Pillar A: Computational materials science - The NOMAD Laboratory
Computational materials science develops and uses highly sophisticated computer programs to investigate, characterize, and predict materials at the atomic level. It thus provides insight into materials properties and functions and the design and development of new materials that meet specific requirements. Providing a forefront infrastructure for computed materials data, that enables the explotation of this significant raw material, is the aim of Pillar A.
The NOMAD Repository accepts, more precisely requests, input and output files of all important codes. It is the largest repository of input and output files worldwide and keeps data for at least 10 years. Open access can be delayed by up to three years. DOIs are provided to make the data citable. See also here.
The NOMAD Repository is the world’s largest collection of computational materials science data, hosting raw data contributed by the community of electronic-structure theory. NOMAD Repository is not restricted to a single code or a closed group of researchers. Already now, more than 30 of the dominant ab initio codes are supported, and more codes will be added on-demand, in collaboration with the corresponding code developers. In the NOMAD Repository, the full input and output files of calculations are stored. NOMAD R also contains the data of the most important computational materials databases worldwide. NOMAD issues DOIs which make the uploaded data citable, a feature that NOMAD offers to the other pillars.
This image shows the number of uploaded open-access total-energy calculations in the NOMAD Repository and Archive as of March 15, 2018. Codes with more than 80 uploads are shown. The stamp “Supported by NOMAD” can be found at web pages of most ab initio computer codes. Currently, the NOMAD Repository and Archive contain > 50 million open-access total-energy calculations.
The NOMAD Archive consists of the open-access data from the NOMAD Repository in a normalized form, i.e. converted regarding units, formats, etc. This ensures that data from different sources can be compared and collectively operated upon. It relies on the meta information and parsing infrastructure developed by the NOMAD Laboratory CoE. The NOMAD Archive also sets the stage for big-data analytics. Thus it forms the basis for collaborations with the NOMAD data-analytics team.
The NOMAD Encyclopedia is a publicly available, web-based infrastructure that provides a materials-oriented view on the computational materials data of the NOMAD data collection. All properties of a given material that have been computed are accessible through a user-friendly graphical user interface (GUI). It thus provides a comprehensive picture of the multitude of materials and their properties that are available in the NOMAD Archive. NOMAD Encyclopedia makes use of advanced visualization techniques, comprising remote visualization and virtual-reality tools, and provides an API to its data.
The NOMAD Encyclopedia, as the face of NOMAD and thus of this pillar of FAIR-DI, will serve as an umbrella, providing a link to bio-molecular simulations (Pillar B) and experimental materials science (Pillar C). It also supports engagement with the academic and industrial communities.
Within FAIR-DI, Pillar A will take care of the following tasks:
- Maintenance and extensions of upload procedures
- Maintenance and extensions of the parsing infrastructure and parsers
- Maintenance and advancement of the NOMAD metadata
- Maintenance and extensions of APIs
- Curation of raw and normalized data
- Provision of PIDs and DOIs
- Enhancement of tools
- Hardware updates, including possible porting of the software stack to new hardware
- User support