BIG DATA
Balancing computing power, storage demands goal of Virginia Tech CAREER project
Ali R. Butt, an assistant professor of computer science in Virginia Tech’s College of Engineering, has received a $400,000 Faculty Early Career Development (CAREER) Award, which is the National Science Foundation’s most prestigious award for creative junior faculty who are considered to be future leaders in their academic fields. The goal of Butt’s CAREER research is to address the increasing performance gap between computing power and storage technology, especially for high-performance computing (HPC) environments. 
Modern scientific computations often require analysis of information from a large number of devices, such as measurements from temperature and humidity sensors distributed across a region for monitoring the climate and forecasting environmental impacts. These complex applications require powerful computing resources and entail managing an ever-growing amount of data.
“In terms of high-performance computing power, we are seeing systems with tens-of-thousands or more processors that reach terabyte speeds,” Butt said. Terabyte speed is the processing of one trillion instructions per second.
“Soon computing systems will have the capability of processing a petabyte—or 1,000 terabytes—per second,” he noted. “Unfortunately, this is not matched by a corresponding improvement in the input/output systems that enable the processors to interact with external devices, such as disks and networks.”
Employing a Scalable Hierarchical Framework, which is designed to provide reliable high-performance storage, Butt is attempting to develop a data storage framework attuned to the ever-increasing demands of modern HPC environments. If he is successful, this framework will bridge the gap between modern computing power and storage, and will support efficient and reliable data management.
“The CAREER grant will enable me to engage quality graduate students and together we will design, develop, and deploy tools and systems for improving the input/output performance of modern HPC systems,” Butt said. “We will adopt a holistic approach—all system components should interact to provide an efficient HPC data management system.”
In addition to helping Butt realize his research goals, the CAREER award “is also is a symbol of the trust that my peers have in this project,” he said. “These are exciting times in the field of HPC research.”
Butt is director of Virginia Tech’s Distributed Systems & Storage Laboratory, where research is focused on tailoring next-generation storage and file systems for the growing data demands of modern high-performance computing applications.
Before joining the Virginia Tech faculty in 2006, Butt completed his Ph.D. in electrical and computer engineering at Purdue University. He received his bachelor’s degree in electrical engineering from the University of Engineering and Technology in Lahore, Pakistan.
Modern scientific computations often require analysis of information from a large number of devices, such as measurements from temperature and humidity sensors distributed across a region for monitoring the climate and forecasting environmental impacts. These complex applications require powerful computing resources and entail managing an ever-growing amount of data.
“In terms of high-performance computing power, we are seeing systems with tens-of-thousands or more processors that reach terabyte speeds,” Butt said. Terabyte speed is the processing of one trillion instructions per second.
“Soon computing systems will have the capability of processing a petabyte—or 1,000 terabytes—per second,” he noted. “Unfortunately, this is not matched by a corresponding improvement in the input/output systems that enable the processors to interact with external devices, such as disks and networks.”
Employing a Scalable Hierarchical Framework, which is designed to provide reliable high-performance storage, Butt is attempting to develop a data storage framework attuned to the ever-increasing demands of modern HPC environments. If he is successful, this framework will bridge the gap between modern computing power and storage, and will support efficient and reliable data management.
