EDA’s Need to Focus on Research
Al Dunlop
President
IEEE Council on Electronic Design Automation
As the EDA industry evolves and works to heighten its stature and importance within the Semiconductor industry, strategic investment in research –– and not just development –– has never been more of an imperative.
Let me draw a parallel to the Semiconductor industry to show why this investment strategy has moved from important to critical. As we all know, Semiconductor research and processing started within systems companies, with some research coming from academia. As time went on, the industry lived through a host of mergers and we saw the research in silicon processing consolidate within the foundries, the largest systems houses and academia. This research transitioned along with the processing in a systematic fashion.
By contrast, over the past 25 years, the business of EDA has moved from systems houses to a new commercially viable industry, but the research did not make the complete transition. Instead, much of it was picked up in academia where the link to commercial endeavors is not as tight or as seamless as it is in the case of the Semiconductor industry.
Of course, many readers will note that EDA has used entrepreneurial startups as the point to connect research to development, a strategy that has worked for many years and follows a long-standing Silicon Valley tradition. Nevertheless, tough, real-world problems remain to be solved by EDA and the way to solve them is to apply theoretical research to practical applications through close partnership between research and development.
We are at a point where the current way of doing business needs to change for EDA to be able to support future design requirements.
Clearly, an optimization of the entire design flow, all the way from the design house specification to the processing lines, is needed and is best achieved by multi-disciplinary industrial laboratories. Such laboratories could also coordinate efforts between academia, entrepreneurial startups, and internal research to create ingenious, commercial solutions to critical problems. The creation of an applied research lab for advanced technology research under a corporate umbrella could be just the place for this to occur.
Within this structure, R&D can be a seamless collaboration of talent, ideas, information and resources that benefit a company and help establish it as an industry leader. Cadence Design Systems recently made a move in this direction with the dedication of Cadence Research Labs, which is taking over Cadence Berkeley Labs.
The yearly Design Automation Conference has proven to be a catalyst for EDA breakthroughs. Much of the research and industry continuity is found at the conference, where a strong technical program highlights recent developments. With a re-energized focus on research, more breakthroughs would be presented, firmly positioning EDA at the forefront of advancing technologies. More important, an acceleration of breakthroughs in EDA could help spur emergence of the more dynamic economic engine that is urgently needed in both the semiconductor industry and EDA.
Increased investment in and attention to research speaks positively of our industry, and sends an optimistic and positive message about the future of EDA. I urge companies of various sizes and disciplines to make a strategic decision to invest in the long-term future of EDA.