Cadence Deep Dive: System-Level Multiphysics EDA and the Millennium M2000 Supercomputer Anchor Physical AI, with AuraStack Agentic AI and Fidelity GPU CFD as Key Levers
Cadence (CDNS) enters physical AI via system-level multiphysics EDA and the Millennium M2000 supercomputer: AuraStack agentic AI, Fidelity GPU CFD and Clarity/Celsius/Nastran multiphysics signoff, on $5.2968B 2025 revenue (+14%).
Executive Summary
Cadence did something in 2026 that no one had done before: it placed chip design and structural, acoustic and multibody dynamics inside the product matrix of a single company. On 23 February 2026 it completed the acquisition of Hexagon's Design and Engineering business for about 2.7 billion euros, including roughly 150 million euros of transaction-related taxes, structured as 70 percent cash and 30 percent stock. The deal brought MSC Nastran, Adams and the AMendate, Romax, CADLM and CAEfatigue assets. Combined with the BETA CAE pre- and post-processing capability absorbed in 2025, Cadence could claim end-to-end multiphysics for the first time.
This report makes five arguments.
First, Cadence physical AI runs on two wheels with different maturity. Chief executive Anirudh Devgan describes simultaneous acceleration on Design for AI and AI for Design. One side is customers using Cadence tools to design AI chips and AI systems; the other is Cadence using AI agents to accelerate design itself. The first already contributes real revenue, with core EDA up 20 percent and IP up 24 percent in the third quarter. The second has moved from portfolio to deployment, with more than 100 active customer engagements across digital, analog and custom, and system design workflows, plus the launch of Maestro, a portfolio of domain-specialised AI models purpose-built for chip and system design.
Second, the financial profile shows the strongest growth among the EDA and CAE peers, with a margin caveat. Third quarter 2026 revenue was 1.633 billion dollars, up about 22 percent. The GAAP operating margin was 26.2 percent against 31.8 percent a year earlier, the non-GAAP margin 44.1 percent against 47.6 percent, GAAP earnings per share 1.05 dollars against 1.05, and non-GAAP 2.11 against 1.93. Recurring revenue grew 18 percent and backlog reached a record 8.2 billion dollars, with 4.2 billion of remaining performance obligations to be recognised within twelve months. Full-year guidance was raised again to 6.387 to 6.417 billion dollars, a midpoint growth of 21 percent, with non-GAAP earnings per share of 8.37 to 8.43 dollars. Growth stayed strong while margins contracted, so the revenue story and the profitability story now point in different directions.
Third, System Design and Analysis is the fastest segment and a direct product of acquisition. Third quarter SDA revenue grew 30 percent, driven by printed circuit board and advanced packaging solutions, adoption of the Millennium platform and momentum across the physical SDA portfolio, plus continued integration of Hexagon D&E. How much of that 30 percent is inorganic is not disclosed, and this report treats it as the most important verification gap.
Fourth, AuraStack completes the agentic coverage of the flow, and the layer now has deployments behind it. Launched in the second quarter, AuraStack AI Super Agent is positioned as the industry's first agentic AI platform for advanced packaging and PCB. Together with ChipStack, ViraStack and InnoStack it spans the full electronic system design flow from chip through verification and IP to packaging and board. The industry-first label is a vendor claim without a third-party benchmark. The third-quarter disclosure gives the layer a scale reading that the previous edition lacked, with more than 100 active customer engagements across digital, analog and custom, and system design workflows, and the launch of Maestro, a portfolio of domain-specialised AI models built for chip and system design whose initial workflow results are stated as up to 30 percent higher quality and five times greater token efficiency than general-purpose foundation models, with several customer deployments underway.
Fifth, the NVIDIA relationship is technical complementarity without equity binding. Cadence appears as a partner in NVIDIA's July 2026 Agent Toolkit announcement, where AuraStack and the Millennium M2000 platform claim up to 20x multiphysics performance improvement in advanced packaging and PCB flows. Cadence received no NVIDIA equity investment, in contrast with the 2 billion dollars invested in Synopsys.
Evidence grading. Quarterly reports and press releases are primary. Transaction terms and earnings dilution guidance are primary. The 20x multiphysics figure, the 30 percent, 24 percent and 20 percent segment growth rates, and the Maestro 30 percent quality and five times token efficiency numbers are company statements. The industry-first label is marketing language.
Company Overview
Cadence Design Systems, Inc. is headquartered in San Jose, California, trades on Nasdaq as CDNS, and uses the calendar year as its fiscal year. Its narrative is Intelligent System Design: from chip to system, from electronics to physics, accelerating the full engineering design flow with computational software.
On the third quarter 2026 basis the business splits into four parts. Core EDA, covering digital, analog, verification and signoff, grew 20 percent year on year, with sixteen new digital and signoff full-flow wins and growing adoption among memory and hyperscaler customers. IP, the Star IP portfolio, grew 24 percent on demand across interface, memory and foundation IP. Hardware, meaning emulation and prototyping, delivered another record quarter on meaningful expansions with marquee customers and strong growth in the hardware cloud offering. System Design and Analysis, covering PCB, advanced packaging and multiphysics, grew 30 percent.
Financial quality in the quarter deserves separate attention, and the direction has changed. In the second quarter the GAAP operating margin rose from 19.0 to 28.4 percent, almost ten points, and the non-GAAP margin from 42.8 to 45.5 percent. In the third quarter both reversed: GAAP operating margin was 26.2 percent against 31.8 percent a year earlier and the non-GAAP margin 44.1 percent against 47.6 percent. The reconciliation from 26.2 to 44.1 percent attributes 8.9 points to share-based compensation, 4.9 to acquired intangible amortisation, 1.5 to acquisition and integration costs and 2.6 to restructuring. Acquisition accounting for Hexagon D&E and a revenue mix tilting toward software both contribute, and the expense line, particularly integration costs and the treatment of share-based compensation, matters more than the gross margin. The practical reading is that a company whose revenue is growing above twenty percent is now spending that growth on acquired-intangible amortisation, equity compensation and integration, which is a choice about accounting and about reinvestment rather than a deterioration in demand.
Backlog of 8.2 billion dollars is the highest in company history. That metric matters especially for EDA companies because contract cycles are long and visibility is high, so backlog largely sets the revenue floor for the next four to eight quarters. The 4.2 billion dollars of remaining performance obligations to be recognised within twelve months is about two-thirds of the raised full-year 2026 guidance midpoint of 6.402 billion dollars.
Acquisition Timeline: From BETA CAE to Hexagon D&E
Understanding Cadence in 2026 requires lining up three acquisitions.
The first is BETA CAE, completed in 2025, which brought structural analysis pre- and post-processing, the most labour-intensive part of a simulation workflow. The ANSA and META tools have deep user bases in automotive and aerospace, and what they add is pre-processing efficiency.
The second is Hexagon D&E, completed on 23 February 2026 for about 2.7 billion euros. It brought MSC Nastran, a de facto standard in structural finite elements; Adams, a de facto standard in multibody dynamics; AMendate for additive manufacturing simulation; Romax for drivetrains; CADLM for machine-learning surrogates; and CAEfatigue for fatigue. This transaction gave Cadence a complete chain across structures, multibody and fatigue for the first time.
The third is the integration of Hexagon D&E, running through 2026. Company guidance is about 160 million dollars of incremental 2026 revenue, roughly 0.28 dollars of non-GAAP earnings per share dilution, and accretion from 2027.
Read together, the path is clear. Take the workflow entry point through pre- and post-processing, add physics depth through solvers, then connect the flow with agents. Siemens buying Altair, which added optimisation and discrete elements, and Synopsys buying Ansys, which added multiphysics, are variants of the same play. The difference is the starting point: Cadence begins with chip design, so its notion of system means electronic and electromechanical systems rather than buildings or process industries.
Three Common Misreadings
Misreading one: treating total growth as EDA growth. Third quarter total revenue grew about 22 percent, core EDA grew 20 percent, IP grew 24 percent and SDA grew 30 percent. The four rates differ, and conflating them misjudges the EDA cycle. The gap also moved across the year, from 24.2 against 18 percent in the second quarter to 22 against 20 percent in the third, which means the spread between the fastest and slowest segments narrowed.
Misreading two: treating the 160 million dollars of Hexagon D&E revenue as its full contribution. That is the partial-year 2026 figure for a deal closed in February, covering about ten months, and it excludes BETA CAE. Assessing synergy requires the absolute SDA revenue and a two-year compound growth rate.
Misreading three: treating the AuraStack industry-first claim as a benchmarked conclusion. Synopsys released Multiphysics Fusion and an L4 multi-agent workflow in the same period, Siemens released Intelligence Center X, and Dassault released three virtual companions. Four companies claimed a first in 2026, which shows the label has no common definition.
Technology
Core EDA: First Beneficiary of AI Chip Demand
Core EDA grew 20 percent in the third quarter, driven in the company's words by broad-based strength across the portfolio, including sixteen new digital and signoff full-flow wins and growing adoption among memory and hyperscaler customers. Those three customer groups buy for different reasons.
Hyperscalers design their own AI accelerators and need signoff convergence speed and multi-die co-design, so they buy digital implementation, timing signoff and 3D-IC capability. Top semiconductor companies need methodological completeness at advanced nodes and buy through long-term agreements with large seat counts. Startups need fast onboarding and elastic cloud capacity, buying subscriptions plus on-demand compute. That third group contributes little revenue today but is the future customer pool.
Hardware delivered another record quarter on meaningful expansions with marquee customers and strong growth in the hardware cloud offering, which is the most direct read on AI chip expansion, because every large chip runs enormous verification volume on emulation before tape-out.
IP: What Sits Behind 24 Percent Growth
IP revenue grew 24 percent in the third quarter, on demand across interface, memory and foundation IP and continued engagement with leading hyperscalers, semiconductor companies and foundries. The growth rate is lower than the more than 40 percent recorded in the second quarter, when a significant agreement with Intel was named explicitly. That earlier agreement still matters beyond its quarter, because Intel has historically built in-house, and large-scale purchasing from an external IP vendor signals a change in product cadence and resource allocation.
The Star IP portfolio covers interfaces including PCIe, DDR, UCIe and Ethernet, plus memory controllers and security IP. In the chiplet era IP becomes more valuable because every chiplet needs silicon-proven interfaces. This is a structural advantage Cadence holds over pure EDA vendors.
System Design and Analysis: End-to-End Assembled Through Acquisition
The SDA segment now contains three layers.
The first is electronics and electromagnetics: Sigrity for power integrity, Celsius for thermal, Clarity for electromagnetic, and AWR for microwave. This layer was already a Cadence strength serving PCB and packaging.
The second is structures and multibody: MSC Nastran for linear and nonlinear structures, Adams for multibody dynamics, CAEfatigue for fatigue, Romax for drivetrains, and BETA CAE ANSA and META for pre- and post-processing. This layer came from acquisitions.
The third is surrogates and optimisation: CADLM machine learning plus the in-house Optimality optimiser. This layer is the technical interface to physical AI, using solver-generated data to train surrogates and embedding those surrogates back into design.
With all three layers, Cadence can claim end-to-end multiphysics. Two cautions apply. First, Nastran and Adams user bases overlap only partially with the existing electronics customer base, so cross-selling takes time. Second, solver validation depth is built over decades, and an acquisition brings assets rather than automatically acquired trust.
AI Super Agent: The Layering of Agents
The AI product line is layered by design stage. ChipStack addresses digital design and verification. ViraStack addresses verification including formal and simulation. InnoStack addresses analog and custom design. AuraStack, added in the second quarter, addresses advanced packaging and PCB.
All four share a design logic: agents orchestrate multiple tools with human confirmation at key checkpoints. Technical support combines the in-house Optimality engine with external models. In NVIDIA's July 2026 announcement, AuraStack and the Millennium M2000 platform appear as partners using Nemotron and CUDA-X, claiming up to 20x improvement in advanced packaging and PCB multiphysics flows.
That 20x needs careful handling. It names no baseline, whether manual flow or the previous generation of tools, no task type, whether place, route or simulate, and no accuracy comparison. It is a vendor claim in the classic form.
Maestro is the second layer of the agentic story and the first one with a measured claim attached. Rather than orchestrating tools, Maestro is a portfolio of domain-specialised AI models purpose-built for chip and system design. The company reports initial workflow results of up to 30 percent higher quality and five times greater token efficiency against general-purpose foundation models, with several customer deployments underway. Three cautions apply. The comparison object is a class of models rather than a named baseline, the metric is described as initial workflow results rather than a benchmark, and the deployment count is given without names or volumes. The commercial relevance is nonetheless real: token efficiency is a cost line for every design run, and a five-times reduction would change the economics of large-scale agentic design if it holds in production.
Digital Twins: Where Reality Sits
Cadence describes itself as a market leader in AI and digital twins, and its Reality digital twin platform addresses thermal and airflow simulation for data centres and facilities. The link to physical AI is that AI factory design needs accurate thermal and power simulation, which Celsius and Reality cover.
That is also one of the interfaces to the NVIDIA DSX ecosystem, which needs physically accurate AI factory digital twins while Cadence supplies the thermal and electrical simulation layer.
Layering Credibility
Verifiable: financial data; transaction consideration and structure; the Hexagon D&E asset list; full-year guidance; and backlog.
Partially verifiable: the organic versus inorganic split inside 30 percent SDA growth, which is undisclosed; the AuraStack 20x improvement, which has an NVIDIA endorsement without independent testing; and the customer composition behind 20 percent core EDA growth.
Not verifiable: the definition and benchmark behind industry-first agentic platform for advanced packaging and PCB; actual customer counts and deployment scale for AI Super Agents; and cross-selling results between Nastran and Adams and Cadence electronics customers.
Product and Commercial Model
The model combines software subscription, hardware sales and IP licensing. This combination is standard in EDA, but the three parts differ widely in margin and cyclicality.
Software subscription. Core EDA and SDA are predominantly multi-year subscriptions with high predictability. Backlog of 8.2 billion dollars, recurring revenue up 18 percent and 4.2 billion of twelve-month remaining performance obligations give more than a year of visibility.
Hardware. Emulators such as Palladium and Protium carry high unit prices and clear cyclicality, but once a customer selects a platform the marginal cost of expansion is low. A record quarter with twelve new customers shows AI chip companies are expanding capacity.
IP. Licensed per project, with revenue recognition tied to tape-out, so it is volatile but high margin. The Intel agreement is a landmark.
The combination effect. A single AI chip customer may buy EDA tools, verification hardware, interface IP and packaging simulation at once, which is the advantage Cadence holds over pure IP or pure EDA vendors and the reason it can reach a 45 percent non-GAAP operating margin.
Pricing opacity. Like every EDA company, Cadence does not publish unit prices, so outsiders infer from revenue structure, backlog and remaining performance obligations. Whether AI features are separately charged is also unclear. Public information suggests AI Super Agents are bundled into existing subscriptions rather than sold as separate SKUs.
Competitive Landscape
Against Synopsys including Ansys, the contest is direct and extends from EDA into multiphysics. Synopsys embeds Ansys solvers into signoff through Multiphysics Fusion, while Cadence adds structures and multibody through BETA CAE and Hexagon D&E. The difference is the starting point. Synopsys gains physics depth from Ansys, which is strong across electromagnetics, fluids and structures, while Cadence gains it from MSC, which is strong in structures and multibody but weaker in fluids. In computational fluid dynamics Cadence relies on in-house and partner solvers and trails Ansys Fluent.
Against Siemens EDA with Calibre and Simcenter, competition runs on both physical verification and multiphysics. Siemens has the advantage of automation hardware and factory data, while Cadence has EDA share and proximity to AI chip customers.
Against Dassault Systemes, overlap begins at system-level simulation but customer structures differ. SIMULIA is stronger in aerospace and automotive structures and multiphysics, and with Nastran and Adams, Cadence competes head-on in structures and multibody for the first time. Dassault binds to the product lifecycle through 3DEXPERIENCE while Cadence binds to the design flow.
Against the former Ansys cohort and COMSOL, Cadence presence in general CAE rises after the acquisition. Its sales organisation remains oriented to semiconductor customers, so selling Nastran and Adams into traditional manufacturing is an execution question.
Against NVIDIA, the relationship is technical complementarity without equity. Cadence was an early partner in the Agent Toolkit, and AuraStack uses Nemotron and CUDA-X. But NVIDIA works simultaneously with Siemens, Synopsys and PTC, so Cadence cannot obtain exclusivity.
Against Chinese EDA vendors, the gap is generational today and policy-driven tomorrow. Tighter export controls accelerate local substitution, and Cadence lists export controls among its risk factors.
Financing and Capital Structure
Cadence does not use venture financing, and its capital activity centres on acquisitions, buybacks and debt management.
The acquisition bill. Hexagon D&E cost about 2.7 billion euros, 70 percent cash and 30 percent stock. It is among the largest transactions in Cadence history and the first with a large stock component, which preserves cash while turning the seller into a shareholder and aligning interests.
Financial impact. About 160 million dollars of incremental 2026 revenue, roughly 0.28 dollars of non-GAAP earnings per share dilution, turning accretive in 2027. The transaction includes about 150 million euros of transaction-related taxes owed by the acquired entities.
Buybacks and debt. Cadence has run repurchases for years and uses free cash flow to service acquisition debt. The third quarter used 400 million dollars of repurchases and the board authorised an additional 2.5 billion dollars. Operating cash flow for the first nine months of 2026 was 1.414 billion dollars against 1.175 billion a year earlier. Full-year operating cash flow guidance of about 2 billion dollars funds both buybacks and debt service.
Key financial sequence, GAAP unless noted. Third quarter 2026 revenue 1.633 billion dollars against 1.339 billion; GAAP operating margin 26.2 percent against 31.8; non-GAAP 44.1 against 47.6; GAAP earnings per share 1.05 dollars against 1.05; non-GAAP 2.11 against 1.93; recurring revenue up 18 percent; backlog 8.2 billion. Full-year guidance: revenue 6.387 to 6.417 billion at a 21 percent midpoint growth, GAAP operating margin 28.2 to 29.2 percent, non-GAAP 44.2 to 45.2 percent, GAAP earnings per share 4.84 to 4.90 dollars, non-GAAP 8.37 to 8.43, and operating cash flow of about 2 billion. Fourth quarter guidance: GAAP operating margin 30 to 31 percent, non-GAAP 44 to 45 percent, GAAP earnings per share 1.23 to 1.29 dollars, non-GAAP 2.19 to 2.25; no fourth-quarter revenue range was given.
Valuation note. The 44.1 percent non-GAAP operating margin excludes share-based compensation, acquired intangible amortisation, acquisition and integration costs and restructuring, which together are 17.9 points of the gap in the third quarter. The GAAP figure of 26.2 percent is more conservative, provided the reader remembers that GAAP earnings include non-cash amortisation. The margin is now moving against revenue growth, which is the second consecutive quarter in which the acquisition-related cost base grows faster than the non-GAAP presentation makes visible.
DISCREPANCIES
1. The Hexagon D&E closing date appears in three versions. (Source grade: OFFICIAL versus MEDIA) The company release dated the closing 23 February 2026. One trade outlet reported 24 February, and another page carries 26 March, which is the republication date. This report uses 23 February and explains that the other two are a time-zone difference and a republication date.
2. The transaction currency is inconsistently stated. (Source grade: OFFICIAL versus MEDIA) The company states about 2.7 billion euros including roughly 150 million euros of transaction-related taxes. Some English headlines write 2.7 billion without naming the currency, which invites a misreading as dollars. This report uses euros and notes the inclusion of taxes.
3. The organic and inorganic split inside 30 percent SDA growth is undisclosed. (Source grade: OFFICIAL) The company gives segment growth and a qualitative note about continued Hexagon D&E integration without an absolute inorganic contribution. The only related number is guidance of about 160 million dollars of 2026 incremental revenue. This report does not estimate the split and records the gap.
4. Total growth and core EDA growth are conflated. (Source grade: OFFICIAL versus MEDIA) Third quarter total revenue grew about 22 percent, core EDA 20 percent, IP 24 percent and SDA 30 percent, while the second quarter figures were 24.2 percent, 18 percent, more than 40 percent and 37 percent. Some coverage describes EDA demand using the total rate. This report lists all four.
5. The industry-first agentic platform claim has no comparable benchmark. (Source grade: OFFICIAL) Similar claims appeared densely in 2026, including the Synopsys L4 multi-agent workflow, the Siemens Intelligence Center X and the Dassault virtual companions. No vendor published a comparable grading standard. This report treats it as a vendor statement.
6. The 20x multiphysics improvement lacks a baseline. (Source grade: OFFICIAL) The figure appears in NVIDIA's July 2026 Agent Toolkit announcement describing AuraStack and the Millennium M2000 platform, without naming the comparison object, task type or accuracy condition. This report keeps the figure and labels it a vendor claim.
7. The BETA CAE consideration does not appear in the primary materials reviewed. (Source grade: OFFICIAL) The company confirms BETA CAE was absorbed in 2025 and cites its structural pre- and post-processing technology in the Hexagon D&E announcement, but no consideration figure appears in the materials reviewed here. This report treats the amount as unconfirmed in primary sources.
8. The Maestro quality and token-efficiency figures have no stated baseline. (Source grade: OFFICIAL) The third-quarter release reports up to 30 percent higher quality and five times greater token efficiency than general-purpose foundation models, with several customer deployments underway, without naming the workflows, the task set, the compared model class or the measurement procedure. Ruling: this report keeps the figures and labels them initial workflow results rather than a benchmark.
9. The margin direction reversed between the second and third quarters. (Source grade: OFFICIAL) In the second quarter the GAAP operating margin rose from 19.0 to 28.4 percent year on year and the non-GAAP margin from 42.8 to 45.5 percent; in the third quarter the GAAP operating margin fell from 31.8 to 26.2 percent and the non-GAAP margin from 47.6 to 44.1 percent. Ruling: the margin-expansion narrative does not carry into the third quarter. The company attributes the gap to share-based compensation, acquired intangible amortisation, acquisition and integration costs and restructuring, so this report reads it as an accounting and reinvestment effect rather than a demand problem, and lowers the commercialisation score accordingly.
Risks and Diligence Gaps
Integration risk comes first. Two acquisitions closed within a year, each with a different sales organisation, licensing system and customer base. Nastran and Adams customers sit in simulation departments of automotive, aerospace and heavy industry companies, whose procurement processes differ substantially from Cadence semiconductor customers. Whether cross-selling works will be visible in 2027 organic SDA growth.
Dilution and amortisation create pressure. Fiscal 2026 non-GAAP earnings per share are diluted by 0.28 dollars, with accretion only from 2027. Intangible amortisation depresses GAAP earnings over the same period, and slower integration would extend the dilution window.
Customer concentration and the AI chip cycle. Revenue depends heavily on AI chip and hyperscale capital expenditure. If AI infrastructure spending slows, hardware and IP feel it first.
Export controls. The company lists export controls, tariffs and trade restrictions among risk factors. EDA is a sensitive category, and policy shifts rewrite guidance directly.
The credibility cost of industry-first narratives. Four companies claimed a first in the same year, which dilutes the label. The practical consequence for investors is that such statements cannot be used to judge technical leadership.
Diligence gap list, ordered by importance. First, organic SDA growth excluding Hexagon D&E and BETA CAE. Second, customer retention and cross-sell revenue after Hexagon D&E consolidation. Third, paid customer counts, unit prices and ARR contribution for AuraStack. Fourth, actual usage of the AI Super Agent family, meaning how many customers run it in daily flow. Fifth, hardware customer concentration including top-five share. Sixth, the size and term of the Intel IP agreement. Seventh, the commercial terms of the NVIDIA collaboration. Eighth, China revenue share and sensitivity under export control scenarios.
Composite Assessment
techScore 8.5. Cadence is the only company in 2026 holding signoff-grade EDA, structural and multibody solvers, pre- and post-processing and surrogate modelling together, and AuraStack brings agents into packaging and PCB, previously a gap. The deduction reflects weaker computational fluid dynamics than Ansys and the time still required to build validation depth for end-to-end multiphysics.
bizScore 8.5. Third quarter revenue of 1.633 billion dollars, up about 22 percent, recurring revenue up 18 percent, a record backlog of 8.2 billion and full-year guidance raised again to 21 percent growth keep Cadence the fastest grower in the EDA and CAE peer set. The deduction is the margin reversal: the non-GAAP operating margin fell from 47.6 to 44.1 percent and the GAAP margin from 31.8 to 26.2 percent year on year, so revenue growth of above twenty percent was accompanied by lower profitability.
fundScore 8.5. A 2.7 billion euro acquisition funded 70 percent in cash and 30 percent in stock, nine-month operating cash flow of 1.414 billion dollars against 1.175 billion a year earlier, 400 million dollars of repurchases in the third quarter and a new 2.5 billion dollar authorisation. The deduction reflects acquisition debt, the gap between GAAP and non-GAAP earnings and 2026 earnings per share dilution.
productScore 8.5. From core EDA, IP and hardware to SDA and four AI stacks, the matrix is complete and interlocking, and acquisitions filled structures and multibody. The deduction reflects an opaque strategy for overlapping products across MSC and parts of BETA CAE.
marketScore 8.5. A default option in the AI chip toolchain, with customers across hyperscale, leading semiconductor and startup tiers, and a physical AI narrative that opens a route into traditional manufacturing. The deduction reflects concentration in the AI capital expenditure cycle.
References
- Cadence second quarter 2026 results, published 27 July 2026: revenue 1.584 billion dollars against 1.275 billion; GAAP operating margin 28.4 percent against 19.0; non-GAAP 45.5 against 42.8; GAAP earnings per share 1.33 dollars against 0.59; non-GAAP 2.11 against 1.65
- Same filing: record backlog of 8.1 billion dollars and 4.2 billion dollars of remaining performance obligations to be recognised within twelve months
- Same filing business highlights: core EDA up 18 percent, IP up more than 40 percent including a significant Intel agreement, hardware at a record with twelve new customers, and System Design and Analysis up 37 percent
- Same filing full-year guidance: revenue 6.26 to 6.34 billion dollars at a 19 percent midpoint growth, GAAP operating margin 27.75 to 28.75 percent, non-GAAP 43.75 to 44.75 percent, GAAP earnings per share 4.76 to 4.86 dollars, non-GAAP 8.05 to 8.15, and operating cash flow of about 2 billion
- Cadence press release, Cadence Completes Acquisition of Hexagon Design and Engineering Business, Advancing Leadership in Physical AI and Multiphysics, 23 February 2026: consideration about 2.7 billion euros including roughly 150 million euros of transaction-related taxes, 70 percent cash and 30 percent stock, about 160 million dollars of 2026 incremental revenue, about 0.28 dollars of non-GAAP dilution, accretive from 2027
- Asset list from the same release: MSC Nastran and Adams, alongside AMendate from 2019, Romax from 2020, CADLM from 2020 and CAEfatigue from 2020, integrated with Cadence electronics and electromagnetic capability, computational fluid dynamics and BETA CAE structural pre- and post-processing
- Hexagon background from the same release: a global measurement technology company with about 24,800 employees in 50 countries and net sales of approximately 5.4 billion euros
- Cadence second quarter 2026 business highlights: launch of AuraStack AI Super Agent for advanced packaging and PCB, together with ChipStack, ViraStack and InnoStack covering the full electronic system design flow
- NVIDIA Agent Toolkit announcement, 26 July 2026: Cadence using Nemotron, accelerated computing and CUDA-X with AuraStack AI Super Agent and the Millennium M2000 platform, claiming up to 20x improvement in advanced packaging and PCB multiphysics flows
- Cadence company description: positioning in AI and digital twins, the Intelligent System Design strategy, customers across hyperscale computing, mobile communications, automotive, aerospace, industrial, life sciences and robotics, and the Reality digital twin platform with Celsius thermal analysis
- Cadence management statements: Anirudh Devgan, president and chief executive, on simultaneous acceleration of Design for AI and AI for Design and on being the only provider with agentic solutions across the full electronic system design flow; John Wall, senior vice president and chief financial officer, on the raised outlook
- Trade coverage recording different Hexagon D&E closing dates of 23 February, 24 February and 26 March 2026, addressed in DISCREPANCY 1
- Cadence third quarter 2026 results, published 26 October 2026: revenue 1.633 billion dollars against 1.339 billion; GAAP operating margin 26.2 percent against 31.8; non-GAAP 44.1 against 47.6; GAAP diluted earnings per share 1.05 dollars against 1.05; non-GAAP diluted 2.11 against 1.93; recurring revenue up 18 percent
- Same release business highlights: core EDA up 20 percent with sixteen new digital and signoff full-flow wins, IP up 24 percent, hardware at another record quarter, System Design and Analysis up 30 percent, launch of Maestro with up to 30 percent higher quality and five times token efficiency versus general-purpose foundation models and several customer deployments underway, and more than 100 active AI Super Agent customer engagements
- Same release business outlook and fourth-quarter guidance: fiscal 2026 revenue 6.387 to 6.417 billion dollars at a 21 percent midpoint growth, GAAP operating margin 28.2 to 29.2 percent, non-GAAP 44.2 to 45.2 percent, GAAP diluted earnings per share 4.84 to 4.90 dollars, non-GAAP 8.37 to 8.43; fourth quarter GAAP operating margin 30 to 31 percent, non-GAAP 44 to 45 percent, GAAP diluted earnings per share 1.23 to 1.29 dollars, non-GAAP 2.19 to 2.25, with no fourth-quarter revenue range disclosed
- Same release capital return and cash flow: 400 million dollars used for share repurchases in the third quarter, an additional 2.5 billion dollar authorisation from the board, and operating cash flow of 1.414 billion dollars for the first nine months of 2026 against 1.175 billion a year earlier
This report is compiled from public sources for informational purposes only and does not constitute investment advice. Data is current as of 2 November 2026; please refer to company announcements and periodic reports.
Related papers
Related tools
Related cases
- DrivAerML High-Fidelity Road-Car Aerodynamics Dataset and Surrogates
- AirfRANS High-Fidelity Airfoil Surrogate Benchmark Case
- Physics-Informed CNN for Battery Pack Temperature Fields
- IASC-ASCE Structural Health Monitoring Benchmark and Physics-Informed Damage Identification
- OpenFOAM Industrial Multiphysics Pipeline: From Meshing to Data Supply
Related reports66
- 51World Deep Dive: A Listed Digital Twin Company Where 51Sim Becomes the First Growth Engine While 51Earth Nearly Vanishes, and Two Irreconcilable Financing Naming Conventions Must Both Be Kept
- AMI Labs: A USD 1.03B Seed Bet on World Models, LeCun's Non-Generative Path and Its Blank Scorecard
- Anshi Pacific (PERA) Deep Dive: A Beijing Domestic-CAE Leader Founded in 2003 Whose PERA SIM General Simulation, AIROM Reduced-Order Models, AgentX Multi-Agent and Jingzhi iGPT/Whale Span from Industrial Simulation to Physics AI
- AspenTech Deep Dive: The Process-Simulation Incumbent Acquired by Emerson and Its Hybrid Industrial-AI Route
- Autodesk Deep Dive: Neural CAD Foundation Models and a $200M World Labs Bet on Physical-World AI, with NAVASTO Adding CFD Surrogate Capability
- Bentley Systems Deep Dive: iTwin Digital Twin and Generative AI Empower Full-Lifecycle Infrastructure Engineering
- BeyondMath Deep Dive: A Foundation Physics AI on the Neural-Operator Route, a $19M Consortium Budget, and Three Misread Facts
- COMSOL Deep Dive: Four Decades of Multiphysics Solvers, Entering Physical AI via DNN Surrogate-Model Training and All-Physics GPU Acceleration
- Dassault Systèmes Deep Dive: 3DEXPERIENCE and SIMULIA Virtual Twin Physics Behavior Anchor Industrial AI, with LEO Virtual Companion and GPU Explicit Dynamics as Physical-AI Levers
- DP Technology: From the Gordon Bell Prize to National-Level Unicorn Status and the Door of a Hong Kong Listing, the AI-for-Science Leader's Technical Delivery and Commercial Blank
- Emmi AI: From Linz Spin-out to a EUR 330M Mistral Acquisition in 17 Months
- Fainite Deep Dive: A Zurich Physics-AI Simulation Startup Whose Caltech, ETH Zurich, Cambridge and Google Team Uses Physics-Aware Deep Learning to Cut Simulation from Tens of Hours to Seconds, With Only Venture Kick Grants and No Priced Round
- Fysics AI: The PhysX Founder's Second Act, a Domestic Differentiable Physics Engine and Its Ecosystem Play
- Frontop (301313) Deep Dive: From Digital Exhibition Halls to a Physical-AI Data Foundation - FTE Engine, FunSim Simulation and the Gap Between Positioning and Revenue
- Godela Deep Dive: A YC X25 Physics-AI Simulation Startup Whose Ex-Apple, Google and Intel Team Uses Physics-Constrained Surrogate Models to Cut Simulation from Days to Seconds, at Seed Stage with Pilot-Only Access
- HollyWave (688682) Deep Dive: RDSim AI Electromagnetic Simulation and Engineering Agents Close the Simulation-Measurement-Product Loop
- IAT (Alt) Deep Dive: China's First Listed Independent Automotive Design Company Pivots to AI-native R&D with the Yuanfang, Yufeng and Taiyi 3.0 Product System and an NVIDIA Omniverse Partnership, Reporting RMB 975M Revenue and a Widened RMB 250M Net Loss in 2025
- Inductiva Deep Dive: A Closed Case in Open-Solver Orchestration, Eight Premises That Must Be Corrected, and Where Co-Founder and CTO Hugo Penedones Went Next
- Intesim (INTESIM) Deep Dive: A Home-grown Chinese CAE Leader Founded in 2009 with Self-developed Solvers for Structure, Fluid, Electromagnetics, Thermal and Acoustics and Multiphysics Coupling, a ~RMB 100M B Round from CNII and Shenzhen Capital, and 2025 Revenue and Bookings Both above RMB 100M
- iSoftStone (301236) Deep Dive: A ChiNext-Listed IT Services Giant Whose TianShu Platform Builds a Three-Layer Physical-AI Stack with Self-Developed TwinHub Digital Twin, FSim Industrial Simulation, CAE Multiphysics and Phyxis Embodied Control
- Jua Deep Dive: The Earth Physics Transformer as a Large Physics Model for Energy Trading
- Keysight (with ESI Group) Deep Dive: Virtual Prototyping and Hybrid AI Bridge Measurement and Simulation, with Natural-Language Assistants in EDA
- Lightwheel Deep Dive: A Physical AI Data and Evaluation Infrastructure Startup Founded by a Former NVIDIA Simulation Lead, Seven Rounds Totalling Roughly CNY 2.65 Billion at a Valuation Above USD 2 Billion, Seats on Both the Newton and EgoVerse International Standards Committees, With Two Incompatible Accounts of Q1 Orders
- Luminary Cloud: The Physics AI Model Factory Backed by NVentures
- MathWorks Deep Dive: MATLAB/Simulink and Simscape Multiphysics Power Physical AI, with R2026a Copilots and an Agentic Toolkit for Edge Intelligence
- Maniformer (Mifeng) Deep Dive: An Agibot-Spun Physical-AI Data Infrastructure Firm Whose Bodyless MEgo Capture Hardware Is in Mass Production and Which Closed Three Rounds in Half a Year Led by China Telecom
- Mistral AI: How Physics AI Moved From Language Models to EUV Lithography and Nuclear Power
- Monolith AI: The British Engineering-AI Firm That Cut Physical Test Counts Instead of Accelerating Simulation, Now Inside CoreWeave
- Monumo Deep Dive: A Cambridge Deeptech Company Redesigning Electric Motors with the Anser AI Engine, Tens of Millions of Multiphysics Simulations per Day, GBP 10.5M Raised in Total, and a Series A Being Raised
- NAVASTO: Moving Solver Insight Onto the Designer's Desk, and How Autodesk Bought That Entry Point
- Navier: Handing CFD Workflows to Agents, a Seed-Stage Bet on AI for Hardware Engineering
- Neural Concept: An EPFL Spin-out Betting on Neural Surrogates, with $138M Raised
- NP Co. (NP Company) Deep Dive: A 2025-founded Paris Inria-born AI-native simulation startup that raised a EUR 6M pre-seed led by Partech with Mistral AI co-founders as angels, claiming a pre-trained physics foundation model at 1,000x speedup — most claims unverified
- NVIDIA Deep Dive: Building the Full-Stack Physical-AI Infrastructure with PhysicsNeMo, Warp, Cosmos World Foundation Models and Omniverse
- Physical Intelligence Deep Dive: A $2.1B Bet on the GPT Moment for Robotics, With pi0.7 Self-Reporting 60-80 Percent Success on Unseen Tasks and Neither Pricing nor Revenue Disclosed for the Pi API
- Physical Superintelligence Deep Dive: A $58M Seed Bet on Virtual Physicists, With Emmy Targeting AI Data Center Multiphysics Optimization as Its First Commercial Proof Point
- PhysicsX: Pretrained Geometry and Physics for Industrial Engineering at a $2.4B Mark
- Prometheus: A $41B Mark, $18.2B Raised, and Not Yet a Product
- PTC Deep Dive: Creo/Onshape and the ThingWorx Digital Thread Power Physical AI, with NVIDIA Isaac Sim Bridging Robotics Design to Simulation
- Qianjue Technology Deep Dive: How a Body-Agnostic Embodied Brain Pairing the Predictive World Model Polibrain with Brain-Inspired Partitioning and Polynomial Representation Reached a Hundred Thousand Cleaning and Service Robots
- Quanscient Deep Dive: Cloud-Native and Quantum-Powered Multiphysics Simulation as the Data Engine for AI-Native Hardware Engineering
- Rescale Deep Dive: An HPC Operating System and Desktop Engineering Workbench in the Cloud, a $115M Series D, and Five Misread Facts
- Schrödinger Deep Dive: Physics-Based FEP+ Simulation and Machine Learning Power Molecular Discovery, with the 2026 Bunsen Agentic Co-Scientist
- Secondmind Deep Dive: A Cambridge Engineering-AI Firm Whose Bayesian Active Learning Cuts CAE Simulation and Test-Bench Runs by up to 90%, with Roughly USD 73M Raised and Mazda Both Customer and Investor
- Shichuang Technology Deep Dive: SupreCAST Cloud-Native Material-Processing Simulation and Industrial Agents Advance the CAE+AI Strategy
- ShonCloud (Shunyun) Deep Dive: Meshless Particle-Method CFD for EV and Nuclear Fluid Multiphysics
- Shuzhi Maijin Deep Dive: A Shenzhen Fully Self-Developed SNN-plus-LBM CFD Startup Claiming Second-Level Results on Hundreds-of-Millions of Grids on Consumer Hardware, With a Tens-of-Millions-RMB Angel Round From a Ministry-of-Finance-Affiliated Fund and an FPGA Migration Plan
- Siemens (incl. Altair) Deep Dive: The 2025 Altair Acquisition Unites Simcenter and Altair, with PhysicsAI Geometric Deep Learning and PhysicsAI Generate as the Physical-AI Core
- Silurian Deep Dive: A Generative Forecasting Transformer Earth-System Foundation Model for Weather Risk on Energy and Infrastructure Assets
- SimScale Deep Dive: A Cloud CAE Platform Built on Core-Hour Subscriptions, EUR 52M of Equity Kept Separate from EUR 25M of Debt, and Five Misread Facts
- Skild AI Deep Dive: A General Robot Foundation Model Incubated at CMU's Robotics Institute, With a USD 1.4B Series C at a USD 14B Valuation, a Conflicting Series B Size, and an ARR Figure That Is a Run-Rate Rather Than Revenue
- Songying Deep Dive: A Huawei-Cloud-Alumni Startup Building a Domestic Physical AI Operating System, With ORCA OS Positioning Against NVIDIA Omniverse, Three Incompatible Financing Narratives, and a Verifiable Third-Party Anchor in a CNY 20 Million Minority Investment
- Sudo Tech Deep Dive: Zero-Real-Data Simulation Training, the Fully In-House Sudo R1 Stack, and an Atomic-Skill Commercialisation Strategy from a Startup Valued Above CNY 20 Billion in 17 Months
- Suochen Technology: The Pain of Becoming a Physics-AI Platform from a CAE Vendor
- Synopsys (Ansys) Deep Dive: The 2025 Ansys Acquisition Turns the EDA Leader into a Silicon-to-Systems Engineering Powerhouse, with Ansys SimAI and the NVIDIA Physical-AI Partnership as the Core Physics-AI Levers
- TenFong Deep Dive: An Academician-Led Chinese Industrial Simulation Software Vendor Whose Official Timeline and Database Records Use Incompatible Round Names, With an Official Financing PDF and an Investment Database Both on File
- Tianfu (Nanjing Tianfu Software) Deep Dive: A 2011-founded Chinese Intelligent-CAE Veteran That Closed a Several-Hundred-Million-RMB Series C, Pivoted to an Industrial World Model in 2024, With Unreconcilable Financing and Valuation Records and No Disclosed Revenue
- Tongyuan Soft Control Deep Dive: MWORKS (Syslab/Sysplorer) as a Domestic MATLAB/Simulink Alternative, with MWORKS AI and the MOGICK Engineering Agent Advancing Home-Grown Physical AI
- Vinci AI: Turning Physics Simulation from a Checkpoint into a Continuous Constraint
- Westlake Robotics Deep Dive: Westlake University's First Robotics Spin-out Builds a Dual-Pretrain Brain-and-Cerebellum Architecture with WR1 Whole-Body Foundation Model and GAE Behaviour Expert
- World Labs: Fei-Fei Li's Spatial-Intelligence Unicorn, Acquired by AMD for $8.2B
- XVERSE Deep Dive: A Simulation-World Route to Physical AI Built on the XVERSE Model Family, an Edge-Cloud 3D Engine and the Physics-Based MotionGen Model
- SimWE: From In-House Solvers to a Physical-AI Engine, the Domestic CAE Leader and What Comes Next
- Yunjing Zhifang Deep Dive: A Peking-University-Incubated Cloud-Native GPU CAE Company with the Simversus Platform and the YJ-DORGP Small-Sample Intelligent Solver, Reporting 275 Percent Order Growth, 160 Percent Revenue Growth and Breakeven in 2025
- Zhongke Diwiji (Five Ages) Deep Dive: A CAS-Institute-of-Automation and Tsinghua Spin-out Whose FAM Embodied Foundation Model Uses Action Optical Flow for Cross-Embodiment Generalisation, Reaching 97% Success from 3 to 5 Demos, with EraBrain and Jiling Robots Serving Sinopec, Leapmotor and Midea
- ZWSOFT (688083) Deep Dive: Overdrive Kernel and ZW3D Wukong 2027 Underpin an Integrated CAD/CAE/CAM Stack, with a CAx Agent Platform as the Practical Route to Engineering AI
This report is compiled from public sources. It is for reference only and does not constitute investment advice. Data is current as of the publication date; please refer to the company’s official filings.