Cadence Design Systems (CDNS) Q3 2026 Earnings Call: Q&A Summary

Q: Jason Celino from KeyBanc asked for an explanation of the Maestro portfolio, whether Cadence is developing domain-specific models, and how it will be monetized.

A: Anirudh Devgan explained that Maestro is a portfolio of domain-specialized models that uses model routing to select the most efficient model for a specific task, improving efficiency by up to 5x. It allows for post-training on proprietary data and integrates with Cadence tools to provide customization and performance that general-purpose models cannot achieve.

Q: Charles Shi from Needham asked about the strategy for collaborating with Frontier labs versus using Cadence's own models and how customers should choose between Cadence and competitors.

A: Anirudh Devgan stated that Cadence collaborates with leading frontier labs while maintaining control over the primary customer interface and data confidentiality. Maestro offers flexibility, allowing customers to use frontier models when appropriate while also enabling on-prem or cloud deployment and further post-training to ensure the best results for specific design tasks.

Q: Jim Schneider from Goldman Sachs asked for an update on Agentic AI monetization and the dynamics behind recurring revenue growth.

A: Anirudh Devgan and John Wall noted that recurring revenue grew 18% year-over-year, driven by strong demand for Agentic AI solutions which also increase usage of base tools. The company expects strong performance in 2026.

Q: Kelsey Chia from Citi asked if the strong EDA growth is primarily driven by hardware and if Agentic AI revenue contribution is still small.

A: Anirudh Devgan and John Wall clarified that recurring revenue growth of 18% is distinct from hardware, which is non-recurring. Agentic AI acts as a demand accelerator for core EDA tools, as customers use these agents to perform more design exploration, leading to higher consumption of core software.

Q: Vivek Arya from Bank of America Securities asked how Cadence views its growth and royalty components compared to competitors and how it will maintain market share and margins.

A: Anirudh Devgan stated that Cadence expects 21% revenue growth and margins slightly under 45% for 2026. The company is gaining share across IP, hardware, and EDA, and while it uses royalty models in IP, it focuses on delivering overall value to customers.

Q: Joe Quatrochi from Wells Fargo asked for an explanation of the change in the 2026 operating profit outlook.

A: John Wall explained that the 2026 non-GAAP operating margin is expected to be 44.7%. While the Hexagon DD integration impacts incremental margins this year, the company expects it to contribute positively next year, consistent with historical performance.

Q: Siti Panigrahi from Mizuho asked if hardware growth is sustainable into 2027 and if there was a one-time spike in G&A expenses.

A: John Wall and Anirudh Devgan stated that hardware had a record quarter driven by AI and HPC demand, and the company expects another record year. The G&A increase was due to the timing of annual charitable contributions. Anirudh added that hardware demand is tied to design size, which is expected to grow significantly due to 3D IC and technology scaling.

Q: Harlan Sur from JP Morgan asked if positive demand trends are continuing in analog, power, and microcontroller segments and if this is translating into a better book of business.

A: Anirudh Devgan said momentum in analog and mixed-signal is strong, aided by the recovery of the market and the adoption of ViraStack. The company is seeing increased demand for analog productivity tools from both mixed-signal and AI companies.

Q: Lee Simpson from Morgan Stanley asked about the space technology company engagement and if there is anything specific regarding IP services.

A: Anirudh Devgan noted that system companies, including those in space and finance, are increasingly doing their own chip designs, which creates opportunities for IP, hardware, and design services. Cadence is encouraged by this trend of non-traditional companies entering the market.

Q: Joe Vruwink from Baird asked if specialized models are helping address analog and back-end domains and if there is a higher revenue implication.

A: Anirudh Devgan stated that demand for ViraStack, InnoStack, and AuraStack is phenomenal, with multiple customers reporting significant productivity improvements. These agents are tightly integrated with base tools, and as they become more complex and differentiated, they create a strong competitive advantage for Cadence.

Q: Jay Vleeschhouwer from Griffin Securities asked about the components of the share gain in core EDA and if it is concentrated in implementation.

A: Anirudh Devgan explained that Cadence has gained share in digital implementation, but recent growth is broad-based across synthesis, sign-off, and verification. The integration of these tools with Agentic AI flows is driving full-flow deployments at large customers.

Q: Gary Mobley from Stone X asked what is driving the strong bookings and backlog in a light renewal year.

A: John Wall explained that the strength reflects underlying design activity, expanded customer engagements, and add-on opportunities across the portfolio. It indicates that Cadence is becoming increasingly important to customers accelerating their engineering roadmaps.

Q: Lachlan Brown from Rothschild asked if new AI design tools are lowering barriers to entry for non-traditional designers and if new logos could become a meaningful growth driver.

A: Anirudh Devgan stated that making chip design easier leads to more design activity. The company is seeing more system companies, such as hyperscalers and others in automotive and robotics, engaging in chip design, which will continue to drive growth.