Periodic Reporting for period 1 - EXCELLERAT P2 (European Centre of Excellence for Engineering Applications on HPC and associated technplogies)
Reporting period: 2023-01-01 to 2023-12-31
The second phase of EXCELLERAT continues the prolific work carried out in the first phase of the project, where partnerships between Europe’s leading HPC centres, application specialists and supporting partners have been established by sharing their expertise and knowledge with engineering companies and researchers from industry and academia.
In order to prove the benefits of leveraging large-scale HPC applications and deploying pre-Exascale and Exascale HPC systems to solve engineering challenges, seven use cases from the fluid dynamics, combustion and energy domains were considered during the second phase of the project, four of them based on the engineering codes CODA, Alya, Neko and AVBP, which are known from the first project phase of EXCELLERAT P1, where they were optimised in detail for use on Exascale HPC systems. Two more use cases based on the FLEW and m-AIA applications together with a further application based on the codes OpenFoam/Raysect/Elmer joined the Consortium due to the high relevance for exascaling in the aforementioned domains.
EXCELLERAT P2 continues to address the setup of a Centre as an entity, acting as a single hub, covering a wide range of issues, from «non-pure-technical» services, such as access to knowledge or networking up to technical services as e.g. co-design, scalability enhancement or code porting to new (Exascale) Hardware. As the consortium contains key players in HPC, High Performance Data Analytics (HPDA), AI and experts for the reference applications, technological impact (e.g. code improvements and awareness raising) is guaranteed. The scientific excellence of the EXCELLERAT P2 consortium enables evolution, optimisation, scaling and porting of applications towards disruptive technologies and increases Europe´s competitiveness in engineering. Within the frame of the project, EXCELLERAT P2 will prove the applicability of the results for not only the seven chosen reference applications, but even going beyond. Thus, EXCELLERAT P2 will extend the recipients of its developments beyond the consortium and interact via diverse mechanisms to integrate external stakeholders of its value network into its evolution.
In order to prove the benefits of leveraging large-scale HPC applications and deploying pre-Exascale and Exascale HPC systems to solve engineering challenges, seven use cases from the fluid dynamics, combustion and energy domains were considered during the second phase of the project, four of them based on the engineering codes CODA, Alya, Neko and AVBP, which are known from the first project phase of EXCELLERAT P1, where they were optimised in detail for use on Exascale HPC systems. Two more use cases based on the FLEW and m-AIA applications together with a further application based on the codes OpenFoam/Raysect/Elmer joined the Consortium due to the high relevance for exascaling in the aforementioned domains.
EXCELLERAT P2 continues to address the setup of a Centre as an entity, acting as a single hub, covering a wide range of issues, from «non-pure-technical» services, such as access to knowledge or networking up to technical services as e.g. co-design, scalability enhancement or code porting to new (Exascale) Hardware. As the consortium contains key players in HPC, High Performance Data Analytics (HPDA), AI and experts for the reference applications, technological impact (e.g. code improvements and awareness raising) is guaranteed. The scientific excellence of the EXCELLERAT P2 consortium enables evolution, optimisation, scaling and porting of applications towards disruptive technologies and increases Europe´s competitiveness in engineering. Within the frame of the project, EXCELLERAT P2 will prove the applicability of the results for not only the seven chosen reference applications, but even going beyond. Thus, EXCELLERAT P2 will extend the recipients of its developments beyond the consortium and interact via diverse mechanisms to integrate external stakeholders of its value network into its evolution.
During the first year of its lifetime, EXCELLERAT P2 has achieved the following main results:
A shared workspace server has been set up and all necessary management and quality assurance (QA) mechanisms have been established, provided and described in deliverable D1.1 “Project Management Handbook” . Moreover, an extensive coordination and decision-making support for the organisational and technical aspects have been provided throughout the whole first year.
For each of the project’s seven use-cases a development roadmap has been prepared, identifying a number of common themes to be addressed (for more information, please check deliverables D2.1 “Use-Case Execution Roadmap” ). In order to monitor the progress, each use case alternately presents its progress during the WP2 meetings.
The computational efficiency of the simulation code employed in the use case at inter- and intra-node levels were optimised. Advances in code scalability, code optimisation and porting to GPU are reported (please check deliverable D3.1 “Report on Exa-Enabling Methodologies” for more information). Also initial co-design engagements have started.
Significant progress on the in-situ work with the first in-situ visualisation of Alya using Vistle, together with further work on workflow optimisation and development of the various tools to study workflows like marauder's map and SEED has been performed (please check deliverables D4.1 “Workflows for engineering simulations progress report” and D4.4 “Data Management and Data Analytics in EXCELLERAT” for more details).
Re-design the service portal. Initial technical specifications have been defined by task leaders, and information on events and training has been collected and uploaded. Final lists with the redistribution of the EXCELLERAT services have lastly been prepared and reviewed (for more details, we refer to deliverable D5.1 “Services, Training and Portal Progress Report”).
Outline the initial dissemination and communication strategy of the project, including mechanisms and activities planned in terms of community building as well as an initial exploitation plan for the Centre of Excellence (for more details, please check deliverables D7.1 “Plan for dissemination and exploitation including communication activities” and D7.5 “Collaboration plan with definition of common objectives and activities including milestones” ).
In addition, the steps on the road towards the business development of the CoE have been defined (we refer to deliverable D6.1 “Market Assessment” for more details). In total, 19 deliverables were submitted successfully on time.
A shared workspace server has been set up and all necessary management and quality assurance (QA) mechanisms have been established, provided and described in deliverable D1.1 “Project Management Handbook” . Moreover, an extensive coordination and decision-making support for the organisational and technical aspects have been provided throughout the whole first year.
For each of the project’s seven use-cases a development roadmap has been prepared, identifying a number of common themes to be addressed (for more information, please check deliverables D2.1 “Use-Case Execution Roadmap” ). In order to monitor the progress, each use case alternately presents its progress during the WP2 meetings.
The computational efficiency of the simulation code employed in the use case at inter- and intra-node levels were optimised. Advances in code scalability, code optimisation and porting to GPU are reported (please check deliverable D3.1 “Report on Exa-Enabling Methodologies” for more information). Also initial co-design engagements have started.
Significant progress on the in-situ work with the first in-situ visualisation of Alya using Vistle, together with further work on workflow optimisation and development of the various tools to study workflows like marauder's map and SEED has been performed (please check deliverables D4.1 “Workflows for engineering simulations progress report” and D4.4 “Data Management and Data Analytics in EXCELLERAT” for more details).
Re-design the service portal. Initial technical specifications have been defined by task leaders, and information on events and training has been collected and uploaded. Final lists with the redistribution of the EXCELLERAT services have lastly been prepared and reviewed (for more details, we refer to deliverable D5.1 “Services, Training and Portal Progress Report”).
Outline the initial dissemination and communication strategy of the project, including mechanisms and activities planned in terms of community building as well as an initial exploitation plan for the Centre of Excellence (for more details, please check deliverables D7.1 “Plan for dissemination and exploitation including communication activities” and D7.5 “Collaboration plan with definition of common objectives and activities including milestones” ).
In addition, the steps on the road towards the business development of the CoE have been defined (we refer to deliverable D6.1 “Market Assessment” for more details). In total, 19 deliverables were submitted successfully on time.
There are several impacts like the impact through the interest groups or other external stakeholders: The interest group (IG) activity was continued from phase 1 in order to involve external beneficiaries (from academia, industry and public administration) to provide feedbacks. These groups were established to communicate, monitor and validate the overall project goals and to assure the industrial as well as the technological relevance for the user communities. Furthermore, the activities will result in effective (highly scalable) codes, that will be optimized, increased and efficiency and thus be scaled up, so that these existing codes will showcase significant improvements. This impact will at first benefit the code developers, but also their potential users in the background, which in the long term can be from academia, industry or even public administration. The owners of the code will benefit from their improved solution, which they either can use for their own purpose, share it with others or even sell it. Most importantly, the entire existing user community will benefit from the better performance and Exascale capabilities. The impact through services around effective (highly scalable) codes is a more indirect one: having applications on the market that actually can make use of Exascale systems will increase the pressure to other code developers that are addressing the same or similar markets. Part of EXCELLERAT P2´s mission is to further strengthen the community, including all potential stakeholders (e.g.large industry, high tech SMEs etc.). The availability of success stories for different application fields and industry sectors will widely demonstrate what is possible and thus inspire aspirants from the same as well as from other fields and sectors. A success will also be when the developed actions and results can be adapted to be used by external activities. Thus, EXCELLERAT P2 will be as open as possible to requests from outsiders and to collaboration with a wide variety of activities in Europe.