The NoMaD Repository, http://nomad-repository.eu/, offers to store and share input and output files of all popular electronic-structure codes. The uploaded files are saved for at least 10 years. Thus, uploaders and their groups can get back what they have done in the past, and also others can use the data for their own scientific work.

The NoMaD Repository offers DOIs (digital object identifiers) free of charge which makes the scientific data are citeable. It is the only repository in materials science recommended by Nature Scientific Data.

There are many more advantages for Psi-k practitioners. See http://nomad-repository.eu/ and
a two-minute movie at youtube and youku: https://www.youtube.com/watch?v=L-nmRSH4NQM, http://v.youku.com/v_show/id_XMTM0NDA0NDIxMg==.html, and "Making the most of materials computations", by Kristian S. Thygesen, Karsten W. Jacobsen: http://science.sciencemag.org/content/354/6309/180

Currently, the NOMAD Repository contains more than 3 mio entries, which corresponds to more than 18 mio total energy calculations and several billion CPU core hours spent on various high-performance computers worldwide. (The NOMAD Repository also contains the files from OQMD and AFLOWLIB.)

Recently, the repository data (only when declared open access) have been parsed and normalized by the NOMAD Laboratory, https://nomad-coe.eu/. This European Center of Excellence is an effort by 8 scientific PIs – most active in Psi-k – and 4 HPC centres. Its Archive contains normalized, (largely) code-independent data that are, for instance, being used for novel data analytics. For first public examples of machine learning, compressed sensing, subgroup discovery, and more see: https://analytics-toolkit.nomad-coe.eu/. More examples will come soon.

Best wishes,
Claudia Draxl and Matthias Scheffler