SD3: The First Operationally-Proven, Fully Heterogeneous-Accelerated Global Kilometer-Scale Model with Inline Physics Architecture
Abstract
We present Super Dynamics on the Cube (SD3), the first operationally-proven, fully heterogeneous-accelerated global kilometer-scale model. Through a full-stack reconstruction in native CUDA C, SD3 introduces an "Inline Physics" architecture that deeply couples its non-hydrostatic cubed-sphere dynamical core with fast physical processes. This design maintains high physical consistency, ensures energy and mass conservation, and mitigates "Grey Zone" mismatches. Five years of continuous operational forecasting, together with validation against diverse observations, demonstrates the scientific fidelity of SD3. Recently, SD3 achieved an unprecedented 205.7 simulated days per day at global 1.5 km resolution using 24576 GPUs. SD3 establishes a hardware-agnostic, portable, and reproducible paradigm across diverse Exascale ecosystems. Beyond weather and climate research, SD3 offers a transformative paradigm for the HPC community and the next-generation Digital Twin Earth to establish a robust, high-performance blueprint for the future of Earth system modeling.
Related
- Evaluating LLMs for Translating Legacy Fortran Atmospheric Physics to GPU-Accelerated JAX
- Physics, Ecology, Mathematics and Engineering of the Negawatt Economy: A Formal Architecture for Verified Energy Conservation
- Spatial Heterogeneity and Inhomogeneity of Cirrus Microphysical Properties on Kilometer Spatial Scales
- Kilometer-Scale AI Downscaling of Atlantic Hurricanes with Generative Ensembles
- Diffusion-Based Refinement for Kilometer-Scale Probabilistic Precipitation Nowcasting
BibTeX
@misc{liang2026first,
title = {{SD3: The First Operationally-Proven, Fully Heterogeneous-Accelerated Global Kilometer-Scale Model with Inline Physics Architecture}},
author = {Yuan Liang and Xiaomin Hu and Xi Chen and Jingsheng Zhu and Li Yang and Leisheng Li and Qian Song and Xu Wang and Hao Wang and Zhenzhen Wang and Huiyuan Li},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.39334},
eprint = {2610.39334},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.39334},
}
RIS
TY - UNPB
TI - SD3: The First Operationally-Proven, Fully Heterogeneous-Accelerated Global Kilometer-Scale Model with Inline Physics Architecture
AU - Yuan Liang
AU - Xiaomin Hu
AU - Xi Chen
AU - Jingsheng Zhu
AU - Li Yang
AU - Leisheng Li
AU - Qian Song
AU - Xu Wang
AU - Hao Wang
AU - Zhenzhen Wang
AU - Huiyuan Li
PY - 2026
DA - 2026/10/29
PB - arXiv
DO - 10.48550/arxiv.2610.39334
UR - https://arxiv.org/abs/2610.39334
AB - We present Super Dynamics on the Cube (SD3), the first operationally-proven, fully heterogeneous-accelerated global kilometer-scale model. Through a full-stack reconstruction in native CUDA C, SD3 introduces an "Inline Physics" architecture that deeply couples its non-hydrostatic cubed-sphere dynamical core with fast physical processes. This design maintains high physical consistency, ensures energy and mass conservation, and mitigates "Grey Zone" mismatches. Five years of continuous operational forecasting, together with validation against diverse observations, demonstrates the scientific fidelity of SD3. Recently, SD3 achieved an unprecedented 205.7 simulated days per day at global 1.5 km resolution using 24576 GPUs. SD3 establishes a hardware-agnostic, portable, and reproducible paradigm across diverse Exascale ecosystems. Beyond weather and climate research, SD3 offers a transformative paradigm for the HPC community and the next-generation Digital Twin Earth to establish a robust, high-performance blueprint for the future of Earth system modeling.
ER -