Two of the eight questions submitted (WEB #1 and WEB #7 below) ended up being posed to the panelists.
What do the panelists see in the next 1-3 years for ESL? Has enough critical mass built where the move to a new level of abstraction is imminent? If not, what else is needed?
acquisitions vs. in-house/ organic R&D innovation, what do you prefer and why?
WEB #4: With investment in semiconductor startups out of favor, how can a chip startup get off the ground with a survivable budget for EDA tools?
WEB #5: When do you think the process roadmap will stop, for all practical (cost-feasible) purposes?
WEB #6. What are your companies doing to reduce the insane NREs (both design and mask costs) that have caused the reduction of design starts?
WEB #7. What is the most effective way to retain top talent at your EDA companies? Are stock options still useful?
WEB #8. Why can't you guys agree on a power format, move on, and let designers do their jobs with best available tools?
So, what were the most common reasons for not using a question?
First, from the perspective of panel organizer (and 46th DAC general chair), an overriding goal was to have a positive and forward-looking discussion. I wanted the panelists to discuss strategy, not tactics. Questions that had a premise of negativity, or that would obviously lead to panelists sniping at each other in some uninteresting way, were dismissed.
Second, two of the three companies were in their end-of-quarter "quiet period". Questions that dealt with topics such as pricing model evolutions would have been tricky for these companies' CEOs to answer.
Third, I did not put forward questions that would ask CEOs to discuss near-term, tactical competitive plans. Such questions were likely to receive non-answers and hence waste valuable time.
Fourth, some questions simply (i) weren't as interesting (to me) as others, or (ii) would lead only to "obvious" answers ("Should TSMC be a member of EDAC", "Would you support a significant EDA improvement in efficiency", "Do you see EDA heading to Detroit's fate", etc.), or (iii) were redundant with other questions (e.g., many questions about pricing).
Fifth, with each question expected to require up to eight minutes (moderator, plus two minutes per panelist, plus transitions and followup), not too many questions could be asked during the 75-minute panel.
It is to the panelists'and moderator's great credit that they moved the discussion so briskly, and managed to cover 14 questions along with a round of "closing thoughts" commentary.
A final point: With the logistics structure described above, some PowerPoint slides needed to be prepared in advance in order to start off the discussion. Thirteen questions were ready in the pre-session PowerPoint file, including questions taken from the web submissions.
Mainly, it was agreed in advance that while audience questions came in by SMS, the discussion would start off with three "high-level" or "top-down" questions:
1. What impact will the global economic slowdown have on EDA?
2. What do customer CEOs seek from EDA and their supply bases?
3. What impact will reduced fabless VC investment have on EDA?
In closing, I would first like to thank all of the members of the community who submitted questions by Web, or by SMS during the panel session.
Some questions were signed, but I have kept everything anonymous in this article.
I invite authors of questions to take credit for themselves in followup comments.
Second, to the extent that these questions reflect the concerns of our industry at least, the demographic who attended DAC, who attended the session, who were SMS-capable, etc. I hope that they can stimulate further thinking and discussion.
Third, and finally, I would like to again thank the three CEOs and moderator who so graciously and capably opened the 46th DAC program with "Futures for EDA: The CEO View".
At least to me, the discussion reinforced why these gentlemen hold their positions of leadership in our industry.
Andrew B. Kahng is Professor of CSE and ECE, UC San Diego and conference chair of DAC 2009.