Detailed Narrative
Agentic AI leadership and the three-layer platform
Cadence framed the quarter around the arrival of the 'agentic AI era,' presenting its platform as a three-layer cake: accelerated compute/data at the base, principled simulation and optimization in the middle, and agentic AI on top. At CadenceLIVE Silicon Valley 2026 it introduced AgentStack, a head-agent framework for its AI Super Agent, plus two new super agents—ViraStack for analog/custom design and InnoStack for digital implementation and signoff—joining the existing ChipStack for RTL design and verification. Management argued the greatest value comes from tight coupling of agents to physically accurate base engines, which agents invoke at scale to materially expand EDA consumption.
IP business — third year of strong growth
IP grew 22% YoY, driven by AI, HPC and automotive workloads and demand for the Star IP portfolio across interface, memory and foundation IP. Devgan cited three drivers: superior PPA quality from a rebuilt R&D team yielding competitive wins at marquee accounts; portfolio expansion (HBM acquired from Rambus and improved, UCIe developed organically); and new foundries (Samsung, Intel, Rapidus at advanced nodes; GlobalFoundries and others at mainstream). Cadence closed a record IP deal with a leading global foundry—its largest engagement with that customer, explicitly not Intel, on a new 2nm node with broader content.
Core EDA and hardware — best quarter ever
Core EDA revenue grew 18% YoY. Hardware demand accelerated to its best quarter ever, led by AI/HPC customers plus automotive and robotics, with Palladium Z3 driving multiple competitive displacements as the emulation gold standard. Verisium SimAI and Xcelium verification gained momentum and ChipStack generated large numbers of evaluations. Cadence Cerebrus-led digital continues to gain share at advanced nodes; a global semiconductor leader significantly increased Innovus usage and adopted digital signoff, and a marquee AI infrastructure/ASIC company expanded signoff usage in leading-edge designs.
System Design & Analysis and physical AI
SDA grew 18% YoY on multiphysics simulation and 3D-IC adoption. With the closed Hexagon D&E acquisition adding structural and multibody dynamics, plus prior Millennium, Cascade and BETA assets, Cadence says it now has a complete physical-AI middle-layer solution and roughly $1B SDA run rate. Focus areas: integrating a full CFD/structural/multibody flow via an agentic system-design flow; improving solver performance (potential order-of-magnitude gains via GPU acceleration and AI surrogate models); and strengthening go-to-market with Hexagon's team and acquired resellers. Management maintains physical AI will eventually be far larger than data-center AI.
EDA share of R&D spend
Devgan noted EDA has historically risen from ~7% of customer R&D spend to ~11%, and sees real potential for that share to increase further with agentic AI even as customer R&D budgets themselves grow significantly. He said every major customer CEO he speaks with wants to invest more in automation and compute, though Cadence prefers to 'print' results rather than predict the trajectory.
Hexagon integration and 2026 as an investment year
The Hexagon D&E acquisition makes 2026 a dilutive integration year: $160M revenue at a 5-10% margin, ~$0.28 EPS dilution driven largely by lost interest income (30% stock / 70% cash consideration), with accretion expected in 2027. Reported operating cash flow guidance also carries ~$180M of preclose Hexagon tax liabilities, economically part of consideration; adjusting for it, OCF outlook is ~$2.1B, roughly $100M above the original guide. Organic incremental margins are described as closer to 60% than 50%, mirroring the BETA integration arc from 2024 into 2025.