SC20 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Combining Spatial and Temporal Properties for Improvements in Data Reduction


Workshop:DRBSD-6: The 6th International Workshop on Data Analysis and Reduction for Big Scientific Data

Authors: Megan L. Hickman Fulp (Clemson University), Ayan Biswas (Los Alamos National Laboratory), and Jon C Calhoun (Clemson University)


Abstract: Due to I/O bandwidth limitations, intelligent in situ data reduction methods are needed to enable post-hoc workflows. Current state-of-the-art sampling methods save data points if their region is deemed spatially or temporally important. By analyzing the properties of the data values at each time-step, two consecutive steps may be found to be very similar. This research follows the notion that if neighboring time-steps are very similar, samples from both are unnecessary, which leaves storage for more useful samples to be chosen. Here, we present an investigation of the combination of spatial and temporal sampling to drastically reduce data size without loss of valuable information. We demonstrate that by reusing samples, our reconstructed dataset is able to reduce the overall data size while achieving a higher post-reconstruction quality over other reduction methods.





Back to DRBSD-6: The 6th International Workshop on Data Analysis and Reduction for Big Scientific Data Archive Listing



Back to Full Workshop Archive Listing