David Krasowska received first place in the ACM Student Research Competition (Undergraduate) for his work Statistical Prediction of Lossy Compression Ratios for 3D Scientific Data. He explored statistical methods to characterize the data correlation structures and their relationships, through functional models, to compression ratios and quality metrics for 3D scientific data. He views the work as a step toward identifying the theoretical upper limit of lossy compressibility. Krasowska’s research was done during his summer internship in the MCS division with Julie Bessac, Sheng Di, Robert Underwood and Franck Cappello.
His poster is available at https://sc22.supercomputing.org/proceedings/src_poster/poster_files/spostu110s3-file1.pdf. Krasowska did his undergraduate work at Clemson University and is now a Ph.D. student at Northwestern University.
Milan Shah received second place in the ACM Student Research Competition (Graduate) for his work Compressing Quantum Circuit Simulation Tensor Data. Using entropy as a metric, he studied the effects of various lossy compression/decompression strategies on tensor data compressibility, throughput and result error. The study showed greater compressibility with the datasets SZ and SZx, but Shah notes the need for further research on other metrics. Shah’s work was done during his summer internship in the MCS division with Xiaodong Yu, Sheng Di and Franck Cappello.
His poster is available at https://sc22.supercomputing.org/proceedings/src_poster/poster_files/spostg112s3-file1.pdf. Shah is a doctoral student at now North Carolina State University and a visiting student at Argonne.
The ACM Student Research Competition was started in 2003. It provides a forum for undergraduate and graduate students to present their research at ACM conferences.