Skip to main content Skip to navigation

EECS Colloquium: Error-Bounded Lossy Compression Approach for Scientific Data Reduction in Exascale Computing

Engineering Teaching Research Laboratory, Pullman, WA
ETRL 101
View location in Google Maps

About the event

Abstract: The next generation of supercomputers will be exascale high-performance computing (HPC) systems, which are capable of at least 1018 floating-point operations per second. These systems will help researchers tackle increasingly complex problems (such as nuclear reactors and global climate) through modeling and simulating large-scale systems. Due to the gap between the ever-increasing computation power and the limited storage capacity and I/O bandwidth, HPC researchers have to develop intelligent ways to efficiently manage the scientific data. Error-bounded lossy compression has been considered as a promising solution, because it can significantly reduce the data size while maintaining high data fidelity. However, the current HPC community is facing several challenges to effectively and efficiently utilize error-bounded lossy compression in practice. This talk will cover three related research topics, including: (1) developing and optimizing error-bounded lossy compression for extreme-scale HPC applications, (2) exploring error-bounded lossy compression in diverse computing scenarios, and (3) enhancing cyberinfrastructure via hardware-algorithm co-designed lossy compression.

 Bio: Dr. Dingwen Tao is an Assistant Professor in the Department of Computer Science at the University of Alabama. He founded the High-Performance Data Analytics and Computing Lab at UA, and the lab has 3 Ph.D. students and 2 undergraduate students. He received his bachelor’s degree in mathematics from the University of Science and Technology of China in 2013 and his Ph.D. degree in computer science from the University of California, Riverside in 2018. Before joining the university as faculty, he interned at Pacific Northwest National Laboratory, Argonne National Laboratory, and Brookhaven National Laboratory. His research interests include high-performance computing, parallel and distributed systems, and big data analytics. Specifically, he focuses on scientific data reduction and management, resilience and fault tolerance, and large-scale machine learning & deep learning. He has published over 30 peer-reviewed high-quality papers in prestigious HPC and Big Data conferences and journals, such as ACM ICS, HPDC, PPoPP, SC, IEEE BigData, CLUSTER, IPDPS, MSST, TPDS, IJHPCA, including two Best Paper awards. His current research has been supported by U.S. DOE and NOAA.