Are We Looking Beyond the Numbers in the 2026 Functional Verification Study?
For more than two decades, we’ve tracked the evolution of functional verification across the semiconductor industry. Yet the findings from the 2026 Functional Verification Study raise some uncomfortable questions.
Consider this: reported first-silicon success for IC/ASIC projects has fallen to just 5%, while approximately three out of four IC/ASIC and FPGA projects are missing their original schedules.
The real story isn’t any single statistic. It’s what these findings reveal when we examine them together.
Join Me for an Upcoming Webinar
Title: Beyond the Numbers: What the 2026 Functional Verification Study Reveals
Date: Tuesday, November 10, 2026
Time: 8:00 AM Pacific
Presenter: Harry Foster, Chief Scientist Verification, Siemens EDA
Beyond Traditional Verification Complexity
For decades, we’ve managed verification complexity by dividing large problems into smaller, more manageable pieces. We verify individual components, interfaces, and functionality using specialized methodologies.
This separation of concerns has been enormously successful—and remains essential.
But modern SoCs are increasingly software-driven, architecturally diverse, and dependent on interactions across hardware, firmware, power management, safety, and security.
A component can function correctly in isolation, yet its interactions with other components can expose unexpected failures.
The challenge isn’t simply that designs are becoming larger. It’s that the number and diversity of interactions influencing correct SoC behavior are growing.
A Different Kind of Functional Verification Webinar
In this webinar, I’m taking a different approach from our traditional Wilson Research Group study presentations.
Rather than walking through a large collection of charts and statistics, we’ll focus on selected findings and explore what they may tell us about the changing nature of verification.
We’ll examine quality and schedule outcomes, software-driven designs, chiplet and 3DIC architectures, safety and security requirements, and the growing adoption of AI.
Importantly, the study doesn’t establish a single cause for today’s verification challenges. But the combined findings raise questions worth exploring.
Building Confidence in the Complete SoC
My central argument is straightforward.
Separation of concerns helped us manage domain complexity. Our next challenge is managing the interaction complexity that emerges when those concerns converge.
The goal isn’t to abandon proven verification methodologies. It’s to build on them while expanding our ability to establish confidence in the complete SoC.
I hope you’ll join me for this discussion.
