Name: Dennis Georg Age: 46 Place of Birth: Algona, Iowa Cars He Drives: Ford Bronco II and Acura Legend Current Position: Director, Computer Systems Lab, Hewlett-Packard Laboratories Time in Current Position: 1 month Years in the Industry: 24 Favorite Non-work Activities: skiing, hiking, traveling, reading
Pet Open Systems Peeve: "The problem is that the cost of ownership of any computer system is largely dominated by the post-purchase cost-the software, service, maintenance, and upgrades. And among the dominant costs in any computer system, open or not, is the labor part of that. So if I have a pet peeve, it's the slowness with which the industry is responding to addressing the cost of maintenance and operation. I think there's a huge opportunity there. HP has good products, but I think there's a huge difference between where we are and where we'll need to be in the future."
On UniForum: "I've always been impressed, having attended UniForum [conferences] in the past, with the level of informed presentation and informed discussions and panels. A lot of things are going to go on in parallel [at UniForum '95] and you're going to wish you could be at most of them. These are all huge topics, and they're all extremely important."
As a panelist in one of UniForum '95's featured plenary sessions, "Experts Predict the Future of Open Systems," Hewlett-Packard's computing visionary Dennis Georg will be asked some provocative questions about developments that could dominate the industry in the next century.
As director of the computer systems laboratory of one of the world's largest computer manufacturers, Georg should be able to answer those questions. After all, it's his job to know.
Georg is director of the Computer Systems Lab of Hewlett-Packard Laboratories, in Palo Alto, CA, the branch of the company responsible for developing the technologies to be used by HP in its product line beyond the next five years. Each HP business unit has a research and development team that focuses on applying new technologies in a three-to-five-year time span. HP Laboratories consistently looks even farther into the future, aiming to create the basis for new products that will please current customers and bring new ones into the fold.
"We're looking over a broad range of technologies-all the way from how systems might be put together in the future, to imaging and visualization, to compiler optimizations and high-performance systems design," Georg notes. Other areas being worked on are packaging and ultra-large-scale integration (ULSI) technology for microprocessors. "It's a broad set of technologies," he says. "The challenge is that there's no way to compute the future. It's important to be able to see shifts in requirements-where the unmet needs might be today that might give you some prediction of where to go." In addition, Georg needs to anticipate what customers' unmet needs will be three or four years from now, which is more of a guessing game. "You need a good combination of intuition and analysis," he says. "Right now it's a little bit mind-numbing."
Here's a snapshot of some areas of technological breakthrough Georg expects:
After graduation in 1971, he went to work for Procter and Gamble, as part of a small team of people working on computing applications. He worked on several programs, applying computers and computer algorithms in the engineering of new technologies and facilities for factories that Procter and Gamble was developing. "That experience was a tremendous benefit to me in my later career because I had my first taste of real industrial application of computing," Georg says. "It was my first taste of actually applying database technology in a real, practical experience. It was my first experience in hacking on operating systems."
After slightly over a year, he returned to Iowa State, where he "went back to pursuing computer science with a vengeance," eventually earning his master's and Ph.D. He pursued several areas, including numerical mathematics, the development and theory of operating systems, and analysis of mechanisms and policies in scheduling and security. He also got his first solid introduction to UNIX. "What was exciting to me was that you could actually get your hands on it, you could run experiments with it. You could construct different policy alternatives and invest in trying them out."
While a graduate student, Georg became a research associate at Ames Laboratory, then part of the Atomic Energy Commission. "I had research support throughout my entire graduate program," he says. "I also had some really outstanding professors. The early '70s was a time of explosion in quite a few technologies. It motivated me to take on some of the really hard operating systems problems. Getting support and enthusiasm can be infectious."
Subsequently, Georg was part of a team that helped define HP's RISC architecture, called Spectrum. He started HP's RISC microprocessor design team and from 1983 to 1990 he had overall responsibility for developing HP's family of RISC microprocessors. "Back in 1983, I was absolutely convinced that a couple of things were going to be driving the industry," Georg says. "One was microprocessor technology in general, and the second thing was providing solutions on systems. The customers really wanted to buy solutions, so the availability of solutions and portfolios was going to be very important.
"I saw that RISC would rapidly overtake CISC in performance. And the compilers were going to be there to take advantage of it. You were always going to be able to get better, more closely coupled code for an application using a compiler than you were using generic application code that might be put into firmware in a CISC implementation. RISC was trying to provide the facilities that would allow the needs of a particular application to be met at lower cost. And there's the hardware advantage, that you could fundamentally implement a RISC processor using less aggressive technology and get more performance in instructions per second."
In 1990, Georg was asked to be program manager for HP's RISC workstation program, which was to incorporate the company's PA-RISC microprocessor, switching from Motorola 68000 chips, and a new release of the HP-UX operating system. That workstation family became the series 700, which is still a major part of the HP workstation line.
Following that job, Georg held a number of other assignments in HP's workstation business, until joining HP Laboratories 16 months ago as a senior technical contributor. He was appointed to his current position in December.
"One of the things that will happen in the future is that the world will become more heterogeneous. By that, I mean that it's probably unlikely there will be any macrocomputer system that won't contain open systems computers. In other words, there will be much more of a mix and match in terms of computers-proprietary special purpose computers, general purpose computers, and open systems. That may be the hook that lets people know the values they can attain with these open systems, and it may make it palatable for them to gradually move off their proprietary solutions."
The plenary in which Georg will be heard is scheduled for Wednesday, March 15, at noon. According to session chair Alan Paller, director of open systems at Computer Associates International, "We want to open up people's minds to ways of computing that they hadn't thought about, so they will go out of the room saying, 'I hadn't thought about that, but I'm going to explore it because it might be a future for me."