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

Qian Gong; Jieyang Chen; Ben Whitney; Xin Liang; Viktor Reshniak; Tania Banerjee; Jaemoon Lee; Anand Rangarajan; Lipeng Wan; Nicolas Vidal; Qing Liu; Ana Gainaru; Norbert Podhorszki; Richard Archibald; Sanjay Ranka; Scott Klasky

doi:10.1016/j.softx.2023.101590

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

Qian Gong, Jieyang Chen, Ben Whitney, Xin Liang, Viktor Reshniak, Tania Banerjee, Jaemoon Lee, Anand Rangarajan, Lipeng Wan, Nicolas Vidal, Qing Liu, Ana Gainaru, Norbert Podhorszki, Richard Archibald, Sanjay Ranka, Scott Klasky

Research output: Contribution to journal › Article › peer-review

4 Scopus citations

Abstract

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.

Original language	English
Article number	101590
Journal	SoftwareX
Volume	24
DOIs	https://doi.org/10.1016/j.softx.2023.101590
State	Published - Dec 2023

Funding

This research was supported in part by the Exascale Computing Project CODAR (17-SC-20-SC) of the US Department of Energy (DOE), the DOE's Advanced Scientific Research Office (ASCR) research project SIRIUS-2, United States, and the DOE's RAPIDS-2 SciDAC project under contract number DE-AC05-00OR22725. In addition, this research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of DOE under Contract Numbers DE-AC05-00OR22725. This research was supported in part by the Exascale Computing Project CODAR (17-SC-20-SC) of the US Department of Energy (DOE), the DOE’s Advanced Scientific Research Office (ASCR) research project SIRIUS-2, United States , and the DOE’s RAPIDS-2 SciDAC project under contract number DE-AC05-00OR22725 . In addition, this research used resources of the Oak Ridge Leadership Computing Facility at the Oak Ridge National Laboratory, which is supported by the Office of Science of DOE under Contract Numbers DE-AC05-00OR22725.

Funders	Funder number
DOE's Advanced Scientific Research Office
DOE’s Advanced Scientific Research Office
U.S. Department of Energy
Office of Science
Advanced Scientific Computing Research	DE-AC05-00OR22725

Keywords

Data refactoring
Derived quantities preservation
Error-controlled data compression
I/O acceleration

Access to Document

10.1016/j.softx.2023.101590

Cite this

Gong, Q., Chen, J., Whitney, B., Liang, X., Reshniak, V., Banerjee, T., Lee, J., Rangarajan, A., Wan, L., Vidal, N., Liu, Q., Gainaru, A., Podhorszki, N., Archibald, R., Ranka, S., & Klasky, S. (2023). MGARD: A multigrid framework for high-performance, error-controlled data compression and refactoring. SoftwareX, 24, Article 101590. https://doi.org/10.1016/j.softx.2023.101590

@article{727cb8400f1640cc8ecda7c4c12abe24,

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

abstract = "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.",

keywords = "Data refactoring, Derived quantities preservation, Error-controlled data compression, I/O acceleration",

author = "Qian Gong and Jieyang Chen and Ben Whitney and Xin Liang and Viktor Reshniak and Tania Banerjee and Jaemoon Lee and Anand Rangarajan and Lipeng Wan and Nicolas Vidal and Qing Liu and Ana Gainaru and Norbert Podhorszki and Richard Archibald and Sanjay Ranka and Scott Klasky",

note = "Publisher Copyright: {\textcopyright} 2023",

year = "2023",

month = dec,

doi = "10.1016/j.softx.2023.101590",

language = "English",

volume = "24",

journal = "SoftwareX",

issn = "2352-7110",

publisher = "Elsevier",

}

TY - JOUR

T1 - MGARD

T2 - A multigrid framework for high-performance, error-controlled data compression and refactoring

AU - Gong, Qian

AU - Chen, Jieyang

AU - Whitney, Ben

AU - Liang, Xin

AU - Reshniak, Viktor

AU - Banerjee, Tania

AU - Lee, Jaemoon

AU - Rangarajan, Anand

AU - Wan, Lipeng

AU - Vidal, Nicolas

AU - Liu, Qing

AU - Gainaru, Ana

AU - Podhorszki, Norbert

AU - Archibald, Richard

AU - Ranka, Sanjay

AU - Klasky, Scott

PY - 2023/12

Y1 - 2023/12

N2 - 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.

AB - 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.

KW - Data refactoring

KW - Derived quantities preservation

KW - Error-controlled data compression

KW - I/O acceleration

UR - http://www.scopus.com/inward/record.url?scp=85177818535&partnerID=8YFLogxK

U2 - 10.1016/j.softx.2023.101590

DO - 10.1016/j.softx.2023.101590

M3 - Article

AN - SCOPUS:85177818535

SN - 2352-7110

VL - 24

JO - SoftwareX

JF - SoftwareX

M1 - 101590

ER -

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

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this