For years Todd Mowry had planned on becoming a doctor, but five weeks at Carnegie Mellon's Governor's School for the Sciences, a summer program for gifted high school students, abruptly changed his career path. He got his first taste of working with computers that summer of 1983, and he hasn't stopped since.
Mowry, 32, associate professor in the Computer Science Department, was recently named a Sloan Research Fellow for his groundbreaking work in computer architectures. The award from the Alfred P. Sloan Foundation recognizes Mowry as one of 100 outstanding young scientists and economists working at the frontiers of computer science, physics, mathematics, chemistry, economics and neuroscience.
The Sloan Foundation selects award winners each year on the basis of "their exceptional promise to contribute to the advancement of knowledge." Recipients receive a two-year, $35,000 research grant.
With his expertise in computer architecture and compiler optimization, Mowry is working to boost the performance of microprocessor-based systems through various forms of parallel processing. He heads the STAMPede project, an effort to harness the full performance potential of integrating multiple processors onto a single chip.
"Todd is extremely sophisticated in the arts of compiler design and system simulation," said Jim Morris, head of the Computer Science Department. "He was the key architect of the MIPS R10000, one of the industry's best processor chips. Currently, he is working on the opportunity of exploiting chips with a billion transistors. To put these transistors to good use, he has developed an architecture for multi-processing that provides a platform for the execution of many dependent threads. His facility with compiler technology allowed him to make some very clever delegations of responsibility between the compiler and processor tradeoffs that most other researchers don't make," Morris explained.
Mowry, a native of nearby Butler, Pennsylvania, said he had never touched a computer prior to attending the Governor's School. He recalls a tremendous intellectual letdown when he returned to high school after attending the summer program. He tried to keep the momentum going by working on a computer in the physics lab.
"I spent at least 80 percent of my senior year writing programs on that computer, trying to recreate some of the exciting computer research environment that I had experienced here at Carnegie Mellon," he said.
Mowry received a bachelor's degree in electrical engineering from the University of Virginia in 1988, and his master's (1989) and doctor's (1994) degrees in electrical engineering from Stanford University.
As a doctoral student, he participated in several high-profile multiprocessor and compiler projects. He also worked part-time in the architecture group at a startup company called MIPS, where he worked on projects including the R10000 processor. He came away from that experience at MIPS, which was subsequently bought and sold by Silicon Graphics, Inc., with two patents and a lot of insight about shaping his future.
Upon entering the doctoral program at Stanford, Mowry initially thought he'd head into industry, but he soon realized it was his research rather than his work in industry that was most satisfying.
In 1994, he became an assistant professor in the Electrical and Computer Engineering Department at the University of Toronto. In 1997, he joined the faculty at Carnegie Mellon.
"In industry there is a small amount of time at the start of a project when you can think about wide-open research issues, but most of the time is actually spent just making sure the next project works," Mowry explained.
"I decided that I got more enjoyment out of looking further down the road five to 10 years ahead rather than at the 12-18 month horizon in industry. I wanted to change the entire computer industry and not just the fortunes of one company."
Mowry also learned that he likes to teach and enjoys working with graduate students.
"I believe it is completely natural for teaching and research to go hand in hand," he said. "We need to learn in research how to explain our ideas clearly to other people."