1 / 6

Working Session II – Future Issues:

Explore the challenges and potential solutions for achieving interoperability and sharing of models/infrastructure in the context of exascale computing. Discuss the need for governance and the importance of software sharing before establishing community coupler governance.

cstokes
Download Presentation

Working Session II – Future Issues:

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Working Session II – Future Issues: interoperability, sharing of models/infrastructure, governance and exascale?

  2. c h emi s t r y a t m o s P h e r e pe1 pe1 pe2 pe2 pe3 pe3 pe4 pen pen • OASIS3-MCT: • Very successful example of software sharing • OASIS API and upper layer for simpler implementation in model codes • OASIS flexibility and design principles: • user’s defined source or target (end-point communication) • sending or receiving at appropriate time only • automatic averaging/accumulation if requested • automatic writing of coupling restart file at end of run • MCT for parallel regridding • MCT parallel communication and redistribution Sharing of models/infrastructure

  3. IS-ENES2 WP4 T4 Governance of a community coupler • D4.3 Coupler Governance model document (CERFACS) “proactive involvement by the user community, both in OASIS development and its governance, is urgently needed because of the jump in complexity of the coupling problem on future, massively parallel platforms” Governance OASIS user survey to evaluate their engagement in the development and governance of a next-generation community coupler  o   The strategic requirement. o   The appetite for engagement.  o   The structure.  o   Funding model.  o   Need for more software sharing before possible governance • 10 replies: Met Office (UK), EC-Earth consortium, MPI-M (Germany), Météo-France, IPSL (France), CMCC & INGV (Italy), JAMSTEC (Japan), NEMO consortium , 2 individual expert users

  4. The strategic requirement Do you think that sharing common coupling software, like the OASIS coupler used today by a majority of the climate modelling groups in Europe, is a good thing? 9 yes, 1 maybe Do you think the different groups should keep on sharing their coupling software in the mid and long term? 9 yes, 1 no (other plans) • The appetite for engagement Do you think community governance should be established for OASIS on the short term and for its evolution or the use of alternative software in the mid and long term? 6 yes; 1 yes if ; 3 no, sharing at working level more important than governance • The structure If so, what type of governance should be established? Should it include an OASIS user group (feedbacks and advices from the users to the developers)? 9 yes and a higher-level more strategic groups (technical development, quality control, resources and funding) ? 3 no, 2 maybe, 4 yes Governance

  5. Funding model.  If so, what should be the funding model of community coupling software? One group develop the software with the different partners contributing money? 3 possibly Should the different partners contribute efforts to the software development, this resulting in a geographically distributed software development? 2 maybe, maybe not;   5 no Should sustainable funding be sought from some form of research infrastructure? 4 yes; 3 yes with coordination via ENES 2 securing long-term funding is not needed • Need for more software sharing before possible governance. Alternatively, do you think that sharing more software, naturally leading to distributing the software development effort over more groups, is a mandatory condition before trying to establish some community software governance for the coupler? 3 no; 3 no but would help Governance

  6. Things seems to be fine the short term but what about the long term? • How should we interact to address coupling issues for the exascale? • Should we just try to interact more at the working level? • Should we try to revive/redefine the iWCCT (international Working Committee on Coupling Technology): Cecelia DeLuca, Graham, Robert Jacob, Mick Carter, Tony Craig, Rocky Dunlap, Steve Easterbrook, Florent Duchaine, Scott Peckham, Joachim Biercamp, Bryan Lawrence, Mike Hobson, Gerhard Theurich, Ford Rupert? • Should we try to implement an international/European/American funded / more formal project? • How can we establish more concrete links with the computer scientists? • Should we interact with other communities addressing multi-physics multi-scale coupled problems? • … Issue to discuss

More Related