When I started covering EDA or what was then called computer-aided engineering in 1985, there was a sense of excitement. Here was something new, dynamic and rapidly growing and absolutely essential to an emerging generation of semiconductors and electronic products.
That excitement carried over into the next decade. New industry leaders emerged, designers moved up to RTL design and synthesis, and innovations such as formal verification appeared. Startups and established companies alike worked feverishly to solve the problems of deep-submicron design.
But today, when you hear the word EDA, what comes to mind? Failed companies. Flat revenues. Endless lawsuits. Marketing hype. Depressed market values. Vision and excitement seem in short supply, and EDA's reputation among investors and customers is lukewarm at best.
This, despite the fact that EDA remains indispensible for IC and board design, and that thousands of bright people in the EDA industry are working hard to anticipate customers' needs.
The causes of EDA's malaise are many and the solutions elusive. Nonetheless, here are a few perspectives.
First, the EDA industry has limited itself by focusing its efforts on digital ASIC and system-on-chip design a discipline that is just one small part of the challenging job of getting products into customers' hands. At 90 nanometers and below, fewer and fewer companies are even going to try chip design. There aren't all that many custom-chip designers in the world. But there are legions of FPGA, pc-board and embedded-software developers out there, and they have requirements that remain unmet.
Only recently did someone figure out that Matlab and Simulink from The Mathworks, with tens of thousands of users, should actually be viewed as EDA products. Here's a whole group of people who weren't counted before. The EDA industry needs to encompass more of the design flow and get out in front of a much larger group of users.
Second, there's no sense of the next big thing. Design-for-manufacturing? Most designers would rather not deal with it. Electronic-system-level design? ESL's been "new" for 15 years, and in all that time it has yet to catch on in a big way, except for Matlab and some recent usage of SystemC for modeling and verification.
Most established EDA companies' announcements are about incremental tool improvements or competitive posturing. There hasn't really been a major methodology shift since the move to RTL in the early 1990s. A few startups have put forth truly new and innovative technology, but in the main the startups have had a tough time making it in the market. Quicksilver Technologies and Barcelona Design are the most recent casualties.
Lack of technology development is not the problem. I hear about interesting technology all the time from startups, and I see it at conferences like the Tau timing workshop, the International Symposium on Physical Design and the Design Automation Conference. Rather, what's lacking is a way to bring all this stuff to market in sustainable fashion.
Perhaps it is time to step back and take a broader look at the design problems the EDA industry is trying to solve.
For example, what are all the steps involved in getting a new cell phone into someone's hands? What combination of intellectual property, services and tools will help make that an easier process? What new technologies or methodologies, from academia or industry, can be applied to the areas of biggest pain? And how can customers be made to feel they're getting good value?
Whoever can answer those questions, while also understanding that no single vendor will provide everything, will do well. Rather than just another tool, what's needed are solutions to broad-ranging problems.
Perhaps that would restore EDA to its former status as a growing industry that delivers recognizable value to its customers.
By Richard Goering, managing editor of design automation for EE Times.