ACROSS Project is a project funded by the European High-Performance Computing Joint Undertaking (EuroHPC JU) under grant agreement No 955648.[1][2] It is supported by the European Union's Horizon 2020 Research and Innovation Programme and several European countries, including Italy, France, the Czech Republic, the United Kingdom, Greece, the Netherlands, Germany, and Norway. The project commenced on 1st March 2021 and is set to run for 3 years with a total budget of €8.8 million. ACROSS Project aims to design and develop an HPC, Big Data, and Artificial Intelligence convergent platform, supporting applications in the Aeronautics, Climate and Weather, and Energy domains.

Project overview

edit

The ACROSS platform provides access to a software stack[3] that enables application workflows combining numerical simulations, artificial intelligence (AI),  and high-performance data analytics (HPDA) tasks to be executed on High-Performance Computing (HPC) resources.[4][5] The platform is designed to take advantage of diverse hardware accelerators, including GPUs, FPGAs, and neuromorphic-like architectures. Hardware specific libraries and frameworks are part of the stack to enable the use of these accelerators.

Partners

edit

The ACROSS consortium consists of 13 contributing organizations from 8 European countries and is coordinated by the LINKS Foundation. The consortium includes Supercomputing Centers, small and medium enterprises, research organizations, and large enterprises, such as CINECA (IT), IT4I (CZ), Atos/Bull (FR), Avio Aero (IT), Morfo (IT), NEUROPUBLIC (GR), INRIA (FR), CINI (IT), SINTEF (NW), MPI-M (DE), DELTARES (NL), and ECMWF (IO).

Technologies

edit

ACROSS leverages StreamFlow[6][7] for parsing and executing CWL-based workflows, WARP for deterministic allocation of HPC resources, HyperQueue for workload distribution on reserved resources, FMLE/YSTIA components for addressing Cloud-based resource management and ML/DL model training, and Damaris middleware[8] for parallel (in-situ) output capabilities in carbon sequestration pilot. Jobs queue-waiting time predictors have also been investigated. It is further integrated with other EU-funded platforms and infrastructures using dedicated components, including the High-End Application Execution Middleware (HEAppE).

Pilots

edit

ACROSS is conducting pilots in various domains, including aero-engine module optimization, weather, climate, hydrological and farming, and energy and carbon sequestration. These pilots aim to demonstrate the benefits of the ACROSS platform in improving advanced numerical modeling capabilities and enhancing global numerical weather prediction, among other objectives.

Greener aero-engine modules optimization

edit

Avio Aero leverages the HPC resources made available through the ACROSS project to improve advanced numerical modeling capabilities for critical engine components. The pilot objective is to enhance effectiveness in designing key aeronautical components by adopting new methods and workflows, Multi-scale/Multiphysics[9] unsteady approaches and AI.[10] Two aeronautical engineering case studies will be rolled out: one regarding the combustor and another one referring to low-pressure turbines design.

Weather, climate, hydrological, and farming

edit

This pilot aims to demonstrate the benefits of the ACROSS platform in the context of three deeply connected workflows: global scale numerical weather predictions, climatological simulations, regional numerical weather predictions, hydrological simulations, along with farming services performed by NEUROPUBLIC.

Energy and carbon sequestration

edit

The pilot has two use cases, both using the reservoir simulator program OPM Flow:[11] carbon sequestration and direct simulation on seismic cubes. The main objectives of the pilot are to improve the capability of performing large-scale carbon geologic sequestration simulations, enable direct subsurface flow simulations on processed seismic data, and develop cross-stack workflows for subsurface simulation/analysis.

References

edit
  1. ^ https://cordis.europa.eu/project/id/955648
  2. ^ https://eurohpc-ju.europa.eu/research-innovation/our-projects/across_en
  3. ^ Kenneally J and Hoppe H-C (2018). "The technology stacks of High Performance Computing and Big Data Computing: What they can learn from each other". A joint publication between the European associations of www.ETP4HPC.eu and www.BDVA.eu.
  4. ^ https://cordis.europa.eu/article/id/452273-how-ai-can-help-us-tackle-future-exascale-workloads
  5. ^ https://cordis.europa.eu/article/id/452268-growing-europes-supercomputing-ecosystem
  6. ^ Colonnelli, Iacopo; Aldinucci, Marco; Cantalupo, Barbara; Padovani, Luca; Rabellino, Sergio; Spampinato, Concetto; Morelli, Roberto; Di Carlo, Rosario; Magini, Nicolò; Cavazzoni, Carlo (2022-03-01). "Distributed workflows with Jupyter". Future Generation Computer Systems. 128: 282–298. doi:10.1016/j.future.2021.10.007. ISSN 0167-739X.
  7. ^ Colonnelli, Iacopo; Cantalupo, Barbara; Merelli, Ivan; Aldinucci, Marco (2021-10-01). "StreamFlow: Cross-Breeding Cloud With HPC". IEEE Transactions on Emerging Topics in Computing. 9 (4): 1723–1737. doi:10.1109/TETC.2020.3019202. ISSN 2168-6750.
  8. ^ Dorier, Matthieu; Antoniu, Gabriel; Cappello, Franck; Snir, Marc; Sisneros, Robert; Yildiz, Orcun; Ibrahim, Shadi; Peterka, Tom; Orf, Leigh (2016-10-25). "Damaris: Addressing Performance Variability in Data Management for Post-Petascale Simulations". ACM Trans. Parallel Comput. 3 (3): 15:1–15:43. doi:10.1145/2987371. ISSN 2329-4949.
  9. ^ Paccati, Simone; Bertini, Davide; Mazzei, Lorenzo; Puggelli, Stefano; Andreini, Antonio (2021-04-01). "Large-Eddy Simulation of a Model Aero-Engine Sooting Flame With a Multiphysics Approach". Flow, Turbulence and Combustion. 106 (4): 1329–1354. doi:10.1007/s10494-020-00202-5. ISSN 1573-1987.
  10. ^ Vinuesa, Ricardo; Brunton, Steven L. (2022-06-27). "Enhancing computational fluid dynamics with machine learning". Nature Computational Science. 2 (6): 358–366. doi:10.1038/s43588-022-00264-7. ISSN 2662-8457.
  11. ^ Rasmussen, Atgeirr Flø; Sandve, Tor Harald; Bao, Kai; Lauser, Andreas; Hove, Joakim; Skaflestad, Bård; Klöfkorn, Robert; Blatt, Markus; Rustad, Alf Birger; Sævareid, Ove; Lie, Knut-Andreas; Thune, Andreas (2021-01-01). "The Open Porous Media Flow reservoir simulator". Computers & Mathematics with Applications. Development and Application of Open-source Software for Problems with Numerical PDEs. 81: 159–185. doi:10.1016/j.camwa.2020.05.014. ISSN 0898-1221.