MGARD: A multigrid framework for high-performance, error-controlled data compression and refactoring

We describe MGARD, a software providing MultiGrid Adaptive Reduction for floating-point scientific data on structured and unstructured grids. With exceptional data compression capability and precise error control, MGARD addresses a wide range of requirements, including storage reduction, high-perfor...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	arXiv.org 2024-01
Hauptverfasser:	Gong, Qian, Chen, Jieyang, Whitney, Ben, Liang, Xin, Reshniak, Viktor, Banerjee, Tania, Lee, Jaemoon, Rangarajan, Anand, Wan, Lipeng, Vidal, Nicolas, Liu, Qing, Gainaru, Ana, Podhorszki, Norbert, Archibald, Richard, Ranka, Sanjay, Klasky, Scott
Format:	Artikel
Sprache:	eng
Schlagworte:	Application programming interface Computer Science - Computer Vision and Pattern Recognition Computer Science - Numerical Analysis Data analysis Data compression Floating point arithmetic Mathematics - Numerical Analysis Reduction Unstructured data Unstructured grids (mathematics)
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	We describe MGARD, a software providing MultiGrid Adaptive Reduction for floating-point scientific data on structured and unstructured grids. With exceptional data compression capability and precise error control, MGARD addresses a wide range of requirements, including storage reduction, high-performance I/O, and in-situ data analysis. It features a unified application programming interface (API) that seamlessly operates across diverse computing architectures. MGARD has been optimized with highly-tuned GPU kernels and efficient memory and device management mechanisms, ensuring scalable and rapid operations.
ISSN:	2331-8422
DOI:	10.48550/arxiv.2401.05994