High-Throughput Computing and Data storage

At MaX, we advance materials research through a state-of-the-art informatics ecosystem that combines high-throughput calculations with transparent data management. Our commitment ensures that all research data is FAIR — Findable, Accessible, Interoperable, and Reusable. 

To discover novel materials and design their properties, researchers run thousands of HPC simulations, which require sophisticated management of both simulations and data. MaX provides AiiDA, an automated interactive infrastructure and database for computational materials science based on the four “ADES” pillars — Automation, Data, Environment, and Sharing (Pizzi et al, 2016). AiiDA can manage and automate tens of thousands of simulations concurrently, from local clusters to next-generation supercomputers, while automatically storing inputs, outputs, and the complete provenance of calculations, ensuring full reproducibility and traceability of all computational research.

Research results are shared and disseminated through Materials Cloud, where data can be uploaded, visualized, analyzed, and accessed by the scientific community. AiiDA and Materials Cloud provide a streamlined platform for FAIR data management, ensuring that research data is:

  • Findable via DOIs and standardised metadata in the Materials Cloud Archive.

  • Accessible through interactive interfaces, including curated data and raw AiiDA provenance graphs.

  • Interoperable using code-agnostic formats for reuse across different software.

  • Reusable under open licenses, with strong reproducibility guarantees from provenance tracking.

Materials Cloud also serves as a simulation platform, enabling new simulations both locally and in the cloud via AiiDA lab, Quantum Mobile, and interactive online tools, accelerating materials discovery and collaborative computational research.

Our Infrastructure

MaX leverages AiiDA as the workflow engine and Materials Cloud as the cloud-based platform for simulation, dissemination, and data sharing. This powerful combination allows researchers to:

  • Implement, run, and share computational workflows and turnkey solutions.

  • Automatically track, store, and preserve data provenance.

  • Ensure reproducibility, reuse, and advanced data analytics across materials research.

By providing automated high-throughput calculations and graph database storage, MaX strengthens the computational materials science community in Europe and beyond. Our ecosystem accelerates materials discovery, fosters collaboration, and guarantees that research outputs are fully accessible and reusable for future innovations.

AiiDA

…the “operating system” for FAIR materials science simulations.

AiiDA is the backbone of our operations. As an automated interactive infrastructure, AiiDA handles vast numbers of simulations and ensures that everything, from data input/output to provenance tracking, is well managed. The complete history of how data is generated and used is archived, ensuring full reproducibility. AiiDA also empowers our turn-key workflows, allowing even non-experts to compute material properties easily, and extends the scope of reproducibility beyond just rerunning calculations.

aiida_slide
materialscloud_slide

Materials Cloud

The FAIR dissemination platform for data and simulation services

MaterialsCloud is our primary cloud simulation and web dissemination platform, and it plays an integral role in data sharing. Partnering with AiiDA, it facilitates the upload, visualization, and sharing of research results. Materials Cloud stands out for its commitment to FAIR data, ensuring data is easily found through digital object identifier (DOI), accessible via its interactive interface, interoperable with code-agnostic formats, and reusable with strong reproducibility from its provenance graph. Additionally, Materials Cloud offers both curated data and a wide array of interactive online tools.

Promoting Open Science

  • Curated Data: Our Materials Cloud DISCOVER sections present curated scientific outcomes in an interactive visual format, ensuring intuitive understanding. All underlying research data here is available in the ARCHIVE records, linked with DOIs.
  • Data on Demand: With the AiiDA lab web platform, we offer access to intuitive Jupyter apps. Users can effortlessly execute complex simulation workflows and share turnkey solutions, making science more accessible to all.
  • Public Repository: The Materials Cloud ARCHIVE is a free, open-access repository that ensures data preservation for a minimum of a decade post-deposition. Each record has its unique digital object identifier (DOI). The Materials Cloud ARCHIVE is non-commercial and free of charge.
  • FAIR Data Sharing: Beyond traditional publishing, we emphasize comprehensive data and metadata sharing. The Materials Cloud Archive, registered on platforms like FAIRsharing.org and re3data, is globally accessible and guarantees prolonged storage. Our adherence to FAIR principles ensures that data is not simply stored, but that it is meaningful, connected, and ready for further use and application.

The "operating system" for FAIR materials science simulations: AiiDA

To discover novel materials and to design materials properties, researchers need to run thousands of HPC simulations. Managing these simulations and the resulting data is not trivial. The tools delivered by MaX to enable these high-throughput simulations is AiiDA, an automated interactive infrastructure and database for computational science based on the four “ADES” pillars of automation, data, environment and sharing (Pizzi et al, 2016). AiiDA can manage and automate tens of thousands of simulations concurrently, scaling from local resources to next-generation supercomputers. At the same time, it stores automatically input and output data and, most importantly, the whole provenance (i.e., the connection between data and calculations, with a full description of how data was generated) ensuring full reproducibility.

The FAIR dissemination platform for data and simulation services: Materials Cloud

AiiDA simplifies the task of sharing the results between researchers and of disseminating research results thanks to its coupling with the Materials Cloud, where data can be uploaded, visualized, analyzed and shared. Thanks to AiiDA and the Materials Cloud, research data management and stewardship is streamlined, providing an integrated platform for FAIR data sharing, where data is:

  • Findable by DOIs and standardized metadata on the Materials Cloud Archive

  • Accessible via the Materials Cloud interface, giving graphical and interactive access both to curated data and to the raw AiiDA graphs to inspect the provenance

  • Interoperable thanks to the adoption of code-agnostic data formats for the relevant data structures that can be reused across different code plugins

  • Reusable by encouraging the use of open licenses when data is uploaded on the Archive, but most importantly reproducible thanks to the strong reproducibility guarantees of the provenance graph.

In addition to this, Materials Cloud acts also as a simulation platform, providing tools to enable running new simulations both on premises and in the cloud via the AiiDA lab platform, the Quantum Mobile and a number of online interactive tools.

Our collaborations in data management and storage

MaX advances in data management are supported by collaborations with renowned projects and organizations. Key among these are MARVEL, a long-term Swiss effort coordinated by EPFL, and supported by a number of other projects: H2020 MarketPlace, H2020 Intersect, the swissuniversities P-5 “Materials Cloud” project, the EPFL Open Science Fund through the OSSCAR project, the European Research Council (ERC), the Swiss Platform for Advanced Scientific Computing (PASC), the H2020 NFFA, and the H2020 EMMC.