Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces BlockMGARD, a novel GPU-based lossy compressor designed to enhance the efficiency of scientific data reduction through adaptive Region-of-Interest (ROI) error control. By leveraging an In-cache Block decomposition and a hybrid hierarchy, BlockMGARD significantly accelerates both compression and decompression processes while maintaining high compression ratios, outperforming existing methods like MGARD-X. The results demonstrate that BlockMGARD achieves up to 4.2x and 9.1x higher throughput and up to 8.63x better compression ratios compared to uniform-tolerance baselines, showcasing its effectiveness across multiple GPUs with near-linear scaling.
BlockMGARD accelerates scientific data compression on GPUs, achieving up to 9.1x faster throughput while preserving critical features through adaptive error control.
The growing scale of scientific data makes lossy compression essential for reducing data volume under controllable error. Transformation-based compressors using multilevel decomposition, such as MGARD, achieve strong compression ratios but map poorly to GPU architectures. We propose BlockMGARD, an adaptive, Region-of-Interest (ROI)-supported GPU lossy compressor, with four contributions: (1) an In-cache Block decomposition leveraging GPU on-chip memory and constant lookup tables to accelerate decomposition; (2) a hybrid hierarchy combining In-cache Block and global decomposition to balance speed and compression ratio; (3) an end-to-end pipeline with fine-grained ROI error control for feature preservation; and (4) an evaluation against state-of-the-art methods on five real-world datasets. Compared to MGARD-X, BlockMGARD achieves up to 4.2x and 9.1x higher compression and decompression throughput, and up to 8.63x higher compression ratio than uniform-tolerance baselines under ROI-aware error control. Across four GPUs, BlockMGARD achieves near-ideal linear scaling and up to 1.58x I/O cost reduction over MGARD-X.