Mousetraps, alligators and EDA

I didn’t go to the 2011 Design Automation Conference in San Diego because this year I decided to stay home and see what I could pick up virtually from video, audio and news coverage day to day.  It was much more cost effective and focused, despite the usual bandwidth problems with the convention center. My apologies for mixing metaphors but this is how I wrap up the issues arising out of DAC. 


It is said that if you build a better mousetrap the world will beat a path to your door, but when your customers are up to their asses in alligators, vermin are not their critical problem. 


Gary Smith gave his usual glowing report of the future for the industry (he thinks it’s well on the way to being a $6.6 Billion industry by 2015, and once again he claims that this year Electronic system level design (ESL) is going to be really big this year (just like it has every year for the past decade, according to Gary.)  But this year he added a big caveat.


He said that the industry needs to provide tools to the design community that they need to do their jobs.


Um, isn’t that why companies in EDA are in business? Smith’s caveat has the implication that they aren’t.  I agree with him.  Apparently so do the VCs who are pointedly NOT investing in EDA or semiconductor companies because neither can turn profits.


The problem with EDA is a simple one: failure to communicate.  


The industry largely concentrates on making tools that only sort of solve the problems their customers have, which forces the customers to hire more manpower to make those tools actually help them do their jobs, or design and maintain their own tools (Smith has said before that the primary competition for EDA tools is actually the tools their customers are making.)  The tools the customers make actually resolve the specific issues, but it keeps them from concentrating on the job at hand, i.e, making new complex ICs.


So why aren’t the companies making the right tools?  Because they have no idea what the customer really wants.


Most EDA companies start out to solve one problem for one design inside one IC company; either be a design team that spins out of another company or design consultants that discover a nifty resolution to the specific problem.  In the old days (pre 1990) that was enough to get a buttload of money from a VC and launch a new company, but this is not the old days and those one-off solutions are not the money makers they used to be.  The EDA industry, however, is still acting like it is.


Infant industries are always engineering driven. When they reach adolescence they morph to sales-driven philosophy.  Mature industries are market driven. EDA is stuck, however, between infancy and adolescence and can’t seem to break out of it.  People like Smith and investors like Jim Hogan have been telling them what to do do break out of it for years, and they all nod their heads annually to the advice, and when DAC comes to a close they go back to the cubicles and start the process all over again.


The industry needs to mature and do it really fast or there is going to come a reckoning. It will be consumed by more savvy industries that are already making inroads, judging from what I could see, hear and read from the various media portals I was monitoring.  They need to learn how to really speak with, rather than at their customers and stop using the tired old sales tools (trade shows and press releases) that have passed for marketing efforts over the past 20 years. 


If the engineers running these companies can’t figure out what I’m talking about, they need to step aside and bring real marketers into management.  If they don’t, then the investors need to start making some changes in management themselves.