From his work compiling an annual list of the world’s fastest computers to his research expanding the boundaries of supercomputing, Professor Jack Dongarra is an icon of the supercomputing world.
Now, in recognition of those efforts, Dongarra—who directs UT’s Innovative Computing Laboratory—has picked up a pair of vastly different but equally impressive awards.
The Association for Computer Machinery (ACM) recently honored Dongarra with the High Performance Parallel and Distributed Computing Achievement Award at the annual High Performance and Distributed Computing Conference in Kyoto, Japan, while the Institute of Electrical and Electronics Engineers (IEEE) will bestow him with the Super Computing (SC) 2016 Test of Time Award at its conference in November.
“Both of these honors mean a great deal to me personally, but more than that they serve as testament that our group does things the right way and with great impact,” said Dongarra, Distinguished Professor of Computer Science in the College of Engineering and Distinguished Research Staff Member at Oak Ridge National Laboratory.
“This sort of recognition reflects well upon our efforts as a whole, not just the work I do.”
For the ACM award, the recognition came for research done advancing parallel computing—supercomputing in which nearly endless calculations are done simultaneously.
Dongarra’s expertise in that form of computing dates back decades and includes his work as editor of the groundbreaking Sourcebook of Parallel Computing in 2003.
The award also serves to recognize his contributions to concepts dealing with the development, innovation, design, and testing of new devices.
“While the award itself is an honor, it also calls attention to the impact a long-term commitment to experimental computer science work can have on scientific discovery,” he said. “That impact affects everything from economics to high-energy physics, from human health to geology.”
As the name implies, the SC award goes to innovation that has stood the test of time—in this case, a paper Dongarra authored in 2000 with Clint Whaley titled Automatically Tuned Linear Algebra Software (ATLAS).
ATLAS is a particular form of technology known as autotuning, a program that can optimize itself to the highest levels of performance for different computer hardware.
A major step up from preceding programs because of the way it self-optimizes, ATLAS continues to be cited and relied upon as a baseline measurement in the scientific community almost two decades after its release—an eternity in the world of computing.
—
CONTACT:
David Goddard (865-974-0683, david.goddard@utk.edu)