3.22.05 – An academic researcher, educator, innovator, and leader in his field, Tom DeFanti, in his 31-year career at the University of Illinois at Chicago, has helped to create one of the world’s foremost real-time, interactive computer graphics, virtual-reality, and advanced networking environments that extends from the university to national and international borders. Now, as the newest member of Calit2, he will be applying his expertise to advancing the institute’s agenda in large and international-scale visualization, networking, and computing projects.
DeFanti has helped pull together a worldwide community of discipline scientists, computer scientists, artists, computer programmers, and networking engineers who help create the teams, tools, hardware, system software, and human-interface models that enable high-resolution interactive visualizations of extremely large datasets among multi-site, international collaborators over experimental fiber-optic networks. DeFanti is one of the leading architects of high-performance information technology (or “cyberinfrastructure”) for the advancement of science.
DeFanti's current research is empowering scientists to make “10-Gigabit phone calls.” The goal is to provide virtual-reality environments that enable researchers to interact with distant colleagues, data, and computers with the ease and reliability of today’s phone calls. This has been a lifetime pursuit, from developing the first 3-D interactive graphics language for his PhD, to supervising students to develop networked, collaborative, high-resolution, virtual-reality environments. Derivatives of the CAVE, a virtual-reality theater designed by DeFanti and colleague Dan Sandin in 1992, are the dominant modality of virtual-reality delivery worldwide.
Here DeFanti talks with Calit2 about his role, his experience, and where he sees visualization and networking going.
Q: Why were you interested in coming to Calit2?
A: I have worked with Larry Smarr at least half time for the past 19 years, first at the National Center for Supercomputing Applications as his first visiting research scientist in 1986, as associate director for visualization and networking in the 1990s, and then as co-PI of the NCSA Alliance from 1997 on. I spent a sabbatical at Calit2 and SDSC three years ago developing the OptIPuter concept with Larry and Maxine Brown [associate director of the Electronic Visualization Laboratory] and have been visiting San Diego regularly since the OptIPuter grant was awarded. With the new Calit2 building opening soon, the advanced display and networking technology developed with Dan Sandin, Jason Leigh, and Greg Dawe at EVL in Chicago is finding a fertile science, engineering, and art partnership of unprecedented proportions. It was just a matter of time before I rented a beach house in Del Mar and borrowed a BMW motorcycle to get to campus. [Leigh is associate professor of Computer Science and co-director of EVL; Dawe is Calit2 principal development engineer and former design engineer at EVL. –Ed.]
Q: What kinds of visualization and immersive technologies and capabilities do you envision for the new Calit2 building at UCSD?
A: Greg Dawe, Greg Hidley, and I are developing and installing visualization/virtual reality systems in the new Calit2 building at UCSD that better match human visual acuity and perception. We plan to open the building with a stereo HDTV projection system for the auditorium that will allow the audience to see 3-D motion images, a 100+ million pixel display for ultra-high resolution display and collaboration, and two prototype walls for a third-generation, six-wall CAVE virtual reality theater. We are also studying and specifying HDTV portal devices and smart rooms to achieve global telepresence, since no one really wants to leave San Diego . [Hidley is Calit2 chief infrastructure officer.
Q: How did your background prepare you for this?
A: I grew up in New York City and spent all my time doing math, science experiments, and photography. When I went to Ohio State University in 1969, I discovered, to my extreme joy, that I could combine my talents and put them to use in Chuck Csuri's Computer Graphics Research Group. Four years later, at the age of 24, I was an assistant professor at the University of Illinois at Chicago (then called "Circle" campus) and started working with Dan Sandin on real-time, full-color computer graphics systems for use by artists, educators, and scientists. We're still doing this. Along the way, in 1976, California artist Larry Cuba did the computer animation for the first “Star Wars” movie in our lab, I became secretary and then chair of ACM SIGGRAPH, got involved with the first generation of video games, and formed a company to produce software and hardware for video artists. I met Maxine Brown working on SIGGRAPH'77, and, together, we started interacting with Larry in 1986, initially on the first video game show at the Museum of Science and Industry in Chicago, then on visualization, parallel computers, national-scale networking, etc. Based on the success of the I-WAY project at SC'95 which had, as its goal, demonstrating the interoperability of the nation’s research networks, Maxine and I proposed an international hub and connections scheme to the National Science Foundation. Our STAR TAP project, and its successor, StarLight, led international networks in growing thousands-fold from a few 10-megabit links in 1997 to 100s of gigabits in 2005.
Q: Why did you choose to focus on electronic visualization?
A: It held the promise of real-time animation, something obvious now, but, in the 70s, animation was pretty much only stop-action filming and hand-drawn cels. I had a very clear idea of what this would mean to information, education, and entertainment. I went to UIC because there was a huge funded emphasis on electronic media development inspired by the Plato Project from the University of Illinois at Urbana-Champaign. Dan, a physicist who was in the UIC Art department, had invented a video processing computer (the Image Processor) that could color and manipulate graphics generated by my dissertation software system, and we were handsomely supported. The hardware we used in the 70s cost as much in constant dollars as a CAVE or a mid-sized supercomputer does today. We were very lucky young men.
Q: I understand your activities often take you to exotic places. Why is that so?
A: I like to travel. I've been to Japan more than 25 times, Europe at least a dozen times, and I lived in Nancy, France, when I was an undergraduate in 1967-1968. In college, I was a student of several foreign languages before I started writing computer languages.
Visualization surmounts natural language barriers. When I was chair of ACM/SIGGRAPH in the early 1980s, international interest in computer graphics exponentiated. I traveled extensively then and made many lifetime friends. I continued my overseas travels with supercomputing in the late 80s and then networking in the 90s as these too became of great interest to academics worldwide.
I believe that although communication and collaboration are more difficult as the distance and cultural differences increase, they are more compelling to achieve. And, since our international colleagues do not compete for the same pots of money (as national researchers so often do), it's an easy win-win situation to collaborate internationally. The global high-speed networks that NSF has co-funded for the past decade or so have been key enablers of this advanced sharing of international science, engineering, and art.
Q: If you don’t mind speculating, where do you see visualization technology going in the next 5-10 years?
A: We used to call random designs we generated on screens “video wallpaper,” but we will see the ubiquitous wallpapering of our home and workspaces with seamless fabric-thin, ultra-high resolution screens. Most of it will look, as it does now on your desk/lap, like awful strip malls garishly flashing “buy me,” but the potential is there to support wonderful art, science, education, and nature. I expect Calit2 to take the high road here.
Electronic Visualization Lab