Interview with Jen-Yi Wuu

Featured Engineer

Jen-Yi Wuu

Interview with Jen-Yi Wuu - PhD candidate and DFM intern at Mentor Graphic

What are your favorite hardware tools that you use?

As a software engineer, I seldom play with some of those cool gadgets. (I think the last time I played with a hardware tool, a FPGA board, is … 9 years ago.) I spent the majority of my days in front of a computer screen. I consider this a blessing as well as a curse. What excites me is the ever-advancing power of modern computers. I think I always have a child-like attitude inside me, amazed at how the computer is like a living beast, capable of doing more and more each day. I feel like I learn a bit more here and there every day, yet what remains to be learned never seems to become less. This absolutely humbles me and makes me appreciate every change of learning. On the other hand, I also desire to spend less time in front of my computers. I admire my advisor a lot since she usually travels back to her home country for a month every summer. During that period she leaves all her emails behind. I wish to develop that kind of discipline for such a “digital break.”

What are your favorite software tools that you use?

I’d say vi, scripting language, and Google search. The first two are the most commonly used tools for my daily tasks. They save me lots of time and energy. As for Google search, I’ve always told my friends, I can hardly imagine how the life doing research without Google is like. (I am pretty sure my “library muscles” die before they get a chance to develop.) It’s exactly like one of the old old Chinese saying, “A good student can know everything happening in the world without stepping out of the door of his house.” Now, it’s not restricted to good students. Everyone can do it.

What is on your bookshelf?

At my office, I keep some O’Reilly utility books, such as “Linux in a Nutshell” and “C++ in a Nutshell.” At my home, for my personal readings, I have a large pile of books on my to-read list. I have quite a few Christian books that I enjoy very much, such as “A Praying Life” by Paul E. Miller and “Desiring God” by John Piper. I am also reading some photography books, such as “The Photographer’s Eye” by Michael Freeman and “Within the Frame” by David duChemin, for my little leisure time hobby. Did I mention I have lots of cookbooks? I love cooking! When I miss my tasty hometown Taiwanese food, I try to cook them myself. Cooking is an art, plus that it is satisfying to my stomach and saves me money! If you have a chance to travel to Taiwan, don’t forget to taste all our delicious foods! Taiwan is not just famous for its semiconductor industry.

Do you have any tricks up your sleeve? (special way to analyze circuits, special process you use to make something, etc.)

I am a visual thinker. Whenever I encounter a problem, I try to first visualize the problem by drawing pictures or diagrams on my scrap papers or erase board. (I should have included scrap papers and erase board to my favorite hardware tools.) This is also my favorite way to present my thoughts and ideas to other people. I also find this approach quite effective when analyzing experimental data. Sometimes the collected data contain important information. However, if they are not plotted in a smart way, the critical information may be hidden or ignored. In addition, I found that keeping an attitude of open-mindedness enhances my learning. I always try to remind myself that I should embrace any challenges or criticism. People may or may not provide fair comments. But if I keep my mind closed and always assume I am always correct, I will no longer be teachable.

What are you currently working on?

I am currently working on a project building an approximate physical verification tool for lithographic hotspot detection and prediction, based on statistical machine learning techniques. This work will be presented in the upcoming DesignCon in February 2011. It is an “approximate” verification tool because it may not be able to match the accuracy and robustness of a model-based lithography simulator. However, its strength lies in its fast runtime and predictive capability. In other words, it is still on the accurate side but performs with much faster runtime and predicts better than pattern matching-based tools. Pattern matching-based tools are computationally efficient but lacks the capability to accurately predict unseen test samples. Our lithographic hotspot detection and prediction tool will be suitable for use in early design stages, where the hotspot fixing cost is much cheaper.

What challenges do you foresee in our industry?

I feel sort of unqualified to answer this question, since I am only a PhD student with just a few years of experience in this field (EDA industry, DFM area in particular). However, I will share based on my understanding and I always welcome any comments. I believe close collaboration between designers and manufacturers is important for the continuing advancement of feature size reduction. Studies have shown that IC manufacturability and yield are greatly affected by the design style. During the design stages, the issue of manufacturability needs to be carefully considered. Besides, it is also a great challenge to model or abstract the effect of fabrication process in a form that can guide the designers’ decisions. Take the lithographic hotspot problem as an example. It is desirable for designers to implement lithography-friend layout style and hotspots need to be detected as early as possible. When a hotspot is detected, current fixing guideline are mostly based on simple heuristics or local line/spacing minor adjustments. It is a challenge to provide fixing guidelines based on the perspective of manufacturing process; that is, how the fixing operations may actually affect the layout object and its context region.