The Verification Problem Is Changing. Are We Asking the Right Questions
There has never been a more exciting time to be working in semiconductors.
Silicon is rapidly becoming the intelligence underneath almost everything around us from AI infrastructure and autonomous vehicles to factories, robots, connected cities, and systems operating in space.
As silicon takes on more intelligence, more autonomy, and more responsibility, however, something else is happening:
The verification problem is changing.
At DVCon India 2026, Abhi Kolpekwar, Senior Vice President and General Manager of Design Verification Technology at Siemens EDA, will explore that change in his Vision Talk, “The Future of Verification—One Path to Silicon Confidence.”
I believe it is a conversation worth making a special effort to attend.
More verification capability. But are we more confident?
Over the past several decades, our industry has made extraordinary advances in verification.
We have more compute. Better verification engines. More sophisticated methodologies. Greater automation. More data. And increasingly, AI is becoming part of the verification process.
By almost every measure, our verification capability has increased dramatically.
Yet the outcomes are not necessarily moving in the direction we would expect.
New 2026 industry data that Abhi will present at DVCon India points to a troubling pattern: reported first-silicon success continues to decline, schedule pressure remains persistent, and verification still consumes a significant share of project effort.
That creates an important paradox:
Why hasn’t more verification capability automatically translated into more confidence?
Abhi’s talk begins with that question. But getting to the answer requires looking at how the verification problem itself is changing.
Complexity is no longer just about size
For years, semiconductor complexity has been defined largely by scale: more gates, more IP, more software, and more states to verify.
That challenge remains. But another form of complexity is becoming increasingly important: interaction complexity.
As previously separate parts of the system become more tightly connected, verification becomes harder to divide into clean, independent problems.
Abhi will share new 2026 industry data that points to this shift—and why it matters.
Is verification closure enough?
Verification has traditionally been very good at measuring progress toward closure.
But as system behavior becomes more interconnected, an important question emerges:
Are our traditional measures of completion enough to give us confidence in the complete system?
That question is at the heart of Abhi’s talk and leads to the idea of Silicon Confidence.
Not just the problem—a path forward
This will not simply be a talk about why verification is getting harder.
Abhi will lay out a vision for what needs to change and a path forward.
An important part of that discussion will be AI.
AI is already appearing across many parts of engineering, including verification. But its longer-term potential raises questions that go well beyond making individual tasks faster.
How might AI help us address a verification problem that is becoming more interconnected?
How should intelligence fit into established engineering workflows?
And as AI takes on a greater role, how do we ensure that engineering judgment remains where it matters?
Abhi will explore these questions as part of a broader view of how verification needs to evolve.
AI should expand what verification can accomplish
There is another aspect of the AI discussion that I think is particularly important.
When engineering productivity improves dramatically, the outcome is not necessarily that we simply do the same work with fewer people.
Often, we do more.
We explore problems that were previously impractical. We examine more scenarios. We tackle challenges that time and resource constraints previously put out of reach.
We have seen this pattern repeatedly as semiconductor design and verification have become more automated.
AI has the potential to extend that pattern.
The opportunity is not simply to automate today’s verification process. It is to expand what verification and verification engineers can accomplish.
From vision to engineering reality
A Vision Talk should look ahead. But a useful vision also needs a path toward implementation.
Abhi will bring together new industry evidence, the changing nature of complexity, the evolving role of AI, and the future role of the verification engineer into a broader view of where verification needs to go next.
He will also show how elements of that vision are beginning to take shape in Questa One, along with examples of measurable engineering impact.
I won’t give away those results here.
You’ll have to come to the talk.
A conversation worth attending
The intelligent world we are building depends on increasingly sophisticated silicon.
And the more responsibility we place on that silicon, the more important it becomes that our verification approaches evolve with it.
At DVCon India, Abhi will ask some fundamental questions:
Are the approaches that helped us scale verification against design size sufficient for the challenges ahead?
What does confidence mean as systems become increasingly interconnected?
What role should AI play?
And what should verification become next?
This is not simply a talk about what is changing. It is a vision for how verification can change with it. If you are attending DVCon India 2026, I hope you’ll make a special effort to join Abhi for “The Future of Verification—One Path to Silicon Confidence.”
