Simulation of the performance and scalability of message passing interface (MPI) communications of atmospheric models running  on exascale supercomputers

Zheng, Yongjun; Marguinaud, Philippe

doi:https://doi.org/10.5194/gmd-11-3409-2018

Articles | Volume 11, issue 8

https://doi.org/10.5194/gmd-11-3409-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-11-3409-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 11, issue 8

Methods for assessment of models

|

22 Aug 2018

Methods for assessment of models |

| 22 Aug 2018

Simulation of the performance and scalability of message passing interface (MPI) communications of atmospheric models running on exascale supercomputers

Yongjun Zheng and Philippe Marguinaud

Model code and software

Simulation of the Performance and Scalability of MPI Communications of Atmospheric Models Running on Exascale Supercomputers Y. Zheng https://doi.org/10.5281/zenodo.1066934

Short summary

The impacts of collective algorithms, interconnect topologies, and routing algorithms on the performance and scalability of transpositions, halo exchanges, and allreduce operations are significant. The performance of communications can be improved when the congestion is mitigated by a proper configuration of the topology and a congestion-balanced routing algorithm. The scalability of the spectral models could be acceptable for exascale supercomputers.