Scaling Molecular Dynamics with ab initio Accuracy to 149 Nanoseconds per Day
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913800471445504 |
|---|---|
| author | Li, Jianxiong Li, Boyang Guo, Zhuoqiang Li, Mingzhen Li, Enji Liu, Lijun Yuan, Guojun Wang, Zhan Tan, Guangming Jia, Weile |
| author_facet | Li, Jianxiong Li, Boyang Guo, Zhuoqiang Li, Mingzhen Li, Enji Liu, Lijun Yuan, Guojun Wang, Zhan Tan, Guangming Jia, Weile |
| contents | Physical phenomena such as chemical reactions, bond breaking, and phase transition require molecular dynamics (MD) simulation with ab initio accuracy ranging from milliseconds to microseconds. However, previous state-of-the-art neural network based MD packages such as DeePMD-kit can only reach 4.7 nanoseconds per day on the Fugaku supercomputer. In this paper, we present a novel node-based parallelization scheme to reduce communication by 81%, then optimize the computationally intensive kernels with sve-gemm and mixed precision. Finally, we implement intra-node load balance to further improve the scalability. Numerical results on the Fugaku supercomputer show that our work has significantly improved the time-to-solution of the DeePMD-kit by a factor of 31.7x, reaching 149 nanoseconds per day on 12,000 computing nodes. This work has opened the door for millisecond simulation with ab initio accuracy within one week for the first time. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_22867 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Scaling Molecular Dynamics with ab initio Accuracy to 149 Nanoseconds per Day Li, Jianxiong Li, Boyang Guo, Zhuoqiang Li, Mingzhen Li, Enji Liu, Lijun Yuan, Guojun Wang, Zhan Tan, Guangming Jia, Weile Distributed, Parallel, and Cluster Computing 82M37, J.2; I.6.3; C.3 Physical phenomena such as chemical reactions, bond breaking, and phase transition require molecular dynamics (MD) simulation with ab initio accuracy ranging from milliseconds to microseconds. However, previous state-of-the-art neural network based MD packages such as DeePMD-kit can only reach 4.7 nanoseconds per day on the Fugaku supercomputer. In this paper, we present a novel node-based parallelization scheme to reduce communication by 81%, then optimize the computationally intensive kernels with sve-gemm and mixed precision. Finally, we implement intra-node load balance to further improve the scalability. Numerical results on the Fugaku supercomputer show that our work has significantly improved the time-to-solution of the DeePMD-kit by a factor of 31.7x, reaching 149 nanoseconds per day on 12,000 computing nodes. This work has opened the door for millisecond simulation with ab initio accuracy within one week for the first time. |
| title | Scaling Molecular Dynamics with ab initio Accuracy to 149 Nanoseconds per Day |
| topic | Distributed, Parallel, and Cluster Computing 82M37, J.2; I.6.3; C.3 |
| url | https://arxiv.org/abs/2410.22867 |