Amanda Bienz, an assistant professor in the Department of Computer Science at The University of New Mexico, has received a National Science Foundation CAREER Award for a project that seeks to ensure that the performance of parallel computer applications matches the performance of the hardware.
“Towards Exascale Performance of Parallel Applications” is a five-year, $557,000 project that is funded Feb. 1, 2024, through Jan. 31, 2029.
The project, which will fund one graduate student and one undergraduate student, addresses the reality that each generation of supercomputers is more powerful than the previous version, with computing speeds for massive numbers of operations that can be completed in the blink of an eye. Such supercomputers are used for processing complex data needed for computer simulations for everything ranging from weather modeling to cancer treatment.
But as powerful as the hardware on these high-performance computers is, for them to work to maximum capacity, the applications need to match that speed and performance, which has long been a problem. Bienz is tackling that with this project.
“The hardware performance is there, but the applications cannot currently take full advantage of the hardware power,” she said.
Bienz said the inspiration for this project came naturally since this gap in performance between hardware and applications is a common frustration.
“I work with many application developers, and many of the programmers cite this issue as one that needs addressing,” she said.
Although the performance level of hardware will continue to improve as technology develops, Bienz said the goal with this project is to reduce the performance gap between the applications and the hardware.
“The main idea is to change the performance level of the applications so that when hardware advances in the future, the applications can keep up,” she said. “Data is now moving to GPUs [graphics processing units] instead of CPUs [central processing units]. We need to understand why we get the performance we get, and why the applications do not match the expected performance.”
In addition to the research, there is also an outreach component to the project, which will involve reviving the CS4ALL course (a program that teaches coding and computational thinking to students and teachers, previously funded at UNM by the NSF), as well as developing a hackathon that introduces computing topics to a diverse student population across UNM’s main and branch campuses.
Bienz, who joined the UNM faculty in fall 2020 from the University of Illinois at Urbana-Champaign. The Indiana native said that she had an unusually early interest in computer science, thanks to an aunt who was a programmer.
“I wanted to be a computer programmer since fourth grade,” Bienz said.
The CAREER Award is NSF’s most prestigious award in support of junior faculty who exemplify the role of scholar-teacher, through outstanding research, excellent education, and the integration of research and education. She said the award is a major boost for her academic career, the effects of which will be felt for many years to come.
“It will help build a foundation for future research,” she said.
The project is jointly funded by the NSF's Software and Hardware Foundations Program and the Established Program to Stimulate Competitive Research (EPSCoR).