We're driving in a fog

I went along to what used to be John Cooley’s EDA bigwigs panel and is now Peggy Aycinena’s “Is EDA dead or alive?” I had to keep asking myself am I dead or alive for the first part of the panel as each panelist insisted that their company was not suffering at all in the current downturn, there were no layoffs, and every EDA engineer gets a pony for their daughter. I may have misheard that last bit.

Peggy set the tone in her opening by stating that EDA tools are too cheap. I’ve pointed out before that there isn’t enough money flowing into EDA overall, and that it is pathetic that EDA only gets $4B out of the multi-trillion dollar electronics market. But that is different from claiming that everyone is getting tools too cheap.

Of course EDA, in the sense of software, has to continue to exist, as Ajoy Bose of Attrenta pointed out. But it is a fallacy to assume that this means that the EDA industry will continue to exist as currently structured.

Ravi Subranamian of Berkeley Design Automation described his customers as “driving in a fog” and they don’t know how long the fog will last so they are driving very carefully, in particular focusing on cost.

Diana Raggett of Javelin had just been to Taiwan. Customers are all on 4 day work-weeks going down to 3. In Hsinchu science park, three-quarters of the companies are on 1 day work-weeks. Of the 300 fabless companies in Taiwan and 500 or so in mainland China, less than half are expected to survive.

Gary Meyers of Synopsys has the evidence of quarterly results beating expectations. Echoing something I’ve heard Aart say, he related how high level management at customers is rethinking their future and willing to engage in different kinds of relationships. He continued to point out that Princeton has more people in their financial engineering program than electronic engineering program. That problem is obviously self-correcting, but I doubt that electrical engineering is necessarily going to be the beneficiary.

Scott Sandler of Springsoft amitted that customers are struggling and revenues are lower than expected. We were all shocked, shocked to hear it.

Tom Sandoval of Calypto fell on his sword too, admitting that their verification business was constrained but the power-redution business was also benefitting from people wanting low power chips right now.

Peggy pointed out that semiconductor companies are doing more internal development, and that if she was running a semiconductor company she’d want to do all internal development to keep control and to have algorithms optimized for her process. Since her process is almost certainly TSMC (unless she gets Paul Otellini’s job or the CEO of Samsung) then that is the same process as everyone else. There is not much integrated about an IDM these days. The panelists didn’t feel this was any sort of threat or was likely to make any change in the industry structure.

Peggy’s latest column was about using EDA algorithms in biochemistry, nanotechnology and other areas. But she couldn’t get anyone on the panel to really look at this as being a way to enlarge Joe Costello’s bowl of dogfood.

Suddenly the focus shifted to whether EDA was getting any of that bailout money. Did EDAC even communicate with Washington? I’ll be the first to admit that most of the money in the stimulus bill is going to be wasted, but when EDA is lining up behind GM and Chrysler I think we are looking in the wrong place.

Gabe Moretti (in a question that wasn’t a question from the audience) pointed out that EDA really is small at $4B, that electronics is only one part of design and maybe someone should look at the big picture. But it was too near the end and so nobody took the bait, we all went downstairs for a glass of free wine instead.

I think Peggy’s question about internal development, and Gabe’s about tackling the bigger picture are really key questions. Meanwhile the convention of buggy whip and related operators are focused on better and better leather and fighting for market share among their aristocratic customers.