UBC Reports | Vol.
51 | No. 8 |
Aug. 4, 2005
Okanagan Perfect Fit for Microchip Engineer
By Bud Mortenson
Imagine more than 15 million transistors packed onto a wafer the size of a thumbnail. Andrew Labun can tell you precisely how warm one of those transistors is. And that’s pretty cool.
An avid science fiction reader as a kid, Manitoba-born Labun saw engineering as a way to explore new technologies -- to build things and make some of the fiction a science reality. He followed his interests and earned an engineering degree from UBC, and a PhD in electrical engineering from the University of Alberta while working on high-power gas lasers big enough to fill a transport trailer. It was fitting research for a sci-fi fan.
Chemically reacting plasmas involved in laser research are similar to those in which semiconductor integrated circuits are manufactured. That parallel took Labun light years away from the laser lab.
For the past 13 years, he has been in the U.S. delving into the inner world of microelectronics -- bits of silicon and other matter organized into electrical gateways and conducting structures.
Labun’s area of specialization is in modeling electronic structures that make up the microscopic circuits of computer chips. There’s irony in the way he holds his thumb and forefinger slightly apart and explains, “These structures are as big as a micron across.” Not very big at all -- line up 10,000 of them and they’d stretch out for a centimetre.
After working in Massachusetts as a senior engineer for Intel, Compaq and Digital Equipment Corporation, Labun wanted to bring his family to Canada, and he wanted an opportunity to share his specialized knowledge of microprocessor modeling.
It’s the kind of extensive and rare industrial experience to which UBC Okanagan students will have access when Labun begins teaching as associate professor of engineering this September.
“I want to convey some of the excitement of engineering to the students,” he says. “And I plan to work on research in modeling technology -- there’s no doubt the world is full of opportunity for this kind of research.”
Using virtual models and sophisticated computer-aided design tools he developed, Labun can predict exactly where circuits get hot, how much power they’ll use, and their capacitance -- how the tiny wires interact electrically with each other. All without having to build a chip. In fact, he says, there’s no sense building a chip to get these readings. “These chips are so complex, measurement is all but impossible,” he says. “Modeling is the only way.”
His research is changing how computer chips are modeled. A common technique called Finite Element Modeling has its place, but tools Labun has developed are faster and handle big chip designs better. The latest of his published techniques is called Chip-Level Intertwined Metal and Active Temperature Estimator, or CLIMATE. Using this new technique, “what would take four hours to simulate with a finite element model -- still a very useful tool -- takes just a fraction of a second. Of course, multiply the sample by a billion and it takes a bit more time.”
Despite a career immersed in microprocessor technology, Labun says he is intrigued not by the electronics but by the physics and other science that goes into the technology.
“I always hated electronics,” he says, grinning a little. It’s a curious disclosure from someone who happily shows off examples of tiny processors he “and a team of thousands” helped create.
Souvenirs from painstaking development projects, each chip is encased in plastic and each held distinction as the world’s fastest microprocessor at one time or another -- some for several years. Holding in the palm of his hand four generations of these chips, dubbed “Alpha” by maker Digital Equipment, Labun details the magnificent science behind each.
The Alpha EV6 chip, for example, was the first to handle out-of-order instruction processing. “This thing could predict what a more efficient sequence would be -- it’s fiendishly complex microprocessor technology.”
The semiconductor industry is well known for jealously guarding its secrets. Labun came to realize that the sharing of knowledge -- and being able to pursue interesting research not strictly tied to a corporate mandate -- was very important to him.
“Teaching is a great opportunity, and Kelowna is a good fit for me,” Labun says, explaining that he chose UBC Okanagan because he has family in the community, and because he saw a chance to put his experience and interests to work.
“I like to talk to people about the science,” he says. “The academic environment is ideal for that. I see lots of scope for collaborative research with other academics and industry partners.”