Presentation of the 1998 Phil Kaufman Award to Ernest S. Kuh

 

Ernest S. Kuh

Presented on November 3, 1998
A. Richard Newton

Once more, it is both a pleasure and an honor to be selected to present the fifth annual Phil Kaufman Award to Professor Ernest Kuh of the University of California at Berkeley.

Following four years at Bell Laboratories in Murray Hill, Ernie joined the faculty at Berkeley in 1956. As you may know, Ernie started his professional career, in both teaching and research, in the area of circuit theory. I know many of us used the popular text “Introduction to Circuit Theory,” by Charles Desoer and Ernest Kuh, as an undergraduate text. Ernie’s work in network synthesis and approximation, and then system theory, continued until the late 1970’s. His early work on piecewise linear modeling and simulation at the circuit level was an excellent example of rigorous theory applied in a practical way, although even he admits it was just too hard to beat SPICE at that time! Many of Ernie’s graduate students from those early years have made major contributions to electrical engineering. They include Professor Sanjit Mitra, now Chairman of Electrical Engineering at UC Santa Barbara, Professor Ibrahim Hajj at the University of Illinois at Urbana, and of course Dr. Ron Rohrer, another very successful contributor to EDA in both research and practice. As Ron recently told me, “Ernie has had a profound impact on me and on my career.” Ernie’s former student Dr. Ming Chien founded First International Computer in Taiwan, now one of the largest PC motherboard manufacturers in the world.

In 1968 Ernie accepted the role of Chairman of the EECS Department and in 1974 he followed that assignment with a seven year term as Dean of the College of Engineering at Berkeley. If there was a Kaufman Award for service to academia, Ernie would certainly have won that as well. In his administrative roles, Ernie set our Department and our College on a new course in many areas, especially in its relationships with industry. On Ernie’s watch, we were able to develop a much closer and more integrated relationship with the semiconductor and emerging EDA industries, leading to the Berkeley Industrial Liaison Program and culminating in the addition of an entire floor to the Electrical Engineering building in 1982, called the Berkeley CAD/CAM Center. Ernie certainly played a key role in the development of the CAD/CAM Center.

While Ernie has made very significant contributions in all of these areas, the reason we honor him here this evening is for his contributions, both directly and indirectly, to the EDA industry. It was during his term as an administrator that Ernie decided to change the direction of his research and move into a more applied field, in particular the physical design of circuits and systems. Such a radical change is not at all easy! Especially in the middle of one’s career, not to mention moving from a theoretical basis to a very practical one, where you really do have to start almost from scratch, developing new relationships and learning the ins and outs of building software systems. A key insight for Ernie came when he was visiting his friend and well-known circuit and system theorist Brockway MacMillan at Bell Laboratories in 1975. He was introduced to a group working on the automation of PCB placement and routing and he realized how useful his understanding of graph theory might be when applied to these problems. So Ernie began the transition by bringing his theoretical understanding to the physical design arena, initially with printed circuit boards and then to integrated circuits and systems. As Professor Alberto Sangiovanni-Vincentelli recalls: “When I first came to Berkeley in 1975 as a visitor in Ernie’s group, he opened my mind to many interesting and new problems in layout that combined both theory and a practical implementation.” Ernie’s early work on one-dimensional pin assignment and the introduction of the interval graph was critical, not only in the PCB area but also to channel routing as well. His work with Yoshimura of NEC on extending these ideas by introducing the use of the vertical constraint graph and applying them to channel routing resulted in a classical paper in the field and the basis for much work that followed in the area. But Professor Kuh and his students continued and expanded their work in the area, bringing together a variety of different approaches and forming them into a coherent and consistent body, and then polishing the results from both a theoretical as well as a practical perspective until the system really made sense. Soon we were to hear of entire physical design systems like BBL, BEAR, and PROUD. In fact, Ernie’s work on the PROUD system was one of the earliest practical applications of quadratic programming techniques to cell placement. With his student Dr. C. K. Cheng, now a faculty member at UC San Diego, Ernie pioneered the use of a resistive network analog for placement and for partitioning. His most recent contributions have been in the development of timing-driven physical design tools for Deep Submicron VLSI circuits, and for multichip modules, as well as the development of an accurate and efficient circuit and interconnect simulator. This research has also yielded software programs useful both in industry and in academia, including the zero-skew clock routing work with his student Dr. Rensong Tsay (now at IBM), the timing-driven placement work with Dr. Arvind Srinivasan, and the SWEC program, with Dr. Sheng Ling. More than any single contribution, I believe it is the tight and appropriate integration of timing simulation and analysis techniques with detailed layout that is the most significant part of the contribution here.

In the mid 1970’s, Professor Kuh was invited to join the Board of Directors of the young EDA company ECAD. As Paul Huang, the principal technical founder of ECAD, told me, “We wanted to work with Professor Kuh because of his unquestionable personal integrity and because of the great respect he has in the research community throughout the world, but especially in Japan and the rest of Asia. He and his students had a very strong reputation for their research in physical design of integrated circuits at that time as well.” When ECAD merged with SDA Systems to form what became CADENCE Design Systems, Professsor Kuh remained on the Board of the new company and formed the first CADENCE Scientific Advisory Board. As a member of the CADENCE SAB during that period, I can say from personal experience that it was a real pleasure for all of us on the board to work with Ernie. He made sure the company derived all that it could from its SAB!

Of course, Professor Kuh has won many major awards for his research and his teaching as you can see from his resume, including the 1996 C&C Prize in Japan and the IEEE Education Medal. But perhaps the most visible and long-lasting reward for his efforts lives on in the form of his many successful students. As well as those mentioned above, they include Ben Ting and Nan Ping Chen, both of whom worked for many years at Hughes, Chi Ping Hsu and J.T. Li, now at Avant!, and Michael Jackson of Motorola. To his credit, many of his doctoral students have chosen academic careers in the EDA area, including Tim Cheng at UC Santa Barbara, Massoud Pedram at USC, and Wei Ming Dai at UC Santa Cruz.

So, in summary, on behalf of Mrs. Kaufman, EDAC, and all present here tonight, I am both pleased and honored to be able to present the Phil Kaufman Award to Professor Ernest Kuh, another who has played a central role in creating the EDA industry, through his teaching and his research, but perhaps most importantly through the generations of students he has mentored and who have carried his ideals, his values, and his passion for excellence throughout the academic community and the EDA industry.