15 February 2021
AiiDAlab, an intuitive infrastructure for computational workflows
Cloud services and platforms have revolutionized all the fields in which they have found applicability. Notably, in the model of Platform as a Service, users can execute their tasks directly from the web browser. When it comes to computational science, these software should be developed in order to create a common ground for all the interested scientists with different backgrounds, both experimental and theoretical, and different levels of expertise.
AiiDAlab has been conceived and developed as an infrastructure for developing, executing and sharing computational workflows in an intuitive and simple way. The effort, which led the AiiDA team to win PRACE-funded resources, was realized between Prof. Nicola Marzari’s Laboratory of Theory and Simulation of Materials and Prof. Berend Smit’s Laboratory of Molecular Simulation at EPFL, involving also the nanotech@surfaces laboratory at Empa.
The novelty of AiiDAlab includes several relevant sides: (i) collaboration, as it provides an environment where people with different expertise can collaborate on a common computational project; (ii) saving time and avoiding mistakes, as using AiiDAlab one can directly exchange the simulation results in an appropriate format boosting know-how transfer; (iii) modularity, as it is possible to easily build powerful tools by combining compact well-defined components, (iv) reproducibility, as all simulations are run via AiiDA, that keeps track of the full provenance of all simulations in the form of a directed graph, making all simulations reproducible.
Snapshots of the Scanning Probe Microscopy (SPM) AiiDAlab app. (a) Interface for the submission of a workflow to compute Scanning Tunnelling Microscopy (STM) images. (b) Interface to analyze the results of the STM calculations.
MaX centre supported the development of AiiDAlab and of the underlying engines and workflows (AiiDA, and workflows for QuantumESPRESSO, CP2K,…), and promoted the EU networking (e.g. the EOSC Marketplace where AiiDAlab is registered).
DOI: https://doi.org/10.1016/j.commatsci.2020.110165
A. V. Yakutovich et al., AiiDAlab – an ecosystem for developing, executing, and sharing scientific workflows, Comp. Mat. Sci. 188, 110165 (2021).