Richard DeMillo is the John P. Imlay Jr. dean of the College of Computing. In 1972, he earned his PhD in computer science at Georgia Tech, where he was founding director of the Software Test and Evaluation Center, and taught for 11 years before starting a career in industry and government. When he was named dean in 2002, DeMillo was chief technology officer for Hewlett-Packard.
What will the Christopher W. Klaus Advanced Computing Building mean to the college?
The Klaus building is a gateway to the new center of campus. It makes a statement about new technologies that have become important at Georgia Tech and the building reflects what Christopher Klaus is all about. He is an entrepreneur. This is an entrepreneurial building. He is a collaborator. This is a building that supports collaboration.
Georgia Tech is turning traditional computer science education on its head. How?
Think about what has happened in computing over the last 20 years. Computing used to be about the computer. It used to be about the technology. What the computer was going to be used for and the applications were almost afterthoughts. We're turning that model around. Computing is valuable because of the application. We lead with that. Universities are going to play an increasingly more important role in figuring out what the novel applications are — how the computer impacts the real world. We think of it as horizontal education.
The college refers to the "new face of computing." What is that?
The new face of computing is about how computing is going to be used, how it is going to change the world. It's also about the new people that are being drawn to computing as a profession. These are students who would not have considered computing as a major two years ago because they thought that the field was not interesting.
You've revamped the curriculum and introduced "threads." What are threads?
I came here as dean in 2002, just after the dot-com bubble burst. I realized that computer science departments should be doing more to provide the why of computing and not only the how. Around that concept, we built the idea of threads. The whole purpose of a thread is that it doesn't limit you to the idea of a core curriculum. It doesn't limit you to the idea of electives. It allows students to pursue courses of studies that really interest them.
How should American universities respond to the loss of computer science jobs to India and China?
American universities should respond the way Georgia Tech is responding, which is to not continue to educate computer science students in the areas that are being commoditized. When Thomas Friedman talked about the threads program in his book "The World Is Flat," he made the point that you're not going to be able to outsource a great storyteller, a great author, a great poet or a great designer. We want to give the students the technical base so that they can perform at the right level and then give them a set of skills so that they will continue to add value to their employers. Those jobs will never be outsourced.
What is the concept behind the Robotics and Intelligent Machines Center established by the College of Computing and College of Engineering?
It is really part of the new face of computing. We've always had a great robotics group here at Georgia Tech, not just in the College of Computing but also in mechanical engineering and electrical engineering — across the campus really. We decided to get serious about putting all of that expertise together in an interdisciplinary program in robotics and we began attracting international attention. KUKA Robotics, the large German industrial robotics firm, endowed a chair in robotics.
Then we started looking at a PhD in robotics and we attracted senior faculty members who we thought could lead new efforts. At the same time, people like Bill Gates were thinking about how do you educate people to program in this new environment that we find ourselves in. We were hiring some of the best roboticists in the world. Microsoft wanted to fund an activity that was aimed at putting all of this stuff together and the robotics institute came out of it. Microsoft awarded Tech $1 million to bring robotics technology into the computer science curriculum.
Will all students have personal robots in the future?
Yes. We will send students to the bookstore to get their Microsoft-Georgia Tech robot package, which has a little robot in it that connects to their computers. The students will learn how to program by programming these little mobile robots. Robots are intrinsically fun to program. But there is also a deeper purpose. The robotics marketplace is undergoing a profound change. Robots are getting much smaller, much cheaper and they are being used in a variety of environments. In the not-too-distant future the worldwide market leaders in robotics will have not 60,000 installed robots but 600 million. That's a huge change. It is going to make new markets for robotics. It's going to mean there is a group of professionals needed to program those robots to run the systems, to develop the applications, to build the industries around them. We want to see Georgia Tech at the vanguard of that.
What have been the breakthroughs in supercomputing?
Supercomputing is an area that advances like a braid. You make an advance in hardware, you make an advance in software, then you make an advance in application, you make another advance in hardware and then you make another advance in software. Georgia Tech has really been at the foreground at getting this flywheel moving to get hardware, software and applications working together.
What key challenges are ahead?
Part of the challenge is that it is not very cheap to get into this game. The performance of the supercomputers that we are talking about is measured in a unit called FLOPS — floating point operations per second. In order to be a player in the supercomputer game, you have to have a computer that operates at teraFLOPS. A teraFLOPS is a trillion operations per second. We want 100 teraFLOP machines. So far Georgia Tech has bought and deployed 20 teraFLOPS in the last 18 months. We have become one of the top 50 supercomputer sites in the world. As a result, we have become among the top 10 university supercomputer sites. We have our sights set on a petaFLOPS computer — a thousand teraFLOPS. We have a real chance to move our scientific agenda, our technology agenda forward simply because we have unfettered access to these kinds of machines.