Within the ESiWACE2 H2020 European infrastructure project, WP 3 « HPC services to prepare the weather and climate community for the pre-Exascale » provides services to improve performance and portability of climate codes with respect to existing and upcoming tier1 and tier0 computers. In this context,
we provide dedicated TECHNICAL help to design, upgrade or enhance the implementation of XIOS I/O server in your model(s) to set up tailored and computationally efficient I/Os.
XIOS (XML-IO-Server) is a library developed at IPSL dedicated to parallel I/O management in climate codes. XIOS manages output of diagnostics and other data produced by climate component codes into files and offers temporal and spatial post-processing operations on this data. XIOS user-interface (available for C/C++/Fortran) consists of a defined set of subroutines that facilitates its interfacing in model codes. I/O configuration is done though a set of XML files, using a specific syntax. To help the writing of such XML configuration files, a python tool was developed by CNRM for the need of the CMIP6 project. This tool can be used outside of CMIP6 and should prove to be very useful in case of a long list of output variables is requested.
This year, one 2 person-month Dedicated User Support will be offered. Provided by CERFACS and IPSL, this service is open to any weather/climate research laboratory in Europe. The service will consist in remote dedicated support on the user code and if required, an on-site visit by the XIOS support can be organized. The granted applicant will only have to give the XIOS support team an access to its codes and computing platform. All other costs are covered by the project.
This service excludes the scientific analysis and evaluation of the model outputs.
To apply:
Briefly describe your project and your needs filling the questionnaire "Description of your XIOS project” (5 pages maximum)
Send your questionnaire to Marie-Pierre Moine (moine[at]cerfacs.fr).
The deadline for this 1st call for applicants is October 31th 2019
Evaluation criteria:
- Potential of the resulting model interfaced with XIOS to be used on pre-Exascale machines, e.g. compatibility with emerging architectures, increased task/model parallelism, reduced precision computations, new programming languages, etc.
- Quality of the methodology proposed
- Expected scientific impact of the resulting model and its long-term support by the applicant group
The selection will be made by the XIOS Advisory Board. The institutions selected will be notified by the end of December 2019.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement N°823988 |