Q: What is your area of expertise?
A: My area of expertise is in the signal integrity area, which is one in which we worry about how clean the signal is as it travels through electronic equipment – whether it’s computer or communications circuits. We do that for specific types of applications.
In the analog world, [signal integrity] has to do with high-frequency signals. The higher the frequency, the more information you can put in the signal. The digital information has to do with sending bits of ones and zeros. The problem is that when you work at those rates — high speed or high frequency — it turns out that the signal doesn’t behave the same way. The signals become more immune to noise. A lot of bad things can happen when you operate at high frequencies. As a result of that, the signal gets degraded. One example is when you think of DSL. The company does a test to see if the distance between the service provider and the home is such that allows the signal to have integrity, as you would say. Then the company would state your location is OK, and they can provide you with DSL service. They send signals at a very high rate and they look at how these signals look at your home.
I worry primarily on the channels, though. We model these channels to explain how the signal degradation process takes place. It turns out that a lot of our efforts involve coming up with algorithms and models that describe how the signal could be degraded; and that’s an important part of the design today because people using signals at high frequency or high speeds must use computer-aided designs that would be able to predict whether the signal will have to go through a serious level of degradation before it gets to its destination.
Anything that deals with a signal, you have to worry about the signal integrity – wireless communications, cell phones. People did not have to worry about this in the old days because frequencies were very low and the speed at which the ones and zeros were being transmitted, was very slow. Signal integrity is not a problem when you’re dealing with low frequencies and low bit-rate. It becomes a problem when you’re sending things very fast.
Q: Give me a brief synopsis of your education and career.
A: I spent two years in Florida after leaving Haiti and then completed my undergrad at MIT. I attended the University of Miami in Coral Gables for the first year – it was closer to home. After MIT I spent two years in industry at Hewlett-Packard in California and then I came here to get my master’s and PhD. My advisor was involved in doing some of the first University work on packaging and signal integrity. Upon my graduation it turned out that it was a strategic area in which the department didn’t have any faculty. So that’s how I got hired.
In 2001 I took a leave of absence to start a company. I was a founder and the first president of Xindium Technologies. Two of my co-founders are faculty in the department: Professor Milton Feng and Engineering Dean Ilesanmi Adesida. We ran Xindium from 2001 to 2004. The first purpose was to rely on the technologies that Milton and Ade had developed – devices that used indium phosphide. My expertise was the packaging and measuring. In the spring of 2004 I had been out of the University for three years, so it was time for me to come back. The company continued on and moved to Chicago. As of 2006 the company was sold to another company in California. So we are no longer involved with it. That was one of the major accomplishments we made, though. It didn’t turn out to be what we wanted, but we learned a lot and Xindium did bring a lot of money to the University.
Q: Tell me about an accomplishment you’re proud of.
A: What I feel good about as a professor here at Illinois is that I see how useful I am. I like when the students come to me for help. They have a desire to learn, and then the fact that I help them go through their graduate studies – that’s what makes me feel very good. I feel like I’m helping people better themselves and get an education and build a career. I would say a lot of the achievements I have actually are done through my students. Looking at them go out into industry and continue to develop technologies, that makes me happy.
Q: What do you enjoy most about teaching?
A: You can see that the students are extremely motivated. I was discussing in one of my special sections with the TAs that all the students did well. So I’m going to have to give a lot of A’s this semester. They all did very well, they all pushed themselves. And I like that. When I talk, everyone pays attention and they listen to you. That’s something that I find rewarding. The way we select our students, and especially our graduate students, makes it such that we do end up with the high quality of students who do well.
Q: What role do students play in your research and what do you look for in a potential student?
A: Some of the [power distribution network] tools that are currently being used, commercial tools and their methods were developed by me and my students. And these tools are being used all over the world. A transmission line simulator, called the W-element, was developed by one of my PhD students. That’s been very successful. A lot of the students I recruit, I don’t recruit from the mainstream. For instance, some of them came just accompanying their spouse, I had one come as a refugee, a family member recruiting them to me. What I look for is the attitude. I’ve never had a problem with my graduate students. I’m very careful about how I pick them.
Graduate students are essential. If there’s a reason why I’m still here, it’s because of the quality of the graduate students. I had an offer to work at Georgia Tech in 2001. Atlanta is a much more glamorous city than Champaign, but the one factor that made me stay here was because I realized here I would get better students. To me, it is the primary factor. That’s the most important criterion: the quality of the students.
Q: What are you focused on today?
A: We’ve been working with Cadence Design Systems. They design tools for people who work with circuits. There are going to be quite a few interesting things coming out in the next 18 months in our collaboration with Cadence. There is a set of new tools that we have developed for power distribution networks. Those tools are doing things like micro-modeling and frequency domain simulation, time domain simulation. They will help people design better power distribution networks.
Q: What is the impact of your work?
A: Some of the technologies we have developed are being used by companies and some of them have made it to commercial standards. Having an impact is not something that is easy. Just getting a commercial tool adopted in industry is hard. I’m not going to say technologies we’ve developed are something that will change the world, but they make a difference. It could be big, or it could be modest. It all depends on what happens in the next year in the research that we do and how people receive it in the industry. Some of it involves a completely new way of simulating circuits. It definitely has potential and the fact that Cadence is adopting it is an indication that it has potential.
Q: What technology that’s currently under development are you most excited to see completed?
Right now the thing I’m working on is what I think is the fastest circuit solver. When the designers are designing these circuits, they have to simulate them. The way you design a computer, the way you design a microprocessor, you want the computer to design it and you have to simulate it. You need to predict how the current and voltage are going to behave when the signal goes through it. When you do a computer simulation, it takes a long time. Sometimes when engineers are designing the circuits they can spend a whole week to design the microprocessor because just to simulate it takes a long time. And they cannot afford to wait that long. One of the methods that we have developed can simulate these circuits very quickly. For instance, something that would take an hour, will only take 10 seconds. We are working with Cadence to help commercialize it.