13 February 2019

PRESS RELEASE: The path to materials design and beyond eXascale starts with the launch of MaX European Centre of Excellence (CoE)



December 2018 marks the launch of the European partnership project entitled: MaX - “MAterials design at the eXascale. European Centre of Excellence in materials modelling, simulations, and design supported by the European Union’s Horizon 2020 project and funded under the H2020-INFRAEDI-2018 call. MaX sits in prime position, together with another nine recently-launched, highly-motivated Centres of Excellence (CoE), operating in the field of High-Performance Computing (HPC), High Throughput Computing (HTC) and High-Performance Data analysis (HPDA).

Over the following thirty-six months MAX CoE, will build on the successful experience, synergy, and community of the Centre established in the previous project (funded under H2020 EINFRA-5-2015) which focused on the following results and gravitated around three paramount goals:

  • Enabling the transition of materials’ codes to eXascale;
  • Convergence of HPC, HTC and HPDA;
  • Transition to the pivotal role of co-design in hardware and software.

Achievements Today

During the first phase [2015-2018], MaX delivered significant results in expected impact areas, including:

  • Improved access to computing applications and expertise that enables researchers and industry to be more productive, leading to scientific excellence, discovery and economic and social benefit.
  • Consolidated European leadership in applications in eXascale and extreme scale-oriented materials simulation codes that address societal challenges.
  • Improved competitiveness for companies and SMEs through access to CoE expertise and services.
  • Significant societal Impacts, where MaX codes contributed to solving important societal challenges (in energy, health or information), as materials are a key factor in finding viable and sustainable solutions.
  • Capacity training for a large number of scientists and engineers working on the use of computational methods and the development of applications in the materials domain.

Main Challenge Ahead: Driving to eXascale and beyond

The challenges that lie ahead for the fourteen-partner-strong consortia lies in the co-design activities to ensure that future HPC architectures are well suited for materials domain applications and vice-versa. The project, through an ambitious workplan, looks to widen the access to codes, provide workflows and turn-key solutions to empower user communities in materials simulations.  This second phase fosters the engagement and know-how of users communicating in industry and academia. Moreover, the second phase has added members to the original consortia, key players in open source community codes and Tier-0 supercomputing centres, complementing the partnership with strengthened technology and training components.

Prof. Elisa Molinari, MaX Director at CNR Nano at Modena in Italy, the Coordinator also of the second phase going forward enthusiastically highlights that:

 

"The MAX Centre of Excellence will pave the way for the transition to and beyond exascale technologies by bringing the most successful     and the most widely used open-source, community codes in quantum simulations of materials to the exascale and by adopting a   sustainable software development strategy to deal with the sudden technological disruptions to be expected in the coming years.

 

This strategy will allow to build capacity, competence and services that can be transitioned to large and small enterprises, and to other research and academic communities, fostering the entire HPC ecosystem in Europe and reducing the costs and barriers preventing its full exploitation. The achievement of the main goal will be carried out by implementing four key-actions:

  1. Restructure MaX flagship codes towards eXascale and extreme scaling performance, in order to address frontier scientific, industrial or societal challenges.
  2. Co-design activities to ensure that future HPC architectures are well-suited for the applications and their users, providing them with high performance and scalable application base.
  3. Develop a broader ecosystem enabling the convergence of HPC, HTC and HPDA.
  4. Widen the access to codes and foster transfer of know-how to user communities

MaX Pilot Projects

Four Pilot Projects have been identified to demonstrate the use of data produced by flagship codes in the context of HPDA, HPC and HTC calculations:

  1. Predicting Code Performance.
  2. Configuration explorer/data explorer toolkit.
  3. Dissemination of highly-curated computational materials data.
  4. Edge computing: Deployment of data analysis solutions co-located with experiments.

MaX Demonstrators

One of the core objectives of MaX is to define and plan the onboarding of new demonstrators, and has a dedicated task to identify additional demonstrators over the lifespan of the project addressing specific scientific challenges that would not be possible without the technology developments of the codes which are as follows:

  1. Pre-exascale-ready Demonstrators.
  2. Reduction of the complexity in very large-scale systems.
  3. Full DFT simulation of new, particle-like objects in chiral magnets.
  4. RPA and double hybrid-based MD simulations of condensed phase systems.
  5. Coupled electrons and phonon dynamics within NEQ–MBPT in technologically relevant materials.

MaX Partnership

MAX is built around a core of world-renowned research teams in the materials science field, with an extensive and established track-record in the development of domain-specific software and its application to frontier research, that is CNR (IT), MaX coordinator, SISSA (IT), ICN2 (ES), JUELICH (DE), CEA (FR), ETHZ (EH) and EPFL (CH). In addition to these partners focused on specific codes, the team at UGENT (BE) has demonstrated unique expertise in code validation.

The scientific core of the consortium is closely linked to a shell of leading European HPC Centres that is CINECA (IT), JUELICH-Supercomputing Centre (DE), BSC (ES), and ETHZ-CSCS (CH), CEA (FR). These HPC teams have proven capabilities in systems, software, hardware, data management and storage, energy efficiency and security, in addition to extended experience in serving the materials community.

The technical partners are further supplemented by the Arm organisation (UK) and the E4 company (IT), which focus on complementary co-design activities.

The partner ICTP has unique competence in HPC training and dissemination in materials science.

Finally, the partner TRUST-IT is an SME specialized in and responsible for the communication, stakeholder engagement and outreach activities of MAX.

MaX involvement with verticals

MaX is committed to collaborate with the Coordination and Support Action (FocusCoE CSA[) to complement MAX vertical domain-specific activities with transversal initiatives aimed at maximising the exploitation of the CoE results and reaching a critical mass on subjects such as industrial outreach (especially towards SMEs), where individual CoE activities may be less effective.

MaX and synergies with the European Open Science Cloud (EOSC)

MaX, in the context of the European Open Science Cloud- EOSC, the European Commission initiative that looks to reinforce Open Science, Open Innovation and Open to the world policies, is looking to synergise with the developments taking place on EOSC in these months, as currently the project offers simulation data and services, in the true spirit of open science, by allowing to run open-source simulation codes on the cloud, supercomputing centres, and local resources (deploying virtual machines). It will look to introduce its long-term archival repository of computational data in materials science found, into EOSC, where today you can browse the five sections of the Materials Cloud (Learn, Work, Discover, Explore, Archive). Currently, the data may be deposited in free form or as AiiDA databases. DOIs are assigned, 10+ year preservation is insured, which is currently an official repository of Nature "Scientific Data” and are recorded in fairsharing and re3data.


Visit our web pages and Join the MaX community at max-centre.eu

Follow us on our Social media channels: