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