As a high school student, one of today's foremost authorities in EDA loved the classics. Yes, Dr. Alberto Sangiovanni-Vincentelli probably would have followed that track in college if not for the influence of two of his father's friends.
"I loved philosophy, history, and ancient Greek. I actually hated anything that had to do with building things," he says. But in his hometown of Milan, all good students were expected to go into engineering, the most challenging studies—or at least try it, he adds. "Two dear friends of my father, Raffaele Girotti, then CEO of ENI, the Italian oil and energy company, and Raffaele Mattioli, then CEO of the largest Italian Bank at that time, Banca Commerciale Italiana, encouraged me to enter the Politecnico di Milano. It was the best university in Italy at that time."
As a result, the world now looks to the Edgar L. and Harold H. Buttner Chair at the Department of Electrical Engineering and Computer Science at the University of California, Berkeley, for expertise in EDA, embedded system design, formal verification, and hybrid systems.
He is a consultant, scientific director, organizational founder, board member, and advisor to a number of the world's top corporate, academic, and government technical institutions. (See "EDA Around The World" at www.elecdesign.com, Drill Deeper 11202.) He's written 15 books and more than 700 technical papers. And, he and Dr. Richard Newton founded Cadence Design Systems in 1983 and Synopsys in 1987—two of the leading EDA companies.
Once Dr. Sangiovanni-Vincentelli "chose" engineering, he had to select a specific field. He opted for nuclear engineering but then switched to EECS.
"I found out that electrical engineering computer sciences (EECS) had more mathematics than nuclear and that my dearest friends were doing EECS! No deep reason!" he says. "I ended up doing my thesis on network and circuit theory. But I also did some research work in operation research, control, and computer science."
The pattern for multiple focuses was thus set. "These eclectic interests, including classics, were probably the reason why my career has been involved with several disciplines," he says.
During his early years at Berkeley, he worked closely with civil engineering researchers such as Dr. Pister, then Dean of the College of Engineering, looking at methods for designing earthquakeresistant structures. Don Pederson and Richard Newton of the Department of EECS at Berkeley and Bob Brayton of IBM strongly influenced his interests in EDA, which became his focus.
Sangiovanni-Vincentelli subsequently found success working with hybrid systems, a control discipline, and formal verification, a field driven by computer science. Lately, his interests have turned toward embedded system design. He's worked for the automotive industry in this field, which serves as a bridge with mechanical engineering.
The motivation behind this diversity, he says, comes down to "intellectual curiosity." Of course, the professor at the University of California, Berkeley, who held visiting positions at MIT and Italian universities such as Politecnico di Torino, University of Pisa, University of Pavia, University of Bologna, and the University of Rome, also enjoys "the great experience of educating people and seeing them bloom into world-class researchers."
Safety issues to smart dust
Sangiovanni-Vincentelli believes the complexity of software, combined with the pressure for low costs, raises important challenges like safety.
"We cannot afford to do designs the way we do them now and guarantee safe operations of our devices," he says. "Today, embedded software, i.e., the software that is used to program devices that go into common products, such as cars, airplanes, and appliances, is so complex (more than 1 million lines of code for the embedded software for cell phones) that its verification takes a large amount of time and resources.
"We have seen massive recalls, such as the recent Mercedes one that recalled more than 1 million vehicles and cost around $500 million, and catastrophic outcomes such as the Mars Lander failure and the Ariane satellite blowup caused by software errors. There is a significant amount of research devoted to new design methods that should improve substantially the design time and correctness of our embedded devices."
He also expects the rate of new technology introductions to slow down and cross-use platforms to be developed: "In addition to the embedded software problem, the decrease in feature size of ICs that makes embedded electronic pervasiveness possible will also generate significant challenges that may slow down the introduction of new technology. Costs to develop manufacturing plants and to design novel integrated circuits will reduce the number of design starts significantly in favor of programmable platforms that can be reused across application boundaries."
Finally, Sangiovanni-Vincentelli dwells on ad hoc wireless sensor networks, pioneered by Professor Pister of EECS. It was Pister who proposed the widely recognized term "smart dust" to underline the small dimensions and computing power of the nodes of the network.
These devices are bound to influence how we live in ways we can hardly imagine, way above and beyond what the Internet has done in the past. From health and environment monitoring to security, from manufacturing plants to traffic, wireless sensor networks will have a long-lasting impact.
"The social implications in terms of security and privacy are subject to intense investigation and may substantially impact the way we design these networks as well as the legislation that must be developed to regulate their use," he cautions. "The amount of information about our actions and movements that will be available is posing questions that we never before had to answer. What price will we have to pay to be more secure?"
For more about Sangiovanni-Vincentelli, see "The Envelope Please," Drill Deeper 11203, "Four Key Insights," Drill Deeper 11204, and "The Doctor Takes On The World," Drill Deeper 11205, at www.elecdesign.com.