Voices: Freescale’s Lisa Su: embedding future growth
By Suzanne Deffree, Managing Editor, News - March 18, 2008
EDN recentlyspoke with Lisa Su, senior VP and chief technology officer at Freescale Semiconductor, about the future of embedded technologies, differentiation beyond Moore’s Law, and the challenges that come with an increasingly sophisticated electronics market. Su also discusses her career path, one that has taken her from MIT to IBM and now to Freescale and has made her one of the highest ranking women in technology. What follows are excerpts of that conversation.
EDN: What initially got you interested in math and science?
Su: My father was a math major and so when I was a kid we did a lot of geeky type of things. It was something that I just ended up being good at, I think because it was more tangible. It was easier to see cause and effect in math and science than in some of the other areas.
EDN: What has a life full of math and science done for you?
Su: What I like, and particularly where I’ve spent much of my time in engineering, is being able to touch something that you are working on. That’s the really exciting part of math, science and engineering – you can see products ship out the door and say that the stuff that you are doing is in the next Microsoft Zune or the next BMW car.
EDN: As a doctoral candidate at MIT you choose to focus on SOI, which, at the time, was a relatively uncharted field. Why?
Su: It was one of those areas where there wasn’t a whole lot of work that had been done yet. It was a new and emerging technology. What I found interesting – and what I think is interesting about any PhD – is that the topic is important but it’s really a good feeling to be able to be the expert in a field for at least a certain time. At that given time, the goal was to go very deep. Most graduate work, I think, is not about the specifics but more about the problem solving skills and how do you get your arms around the subject. And that particular subject was very interesting.
EDN: Who’s your mentor?
Su: My PhD thesis advisor [Dimitri Antoniadis] was particularly important mentor and still is after so many years. He always gave me good advice on how to think about problems and on business situations. I’ve had numerous mentors throughout my career, both technical and executives, that I feel are very strong role models.
EDN: What’s the best piece of career advice you’ve ever received?
Su: When I was getting out of school, my advisor said, ‘stay technical as long as you possibly can because once you stop it’s hard to go back. That’s particularly [true] for someone like me. I’m interested in a lot of different things. Probably the other piece of advice, [applicable to] any business situation or in day to day life, is to always run toward the problems because there really is an opportunity to do something special. Problems allow us to be better than we normally are.
EDN: You came to Freescale in June 2007 from IBM, where you served as VP of the semiconductor research and development center and were responsible for the strategic direction of IBM’s silicon technologies, joint development alliances, and semiconductor R&D operations. What was behind your decision to transition from IBM to Freescale?
Su: It was one of those tough decisions. I really enjoyed my time at IBM. I had an opportunity to do and learn a lot of different things. I had a view of ‘what I want to be when I grow up’ and semiconductors is very much where I want to be and where I had spent much of my time. I wanted to be at a smaller company that was a little more focused on semiconductors and Freescale gave me that opportunity. But it’s about continuing to learn new things and I think there are still a lot of things to learn in the industry. I’ve been at Freescale for seven or eight months and Freescale is very much a company in transition from the standpoint that it is trying to extend in the markets it is in today but also trying to grow in some of the new markets for us.
EDN: What tech areas hold future growth for Freescale?
Su: Our core markets are in networking, automotive, and wireless. Those are large, good markets where we have good positions. What we are starting to see, of course, is that the consumer market is a significant growth opportunity. The applications of some of the technology we have built for the networking and automotive are making their way in the consumer area, especially with multimedia. Our [goal] is to really put embedded electronics in every application. The everywhere, everyplace, every time. So there are lots of opportunities.
EDN: What challenges are you facing in doing that?
Su: Probably the biggest challenge that I see for the industry is as electronics are getting more sophisticated. It’s upon us as an industry to make [electronic devices] a lot simpler to use. It’s great to have all this technology, but if we can’t get that into applications -- if we can’t get that hooked up and connected and working together -- it’s not as useful to the overall goals. We are spending a lot of time not just on the chip itself, but on the solution and the package around the chip so that we can get it very quickly into new applications, particularly when you are doing things like making chips that will go into next-generation automotive electronics. The electronics right now are still at the high end of some of the more mechanical areas, but we can see it becoming a lot smarter in the sense that they are going into active mode. You can imagine collision detection control, things like that, require a lot different role of integration in the application.
EDN: Automotive is a highly demanding and time-intensive market. How difficult is the automotive market to navigate from the technology and business standpoints?
Su: It is a challenging market from the standpoint that it takes a long time to design in. On the other hand, it is also an opportunity. It’s really a place where we can see that there will be a lot of growth. Even if the number of cars themselves grow at a certain rate, the amount of electronics as a percentage of the car is going up. Probably the largest challenge is to ensure that the quality levels are where they need to be, the idea that there should be zero defects. We can’t afford to have a quality incident.
EDN: What about growth in wireless? Freescale spun out of Motorola in 2004. Motorola stayed as a Freescale wireless customer, but has seen its mobile devices business significantly decline since the spinout.
Su: I think the wireless market is a good market. It continues to be a place where we are working on diversifying our customer portfolio. We continue to work with Motorola, but we are also looking at engaging with other large wireless customers.
EDN: Where do you see wireless technology going?
Su: I think what we find is that certainly the adoption rates of 3G are there. There are a number of standards across the different geographies. Being able to keep in pace with those standards is an important thing. The technology is going to be there. I’m not worried about the technology, per say. It’s really the adoption rate of the infrastructure and ensuring that the competing standards get worked out.
EDN: How important is Linux to wireless growth?
Su: Linux is extremely important, period. It is a great opportunity to really get application developers on a common operating system. It’s important in both wireless and other consumer, multimedia applications.
EDN: One thing I’ve heard from Freescale now several times is ‘moving beyond Moore.’ Can you go into your ‘more than Moore’ strategy?
Su: When you take a look at where you can differentiate in semiconductors, it used to be that scaling was king and companies would spend most of their time ensuring that you could get to smaller, faster, cheaper every two years. What we are finding in our space is that scaling is important – and when I say scaling I mean all the aspects of migrating a chip from generation to generation – however, there are a lot of other ways to differentiate. It’s about the software, the system that it comes in, we're differentiating in packaging, for example. Packaging is sometimes an after thought, but I view it as a place where there are new opportunities to differentiate. ‘More than Moore’ for us is that we will continue pushing advanced technology from a scaling standpoint, but that we are looking for other ways to differentiate and make our products more valuable to the consumer.
EDN: How have you been changing up your R&D strategy and other investments to reflect this?
Su: We have been shifting our R&D strategy from a process-technology centric to more of a solution-centric type of view. On process development, we are doing much more in collaboration with other companies. For instance, we are part of the IBM alliance and that has worked out really well. However, we are investing a lot more in things like software, analog, system solutions, those types of things to really round out our portfolio.
EDN: Can you speak to Freescale’s packaging technology roadmap?
Su: As we make chips smaller and smaller, one of the issues is really when you get limited by the number of connections you need to make, the wire bond pitches. So we developed a new technology that we call RCP, or redistributed chip packaging. The idea of that is to take some of the wafer-level elements, such as lithography and plating, and really build wafer-scale packaging. That’s one of our most innovative technologies and it will reduce package sizes by 30%. It’ll certainly reduce chip sizes as far as what you are able to do in terms of advanced technologies. That’s a technology to watch. It’s going to be very important, I believe, to get much smaller form factors for these devices.
EDN: What’s your ETA for RCP?
Su: I would say 2009.
EDN: How about Freescale’s networking roadmap?
Su: On the networking side, another really important technology for us is multicore. Multicores have been in PCs for a while, but they haven’t broken into the embedded space in a big way. It is one of the areas that we believe there is still a lot of innovation yet to happen. The key is that you have this network infrastructure that is already in place and you need to be able to migrate it to the higher-performance, lower-power design points in a way that’s really very easy to do. Our networking roadmap is aggressively going after multicore in 45-nm technology.
EDN: Any ETA?
Su: Sampling in the second half of this year and in production after that.
EDN: One final question, Lisa. You seem to refuse to make a choice between being a researcher and a businesswoman. Why and how do you balance those two worlds?
Su: I like that question. I believe that technology is really, really exciting, but only when it gets out into the marketplace and when you can see it in a product and an application. The best technologist understand what’s going on in the market and in the business environment. That’s the way I view it. I happen to be CTO of Freescale, but I’m very interested in ensuring that I understand personally what our customers are thinking and where the market is going. That’s what will make us more successful. It’s also one of those things that I think can differentiate technologies.