The Economics of DAC in 2026

The Booth Is Bigger Than the Business. That's Not an Accident.

A series on what DAC 2026 actually told us about the state of EDA.

Walk the floor at DAC long enough and you start reading it less like a trade show and more like a funding announcement.

I said in my LinkedIn post that a lot hasn't changed since the DAC of my twenties and thirties - incumbents defending their position with ever more detailed complicated messaging, whilst hijacking the buzzword of the moment, and on the periphery, smart people, carving out narrower and narrower niches of value in an increasingly complex world. There was something else going on at DAC in 2026, and it's the one that stopped me in my tracks this year.

Somewhere on that floor, a handful of very young companies with few employees, a couple of years old, in some cases barely past seed funding running exhibits with production quality, staffed presentation theaters, and private meeting rooms that rival or exceed what I remember the incumbents doing at DAC when I was on the inside of one of them. Some of these companies are sponsoring at tiers that put them, on paper, above the two companies that between them still define this industry.

And here's the part that should stop you too: if you go looking, as we have, for the thing that's supposed to justify that kind of visibility - a credible customer, a named tapeout, a production deployment you can actually point to - it mostly isn't there yet. What's there instead is a strong story, a few marquee logos who may be pilots as much as customers, and a very confident narrative about where things are headed.

My first instinct is that this is backward. If I were running one of these companies, I'd want every dollar going into the two or three design partners who could actually turn into a referenceable win - the unglamorous work of sitting inside someone's flow, fixing what breaks, and coming out the other side with a chip that taped out using your tool. That's how credibility has always been built in this industry. It's certainly how the incumbents built theirs. So why set that money on fire in an exhibit hall instead?

I've spent some time this year on this specific question - marketing $ vs execution $. Digging into what these companies actually spend at an event like this versus what they've actually got to show for it has been a fascinating exercise. I don't think the answer is that they've forgotten how enterprise trust gets built. I think they're building two different kinds of credibility at once, for two different audiences, on two different clocks, and the DAC floor only shows us one of them.

The room isn't just customers. It's also the next round of investors deciding whether to write a check, the handful of engineers in this industry who can actually do this work deciding where they want to spend the next three years of their career, and the analysts and press who will spend the next twelve months writing the "state of the category" story that everyone else reads before they've even evaluated a tool. A visible, expensive, well-run presence does a job for all four of those audiences simultaneously, and for three of them it does that job whether or not a single customer conversation converts that week.

Being visible is itself the signal, precisely because the proof isn't there yet. This is the part that looks irrational until you sit with it. In most software categories, a buyer can trial something, see if it breaks, and walk away with limited downside. In our world, a chip design or verification tool that fails mid-flow doesn't cost you a subscription - it can cost you a re-spin, a blown schedule, and a very bad year. Buyers here are unusually rational about vendor risk, and unusually reluctant to bet on someone who might not exist in eighteen months. A serious, well-funded, hard-to-fake presence at the one event that matters is a way of answering the question nobody asks out loud: will you still be here when my chip tapes out? It's a substitute for the reference you don't have yet, aimed at exactly the anxiety a missing reference creates.

It's also a land grab on the category itself, not just the deal. Get enough of the room saying "AI belongs in this flow" before anyone has fully proven whose AI it should be, and you've shaped the conversation in your favor regardless of who eventually wins on merit. That's a genuinely different objective than closing the prospect standing in front of you, and it explains a lot about how these budgets get allocated.

And the design-partner work isn't being skipped - it's running on a different clock. Our industry's cycles are long. Eighteen months, often longer, from first serious conversation to a design that's actually taped out. A big presence this year isn't proof of last year's work; it's the top of a funnel whose bottom won't mature until 2027 or 2028. Used well, the quiet, deep integration with a handful of real partners and the loud, visible presence at an event like this aren't competing for the same dollar. They're sequenced. One builds the reference you'll be able to point to in two years. The other keeps the pipeline full of the next fifty prospects while the first five mature.

None of which means every dollar of this is well spent, or that the story being told always matches the substance behind it. It doesn't, always. And I think that gap - between how big a company appears on the floor and how much production reality sits behind it is going to become one of the more useful diagnostic questions in this industry over the next few years, for buyers and for competitors alike. Not "how good is the demo," but a plainer question: name the tapeout. Not the pilot, not the logo on a slide, not the deployment count - the shipped design, the one where their tool was actually in the flow when it went out the door.

That's the yardstick I'm going to keep coming back to as I write more of these. Are our new friends with the huge presence and no track record playing a new long game? Will they appear next year (or the year after) with some mindblowing chip stories (the new phone, the new drone, the new datacenter all run chips developed on our tech - here are our customers to tell you more!).