest argument is the reacceleration of recurring software revenue to double-digit growth, proving that AI is an accelerant rather than a disruptor to EDA. The 'Agentic AI' layer creates a new, high-margin revenue stream that captures value from the exponential increase in design iterations. I would flip to Bear if the 'current' portion of the RPO (Remaining Performance Obligations) stagnates below $3.8 billion, signaling that the hardware cycle has peaked and the AI design tailwind is normalizing.

Key Factors5 rows

Key Factor	Why It Matters	What To Watch	What It Signals	Where/How To Track	Free Alt Data	Paid Alt Data
ChipStack AI Super Agent Adoption and Monetization	Launched in Feb 2026, ChipStack is the first 'Agentic AI' solution for RTL design/verification. It targets manual tasks, creating a new 'virtual engineer' revenue stream. Success here proves AI is a force multiplier for EDA usage rather than a threat to seat counts.	Watch for press releases regarding 'standardization' or 'broad deployment' of ChipStack by marquee customers mentioned in the transcript (NVIDIA, Broadcom, Qualcomm, Altera, Tenstorrent).	Announcement of >2 additional top-10 semiconductor customers standardizing on ChipStack = Bullish; Lack of new customer mentions by Q2 2026 = Bearish (suggests slow ROI validation for Agentic AI).	Cadence Newsroom (cadence.com/go/news); Company presentations at industry conferences like DAC (Design Automation Conference) in June 2026.	Google Trends: Search volume for 'Cadence ChipStack' or 'Agentic AI EDA'; LinkedIn: Tracking job postings requiring 'ChipStack' or 'JedAI' skills.	Thinknum: Tracking 'Agentic AI' or 'ChipStack' mentions in global semiconductor job descriptions.
Backlog Consumption and RPO Growth ($7.8B Baseline)	Cadence entered 2026 with a record $7.8 billion backlog, with 67% of 2026 revenue already covered by beginning backlog. Tracking the conversion of this backlog, especially in the 'upfront' hardware business (Palladium/Protium), is critical for hitting the $5.9B-$6.0B annual revenue guide.	Monitor Remaining Performance Obligations (RPO) in the Q1 and Q2 10-Q filings. Watch for the 'current' portion of RPO (to be recognized within 12 months) to stay above $3.8 billion to ensure 2026 targets are de-risked.	RPO growth exceeding $8.0B by Q2 2026 = Bullish (indicates sustained multi-year demand); RPO falling below $7.5B or 'current' RPO stagnating = Bearish (signals hardware digestion or slowing design starts).	SEC Filings (10-Q/10-K), specifically the 'Revenue' or 'Backlog' notes; CFO Commentary documents released on earnings dates (approx. April and July 2026).	SEC EDGAR: Search for 'Remaining Performance Obligations' in CDNS 10-Q filings.	Bloomberg Terminal: {CDNS US Equity MODL} for backlog and RPO consensus tracking.
Hyperscaler COT (Customer Owned Tooling) Transition	Hyperscalers are moving from using merchant silicon to designing their own chips (COT). A 'marquee hyperscaler' recently adopted Cadence's full digital flow for its first COT chip. This shift significantly increases the footprint for EDA, IP, and hardware systems per customer.	Watch for news of additional hyperscalers (Amazon, Google, Microsoft, Meta) moving from 'ASIC' models to 'Full Flow' or 'COT' models with Cadence. Specifically, watch for 3D-IC platform adoption for AI training chips.	Second marquee hyperscaler adopting 'Full Flow' for internal silicon = Bullish; News of hyperscalers delaying internal chip programs due to cost/complexity = Bearish.	Quarterly Earnings Calls (Q1 2026 expected April 2026); Industry news sites like SemiWiki or EE Times for 'tape-out' reports.	ImportGenius/Panjiva: Tracking shipments of Palladium/Protium hardware racks to hyperscale data center addresses.	Gartner/IDC: Reports on custom silicon (ASIC/COT) market share and EDA spend by system companies.
IP Business Growth (HBM4 and LPDDR6 Milestones)	IP revenue grew ~25% in 2025, driven by AI infrastructure. Cadence is gaining share in high-performance memory IP (HBM, LPDDR6). As AI chips move to 1 trillion transistors, the royalty and licensing potential for these 'STAR IP' titles increases.	Announcements of 'Industry First' silicon-proven LPDDR6 or HBM4 IP titles. Watch for IP segment revenue growth in Q1/Q2 to stay above 20% y/y.	IP revenue growth >20% y/y in Q1 2026 = Bullish (confirms share gains); IP growth falling below 15% = Bearish (suggests Synopsys or internal alternatives are reclaiming share).	Quarterly CFO Commentary (Segment Revenue table); TSMC Ecosystem partner announcements (OIP Ecosystem).	TSMC Press Center: Search for 'Cadence' in 'Partner of the Year' or 'Interface IP' awards.	IPnest: Annual Design IP Market Share Report (typically updated in Q2) to verify CDNS share gains vs. SNPS.
Hexagon D&E Acquisition Closing and Guidance Update	The pending acquisition of Hexagon's Design & Engineering business is the cornerstone of Cadence's 'Physical AI' strategy, targeting robotics and autonomous driving. Management excluded this from the initial 2026 guide, representing a material inorganic upside.	Closing date of the Hexagon D&E acquisition (expected mid-2026). Watch for the subsequent 'Revenue' and 'Non-GAAP EPS' guidance raise in the following earnings call.	Deal close + 2026 revenue guide raise of >$150M = Bullish; Regulatory delays or a guide raise <$100M (suggesting high churn/integration issues) = Bearish.	SEC Form 8-K (Completion of Acquisition); Press releases on Cadence Investor Relations site.	Federal Trade Commission (FTC) or EU Competition Commission: Monitoring for antitrust clearance updates.	Capitol Forum: Analysis of regulatory hurdles for cross-border software M&A.

Key Reported Metrics3 rows

Metric	Why It Matters	Last Period
Core EDA Revenue	Core EDA is the foundational software engine (Genus, Xcelium, Palladium/Protium, 3D-IC tooling) powering customers' silicon design. A 13% YoY rise in 2025 demonstrates sustained demand for flagship tools and AI-enabled design flows, informing how well Cadence can monetize next-gen workloads.	13%
IP Revenue	IP revenue growth reflects the value of Cadence's expanding library and memory/IP portfolio (HBM, UCIe, PCIe, DDR, SerDes) and royalty/licensing potential. With ~25% YoY growth in 2025, this segment is a key driver of recurring revenue and strategic moat as AI demand climbs.	~25%
Total Revenue	Total revenue is the core top-line indicator of Cadence's demand across the full stack (EDA software, IP, hardware). The 2025 full-year growth of 14% signals durable demand and AI-enabled expansion; near-term Q1 guidance will test continuation of this momentum as AI workloads scale.	14%

Key Questions

Can Cadence successfully monetize its new "Agentic AI" solutions, specifically ChipStack, through incremental usage-based and outcome-oriented pricing without c

Can Cadence successfully monetize its new "Agentic AI" solutions, specifically ChipStack, through incremental usage-based and outcome-oriented pricing without cannibalizing its core recurring software revenue?

Question 2: Will the hardware business (Palladium and Protium) sustain its record growth momentum throughout 2026, or will the "prudent" second-half guidance lead to a digestion phase following the massive 2025 cycle?
Question 3: How effectively will Cadence capture the "Physical AI" opportunity in robotics and autonomous driving as it integrates the Hexagon D&E acquisition and expands its System Design and Analysis (SD&A) footprint?

Earnings Transcript SummaryTable

2026-02-17 · 2025Q4 Earnings Call

3 Things Management Is Most Focused On	Call Takeaway & Tone	Prior Quarter'S Y/Y Growth By Segment	3 Things Analysts Most Pressed On (And Mgmt Responses)	Revenue Segments
1. Agentic AI Leadership: Launching 'ChipStack AI,' the world's first Agentic AI solution for chip design, which management claims provides 10x productivity gains and acts as a 'force multiplier' for their underlying tools. 2. AI Infrastructure & Hyperscale Partnerships: Deepening collaborations with NVIDIA, Broadcom, and marquee hyperscalers who are standardizing on Cadence's full flows for next-generation AI, training, and inference chips. 3. Physical AI & System Analysis: Expanding the 'System Design and Analysis' footprint into autonomous driving and robotics through the integration of BETA CAE and the pending acquisition of Hexagon's D&E business.	Takeaway: Cadence delivered a strong finish to 2025, characterized by a record backlog and a clear strategic pivot toward 'Agentic AI.' The company is successfully monetizing the complexity of AI silicon design and expanding into the broader 'Physical AI' market (robotics/automotive). Tone: Highly confident and bullish, with management emphasizing share gains across all major product segments and a robust 2026 outlook.	Based on Q3 2025 results: Core EDA: ~12% y/y; IP: ~22% y/y; System Design and Analysis (SD&A): ~11% y/y; Total Revenue: 19% y/y. Comparison: Core EDA and SD&A saw slight y/y acceleration in Q4, while IP maintained strong momentum; total revenue growth decelerated from 19% to 14% due to the timing of hardware deliveries.	1. AI Disruption/Cannibalization: Analysts asked if AI would reduce the need for EDA tools. Mgmt responded that AI is not a replacement but an accelerant that increases the number of design iterations and 'calls' to base tools, actually driving higher usage. 2. Hardware Cycle Sustainability: Analysts questioned if the hardware ramp (Palladium/Protium) could stay strong in year two of the platform. Mgmt cited a record $7.8B backlog and stated that 7 of the top 10 customers are now 'Dynamic Duo' (hardware + software) users. 3. Monetization of AI: Analysts asked how AI agents would be priced. Mgmt explained they will maintain the multi-year subscription anchor but add usage-based pricing for incremental capacity and outcome-oriented packages for specific productivity gains.	For FY 2025: Core EDA: 13% y/y; IP: ~25% y/y; System Design and Analysis (SD&A): 13% y/y; Total Revenue: 14% y/y. Note: Recurring software business reaccelerated to double-digit growth in Q4.

Transcript TidbitsTable

2026-02-17

About Expanding Eligible Market	About Competition	About The Broader Industry	Where Things Are Headed	Updates On Theme	Broader Themes Emerging	Bullish-Leaning Quotes (Short)	Bearish-Leaning Quotes (Short)	Hiring
Cadence is expanding into 'Physical AI' through the acquisition of Hexagon's D&E business, targeting autonomous driving and robotics. The launch of ChipStack AI Super Agent opens a new market for automating RTL design and verification, tasks that were previously manual. Additionally, the shift of hyperscalers toward internal silicon (Customer Owned Tooling) and the transition to 3D-IC multichip architectures are significantly increasing the footprint for EDA, IP, and hardware systems.	Management stated they are 'taking share in all major product segments,' including hardware, IP, and core EDA. They highlighted 'meaningful competitive wins' in the IP business and claimed a first-to-market advantage with Agentic AI solutions. While they monitor AI-focused startups, they believe their R&D scale (10,000 engineers) and deep integration into customer workflows provide a significant moat.	The semiconductor industry is accelerating, with projections suggesting it could hit $1 trillion in revenue this year, four years ahead of previous 2030 estimates. AI infrastructure build-out is in 'full swing,' driving exponential increases in design complexity as chips move toward 1 trillion transistors. There is a structural shift toward custom silicon and multichip architectures (3D-IC) to meet performance and power demands.	2026 is expected to be another record year for hardware, supported by a record $7.8 billion backlog. The company is moving toward 'Agentic AI' workflows where autonomous agents call underlying tools, acting as a force multiplier for productivity. Monetization is shifting toward pricing 'virtual engineers' or agents, while the core recurring software business has reaccelerated to double-digit growth.	Software:	Physical AI (applying AI to physical systems like robotics and autonomous vehicles); Agentic AI workflows (autonomous agents performing complex engineering tasks); Proliferation of Hyperscaler COT (Customer Owned Tooling).	Finished 2025 with a record backlog of $7.8 billion.; Taking share in all our major product segments.; Semi industry might hit $1 trillion this year... 4 years ahead of plan.; Expect this year to be yet another record year for hardware.	Quite prudent in the second half of the year [for hardware].; China visibility is near term in the first half.; Assumption that export control regulations... remain substantially similar.; We typically try to derisk the guide for things like hardware and China.

NotesTable

Date	Comment	Comment Type	Comment Sentiment	Link	IS CHANGE	Price Reaction
2026-02-17	Cadence's record $7.8 billion backlog and the launch of ChipStack—the first Agentic AI solution for chip design—drove a 7.6% stock surge. Investors cheered the 25% IP growth and accelerating recurring revenue, signaling clear monetization of AI workloads. Management's robust 2026 guidance confirms that AI complexity is expanding the EDA market, with hyperscalers increasingly standardizing on Cadence's full-flow solutions.	Earnings Transcript	Bullish		False	+7.60% (vs SPY: +7.10%)