Parallel I∕O in Flexible Modelling System (FMS) and  Modular Ocean Model 5 (MOM5)

Yang, Rui; Ward, Marshall; Evans, Ben

doi:https://doi.org/10.5194/gmd-13-1885-2020

Articles | Volume 13, issue 4

https://doi.org/10.5194/gmd-13-1885-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Special issue:

The Australian Community Climate and Earth System Simulator:...

https://doi.org/10.5194/gmd-13-1885-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 13, issue 4

Development and technical paper

|

16 Apr 2020

Development and technical paper |

| 16 Apr 2020

Parallel I∕O in Flexible Modelling System (FMS) and Modular Ocean Model 5 (MOM5)

Rui Yang, Marshall Ward, and Ben Evans

Download

Final revised paper (published on 16 Apr 2020)
Preprint (discussion started on 11 Oct 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review of Parallel I/O in FMS and MOM5', Nikolay V. Koldunov, 13 Nov 2019
- AC1: 'Response to referee's comments', Rui Yang, 16 Jan 2020
SC1: 'Executive editor comment: code and data availability.', David Ham, 29 Jan 2020
- AC2: ''Code availability' is updated', Rui Yang, 15 Mar 2020
RC2: 'Review of Parallel I/O in FMS and MOM5', Michael Kuhn, 02 Mar 2020
- AC3: 'response to referee's comments', Rui Yang, 15 Mar 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Rui Yang on behalf of the Authors (15 Mar 2020) Author's response Manuscript

ED: Publish subject to technical corrections (18 Mar 2020) by Olivier Marti

AR by Rui Yang on behalf of the Authors (19 Mar 2020) Manuscript

Short summary

Parallel I/O is implemented in the Modular Ocean Model (MOM) with optimal performance over a range of tuning parameters of model configuration, netCDF, MPI-IO and Lustre filesystem. The scalable parallel I/O performance is observed at 0.1° resolution global model, and it could achieve up to 60 times faster in write speed relative to serial single-file I/O running on 960 PEs.