The Electronic Design Automation Consortium (EDAC) received some criticism over the last twelve months from Michael Santarini, a journalist who has covered the EDA industry for some years first for CMP ( the publisher of EDA DesignLine and many other industry publications both in print and electronic form), and lately for EDN. Mr. Santarini has questioned the inclusion of Intellectual Property (IP) revenue in the total revenue for EDA, and accused EDAC of "cherry picking" by inflating said revenue by including companies that are not members of EDAC in the total. The accusations were prompted by the contents of the 2006 report of the Market Statistics Service (MSS), a publication that EDAC has issued for many years as a service to its members.

I responded to the first point, IP does not belong in EDA, with pieces published both here (see Growth is better than death in the archives of the blogs section) and on www.gabeoneda.com. Clearly IP is a necessary component of ESL. Without design reuse there would be no system-on-chip designs (SoC), practically no transaction-level-modeling (TLM), and design cycles would so long for complex chips to discourage the great percentage of companies from even attempting them.

The accusation of "cherry picking" was more difficult to address without a complete copy of the report. You can get the report legally in one of three ways: if your company is an EDAC members, if you pay the published price for the report, or if you are a member of the press and you declare that you will not reproduce its contents in a substantive manner. I did not know of this last clause and therefore I had no access to its contents until a few weeks ago when I called EDAC and asked them to give me the same access to the report that Mr. Santarini had. To my surprise they assured me that I already had, since Mr. Santarini had turned down their offer to provide him with a copy of the report under the "non republication" clause. So, I assure them I had no intent to reproduce the findings in details, told them I would publish some of the contents dealing with IP, and promptly received a copy of report accompanied with an explanation of the methodology followed to gather the information and compile the report.

The inclusion of IP revenue, which EDAC labels SIP (Semiconductor IP) in the total EDA revenue began in 1997, so it is not new. The revenue is reported separately and can easily be deducted from total revenue if one so desires. Within the SIP sector, the outside revenue is reported separately and is clearly identified as such through footnotes. Had Mr. Santarini bothered reading the report, he would have seen the same things I saw: no collusion, not malicious intent, in fact no wrong or misleading data. No one, especially EDAC, would profit from inflating total EDA revenues because:

1. around 75% of total EDA revenues are generated by four publicly traded companies, so it is very easy to check the math,
2. financial analysts generate their own figures and would spot discrepancies with ease,
3. fabricated numbers would hurt new startups (inflated total available market (TAM)), venture capitalists (unrealistic returns expectations), and members of EDAC whose purpose for belonging is certainly not to be fooled by their own organization.

It turns out that companies that are not member of EDAC, at least not yet, generate the majority of SIP revenue. In order to provide a correct picture of the TAM, EDAC needs to consider revenue from all sources in a particular market. SIP is the only sector covered by EDAC, where its members do not generate the vast majority of revenue, thus a different method must be followed. From an actuarial point of view it is not responsible to provide a picture of a market segment relying on revenues that constitute less than 20% of total revenue. So EDAC included the three largest publicly traded IP companies in the SIP analysis. In doing so, the report clearly identifies those companies, reports their revenue separately from those of EDAC members, and provides footnotes to help the reader interpret the revenue table correctly.

It is also important to know that the input data fort handled by EDAC, but by the accounting firm of PriceWaterhouseCooper LLP. Their professionals generate the tables, so EDAC has no opportunity to "cook the numbers". Thus there is no cherry or other fruits picking involved; just an effort to present the most accurate picture possible about the SIP sector revenue, its trends, and the opportunities for EDAC members.
Just to put the matter to rest I would point out that one of the non-EDAC members included in the SIP segment is ARM, a company that now offers ESL tools and looks more and more like a traditional EDA company. Will we see ARM join EDAC this year? I hope so.

To comment on this article send email to:gmoretti@gabeoneda.com