Editorial

The Future of EDA According to Alberto Sangiovanni Vincentelli by Ann Steffora

Predicting the future of anything is never easy, but there seem to be a select few who make it look that way. When Alberto Sangiovanni Vincentelli and colleague Richard Newton (now Dean of the College of Engineering at the University of California at Berkeley) began their working relationship more than 20 years ago, they sat down to envision the future of chip design - and accurately predicted some of the major twists and turns that technology has taken.

As for this reflection on history, Sangiovanni Vincentelli only replies humbly, "We were lucky."

Sangiovanni Vincentelli has credentials and accomplishments in the electrical engineering field of a magnitude largely unmatched, and it was with this in mind that I witnessed the ceremony for the EDA Consortium's Kaufman Award when it was bestowed on him in early November 2001. It was refreshing to see the emotionally charged response from the industry's top CEOs - with whom Sangiovanni Vincentelli maintains close personal relationships.

Sangiovanni Vincentelli is presently the chief technology adviser of Cadence Design Systems and founded Cadence Berkeley Research Labs, an advanced technology research group. He is a member of the board of directors of Cadence, along with other EDA companies.

The Future of EDA According to Alberto Sangiovanni Vincentelli Continued…

Looking Back

This visionary and highly influential force in the industry explained the evolution of his industry predictions that began back in the late '70s: "At that time, we were scoping out what would be important in design and what we would base our work on. The kinds of things we were expecting were workstations... and in 1979, there were no workstations around except for the first experiments at Xerox Parc."

He and Newton envisioned a single-user type of workstation that a number of companies would provide. The only way in which this vision could become a reality, he believed, would be for it to be based on an open operating system, which was Unix at that time.

"We had very clear ideas about hardware execution times, and so we decided the business model for an EDA company should not be hardware based, but only software. That was revolutionary because at that time, Mentor, Daisy and Valid were making and selling their own workstations," he said.

"As late as 1980, the then-CEO of Mentor said he didn't know of any CAD company that could be successful without selling hardware - and that was 11 years after we said it was the way to go and now there is not a single EDA company that does not have their business model as essentially [providing] software."

Another idea that the Berkeley pair latched on to was that design was not just verification but that synthesis was important. "We first envisioned layout synthesis - place and route. This eliminated the layout artist. There were lots and lots of people sitting at Calma Systems doing polygon editingand by doing layout automatically, you would free the designer from the task of giving his designs to the layout editor and not knowing what was coming out afterwards." In other words, the loop was closed in some sense, he continued.

"The next step was when [Richard Newton and I] believed logic synthesis was going to be the great thing and we founded Synopsys as such," he explained, which proved to be a pivotal time in the EDA industry.

Interestingly, Sangiovanni Vincentelli said he tried very hard to convince Cadence that they needed to do logic synthesis at that time but they didn't believe in it.

"So, what the heck, we founded a new company," with, among others, the current chairman of Synopsys, Aart de Geus, who is by Sangiovanni Vincentelli's definition, "my kind," i.e., a technologist.

Many questions have arisen over the years as to why Sangiovanni Vincentelli ultimately joined Cadence. "The only reason why I chose Cadence was that it had a wider offering and my vision at that time was that EDA would become the only place where a design chain could be supported," and Cadence seemed better equipped to meet those demands.

The time period of 1990 and 1991 was a significant one in Sangiovanni Vincentelli's career as this is when he began his research on system level design/platform-based design. "The electronics industry at that time could not segment and if you assume that design would be broken up into various pieces, then there was only one way to support the design chain. Someplace, somewhere you had to have someone that knew about the entire chain of design, and that could only be an EDA company. Then you had to pick a company that was active in all areas. Cadence was then, and still is, the only company that has everything."

Now, looking back, Sangiovanni Vincentelli acknowledges the changes and maturing the industry has gone through, and credits former Cadence CEO, Joe Costello, with pulling the industry along. "A great deal of this maturing is due to Joe Costello [who] really changed the business model and changed the way in which the EDA industry is today. He was the first one to envision EDA as having great potential in the service sector; and then he was the first one to go from perpetual license to three-year and two-year license. That was a big revolution in terms of business model - it changed the industry a lot."

The Future of EDA According to Alberto Sangiovanni Vincentelli Continued…

Interestingly, Sangiovanni Vincentelli noted that if you look at the software industry today, there are a lot of companies slowly moving into time-based software licensing business models, which he feels should be recognized.

He also believes the business community should take a look at this because he feels it will be the dominant business model for the software industry. The point is, he said, how do you grow (with perpetual software licensing)? Once a perpetual license is sold, there isn't enough incentive and money to sustain the product. A continued stream of R&D money is needed for ongoing development and time-based licensing continues that stream that would be missing otherwise, he suggests.

Grade: Good, But Room For Improvement

In many areas the industry is doing well, Sangiovanni Vincentelli says, but he also sees room for improvement, including an industry leader.

"For every stage of an industry, you need a leader. If you want to mature, at some point you need someone who is very methodical and very good from the financial side," he says. For this reason, he feels Ray Bingham is the best person to have at the helm of Cadence right now. "To grow beyond the $1.5 billion dollar mark and go to the next stage, you need to be very solid in the financial business sense," which Bingham does have.

Next in line for the industry leader position falls naturally to Dr. Aart de Geus. And while he has all of the technical knowledge required, Sangiovanni Vincentelli points out that Synopsys' view on the future is "limited and looks too much at the implementation side of ICs," and not as much focus on the system level side as he would like to see.

What Should Be Next For EDA?

While he is most proud of being visionary about system level issues, it is time for these ideas to become a reality, and the industry isn't quite there yet, Sangiovanni Vincentelli says.

While he was a huge proponent of synthesis at its inception, he would also very much like to see the industry working on system level issues, hinting that it could lose ground by falling behind from a technology perspective.

"The EDA industry has tunnel vision, so all of a sudden its place and route and synthesis, place and route and that's it. Everyone does that and the battleground is on synthesis, place and route. But there's much more to life than that."

That "more" he refers to is the realization that in a short time, circuits will be designed from a higher perspective and the basic paradigm of current design will be going away, therefore, design tools must evolve as well.

"There is all this platform-based design coming up and do we have the technology to be able to support that? Well yeah maybe, but there is still a lot of stuff that needs to be done. We have addressed it but because I twisted arms all over the place. They wanted to cancel the project inside Cadence about platform-based design looking at how to design platforms and how to connect the IC companies to the systems companies, and how this company should develop that product.

Another really important issue, from the business side, is the ability of the industry to be able to articulate a better value proposition to the customer, since right now there is still the feeling from the customer side that they are paying too much for EDA, he believes. "[Many customers] are unclear of the real value that EDA brings to the party and in addition, they spend an arm and a leg to support these tools internally."

The Future of EDA According to Alberto Sangiovanni Vincentelli Continued…

What he would like to see is people say that they see the value, understand it and outsource a lot more today in terms of support and infrastructure.

"If you get to that step as an industry, you've really grown to a place where you are a solid player in the world of electronics. It is very clear to me that solution selling is the first step," Sangiovanni Vincentelli concluded.

To view the original announcement of Sangiovanni Vincentelli's Kaufman Award, click Here

Once again, thank you for reading, and please send any feedback to ann@ibsystems.com.

Alberto Sangiovanni Vincentelli

Sangiovanni Vincentelli is a world-renowned authority on circuit simulation, computer-aided design of integrated circuits, logic synthesis, and system design. With his appointment into the National Academy of Engineering (NAE), in April 1998, he attained the highest professional honor for an American engineer.

In brief, in 1971, Sangiovanni Vincentelli received his engineering doctorate summa cum laude from the Politecnico di Milano, Milano, Italy. He has served on the faculty of UC Berkeley since 1976. Sangiovanni Vincentelli holds the Edgar L. and Harold H. Buttner Chair of Electrical Engineering and Computer Sciences at the University of California at Berkeley and the Vice-Chair position for Industrial Relations. He is on the Advisory Board of the Lester Center of the Haas School of Business and of the Center for Western European Studies, and a member of the Berkeley Roundtable of the International Economy (BRIE). Sangiovanni Vincentelli co-founded Cadence and Synopsys. He was a director of ViewLogic, Pie Design System, and chair of the Technical Advisory Board of Synopsys. He has consulted for a number of prestigious US, Japanese and European companies including: Alcatel, AT&T, BMW, Daimler-Chrysler, Ericsson, Fujitsu, GE, Hitachi, HP, IBM, Intel, Kawasaki Steel (as Chief Technology Advisor), Magneti-Marelli, ST Microelectronics, and Sony. He is the founder and Scientific Director of the Project on Advanced Research on Architectures and Design of Electronic Systems (PARADES), a European Group of Economic Interest supported by Cadence, Magneti-Marelli and ST Microelectronics.

Sangiovanni Vincentelli has attained numerous awards in his career, including a 1981 Distinguished Teaching Award, the highest honor for teaching bestowed by U.C. Berkeley and a 1995 Electrical and Electronics Engineers' Graduate Teaching Award, an international honor given for the inspirational teaching of graduate students. In 1983, he was named IEEE Fellow for contributions to CAD of Integrated Circuits. He also received research awards for best papers and presentations including the IEEE's Guillemin-Cauer.