EDA Numbers Drop Again
Looking for good news in the EDA numbers this week is like looking on the bright side of a multi-car collision. Even if no one gets injured it’s not exactly something you’ll want to forget about as quickly as possible.
The downturn in EDA this year is abysmal. And while it’s true the EDA market didn’t start dropping until the second half of 2008—about a year after the semiconductor market began—semiconductors started rebounding in Q1 while EDA has not. But that’s almost beside the point. Excluding changes in financial reporting, this is really the first downturn in EDA’s history.
Ironically, this comes at a time when designing chips has hit an all-time high for complexity. Digital, mixed-signal and even some analog chips are getting to the point where the complexity is so great that it requires sophisticated tools. Even FPGAs have reached the point where designs are too complex to do everything by hand.
Wally Rhines, chairman of the EDA Consortium and CEO of Mentor Graphics, said EDA is a trailing indicator. He gave no indication of when things might turn around, but did point to a few narrow bright spots. First, signal integrity tools are up 8%. Considering that signal integrity wasn’t a problem at earlier nodes, this is not particularly surprising. It’s now a broad area of concern for both ICs and printed circuit boards.
Second, hardware-assisted verification is up 6%. Designs are simply getting too big to simulate, which is why it’s taking so long to finish all the verification. And that doesn’t even take into account whether the coverage models are complete and accurate.
Finally, resolution enhancement—part of design for manufacturability—is up 50%. That’s built on a rather small starting point, but it still shows the growing difficulty in getting chips from design to manufacturing—an area ripe for additional tools.
Given these demands, our guess is the overall EDA market will turn around relatively quickly. New nodes will require new tools, and even the most sophisticated companies have stopped building tools where they’re commercially available. The big question is which companies will benefit most coming out of the downturn—and which ones won’t be considered necessary at all.
–Ed Sperling
Tags: Mentor Graphics
October 2nd, 2009 at 10:53 am
EDA killed itself.
The all-you-can-eat deals helped to slow innovation. Without innovation in EDA, the design cost shot up, which in turn kills-off new chip companies. Without new chip companies, the foundry business cannot grow.
The next thing you will see is foundries start to kill off EDA in order to lower the design cost. This is critical for their growth, if not survival.
At least one foundry has the ability to do just that. It will suck the air out of the ecosystem for its competitors.
Watch and enjoy!
October 2nd, 2009 at 12:55 pm
EDA is a lagging indicator, EDA is a leading indicator, EDA has it’s own cycle…. I’ve heard the this over and over. I don’t think EDA knows how to measure business.
October 9th, 2009 at 10:51 pm
This is six consecutive quarters of year over year decline in total EDA industry revenue. EDAC’s big initiative for 2009 seems to be anti-piracy, I am wondering if that is what is really going to grow the pie. I would like to see more focus on enabling innovation. EDAC should return to it’s roots as an organization for startups.