Dec. 14, 2012 – The U.S. Department of Defense has awarded University of Utah computer scientist Mary Hall a $1.6 million grant to develop computer software that could dramatically improve how much information can be collected and processed by robotic vehicles and airborne drones.
This grant is part of a $20 million, five-year award to graphics processing company NVIDIA from the U.S. Defense Advanced Research Project Agency, called DARPA, which is the research and development arm for the Defense Department. The University of Utah is a subcontractor to NVIDIA.
This funding is for the exploratory phase of a DARPA program that funds research for processors that are 75 times more energy efficient than current software and hardware systems. The goal of this research is to enable surveillance and computer vision systems in ground and airborne vehicles that collect and analyze vastly more data than can be processed today.
“Our team will develop an ‘autotuning’ software framework, which enables a programmer to write software that improves its performance and energy usage while the software is running,” says Hall, an associate professor in the U’s School of Computing. “We have been working on this idea for seven years, and we have shown it to be very effective. This grant allows us to migrate these concepts into common practice, and partnering with NVIDIA provides a path for this migration.”
The Utah team will participate in computer programming systems design by developing programming techniques that enable skilled programmers to specify and control intensive computations. These models will be designed together with the hardware developed by NVIDIA.