I have learned a lot in the last twenty-four hours. I am referring to the discussions I have had regarding Mentor's assertion that it is the second largest EDA vendor in terms of revenue. As you might remember, I wrote an article, you can find it here, regarding Jay Vleeschhouwer analysis of the four largest publicly traded EDA companies in 1Q10. In it I questioned Mentor's assertion that they are number 2 in terms of revenue. Iwas looking at total revenue for the year. I have since learned that is not what I was supposed to do. Read on for the full explanation.

Number 2 In What?

I was taken to task by Mentor's Ry Schwark who pointed out to me that the claim was corroborated by Gary Smith's analysis of the 2009 fiscal results for Mentor and Cadence. You can read his email to me as a comment to the article referenced above. By the way if you cannot read it, just register or login and you will be able to read it. The bottom line is that Mentor was the second largest revenue producer in 2009 if you deduct the consulting revenue generated by Cadence. And this is done according to Gary Smith's rule of EDA revenue counting.

I could not believe that our industry could measure itself based on rules that choose which revenue are allowed and which are outside the pervue of the industry. After all it was only a few years ago that EDAC was indited by many editors and analysts for counting IP revenue as part of EDA. Now, of course, no one even pays attention to such dribble: of course IP is part of EDA, just ask Synopsys or Cadence.

So, since I found the omission of Consulting revenue, which by the way is called "services" by both Cadence and Synopsys, from the analysis perplexing, I asked Gary about the exclusion of this revenue from his analysis. I honestly expected a rebuttal. Instead here is Gary's answer:

When we (Dataquest) define a market it needs to be measurable (the Data part
of Dataquest). Design consulting is a wide spread market composed of
thousands of small design houses, which is almost impossible to survey. If
we counted the EDA vendors consulting revenue we would need to count the
rest to get our sub-apps market share. Since we can't we do not count that
as EDA revenue. We look at product revenue, which is the core of an EDA
vendors business. That really determines their "market" position.

I have to say that by "product" Gary means not only licenses, but also maintenance and support revenue. So now I must admit that indeed Mentor has the right to claim to be number 2 in EDA revenue as defined by Dataquest previously and Gary Smith EDA now.

It seems to me that Gary measures market percentage as a derivative of revenue, thus he must have a fairly complete picture of the entire market before he can admit a particular type of revenue in his calculations. Fair enough. But this does not satisfy my need to encourage the EDA industry to look at itself, admit it has self restricted itself to a role of tool provider. An industry where companies, especially the large ones, use discounts to enlarge their customers base. An industry that is unusually litigious often just in order to avoid loosing a deal. An industry that still believes that helping to build hardware is where the money is.

The Future Is Now

In my opinion it is very important to be able to track the "services" business within the EDA industry. Gary wrote to me that if he considered the services business, then Wipro might be the largest EDA company. Well, what if Synopsys bought Wipro? Or the other way around?

To Mentor being number 2 in the traditional and countable EDA market has its merit. It shows that the company has sold licenses and maintenance in a mix that has generated a greater revenue than Cadence. Since Cadence felt from heaven in the Mike Fister times, (the times of countable shoes according to John Cooley) surpassing it shows accomplishments. Mentor stayed the course, worked hard, and let Cadence pass them on the way down. All this is good. But it says absolutely nothing about the ability of Mentor to grow in the future. After all, how many more companies can Henry Potts buy? It does not even mean that Mentor has a larger installed base than Cadence, since license prices vary considerably from market sector to market sector within EDA. It does mean that Mentor had greater revenues in the 2009 fiscal year than Cadence in the "traditional EDA market": the countable one.

How much design work is farmed out to third parties is an invaluable measure of the attitudes of system design companies and may in fact help in planning license renewals, license pricing, and employment trends. I agree there are thousand of consultants, many making less than $100k/year in the service business. But then again, EDAC does not pretend to report all of the revenue in the IP business, just enough of them to show a trend. So the fact that the entire membership of service providers is "uncountable" does not mean that "services" are not part of EDA. What are we going to do when EDA companies start selling software IP? That will be another "uncountable" market. This ranking approach will freeze the order to its historical positions when things were as intended by the founding fathers and EDA vendors stuck to only selling tools licenses.

EDA seems anchored to a past where business was conducted in a certain way because the business climate was different and the role of EDA vendors was simply to replace the internal CAD department. As we have witnessed, this approach to business is self confining and results in a size of the EDA market that is a pittance of what it could be. There is a reason that the past has passed. It is not to be repeated, else progress and creativity would stop.

The rest of the electronics industry, in the mean time, has embraced a business model that underlines collaboration, which of course includes consulting. EDA vendors must become partners with their customers, and that includes consulting for them. One of the way that EDA will grow its revenue is by growing its "services" business as one way to become a valued partner in the development of profitable end products and share in the returns. As long as we are nothing more than tools providers, we are open to price pressures that make us look like beggars fighting for the small amount of money our potential customers are willing to allocate to overhead costs. We obviously have forgotten the dog food bowl made famous by Joe Costello. We continue to eat from it, just ignore its size.