AiiDA

AiiDA provides infrastructure to automate, execute, and record computational workflows and full data provenance for computational science, enabling reproducible high-throughput simulations and standardized deposition of computed materials properties.

Key Features:

• Data Provenance Management: Records full data provenance including inputs, outputs, and intermediate states to produce traversable provenance graphs for reproducibility and analytics.
• Workflow Automation: Implements a workflow engine that automates complex computational workflows with error handling and conditional execution paths.
• Scalability and Performance: Supports high-throughput workloads, capable of managing tens of thousands of processes per hour and designed to operate on next-generation exascale supercomputers.
• Interoperability and Integration: Uses a flexible plugin model to integrate with external simulation software and HPC systems and provides a plugin registry for sharing extensions.
• Metadata Standardization and Interoperability: Leverages standardized metadata protocols to support FAIR principles and integrates with the TCOD database to tag and deposit computed materials properties along with their provenance graphs.

Scientific Applications:

• Computational materials science: Automates the deposition of theoretical structures and computed properties into TCOD and manages extensive provenance information to enhance reproducibility and data sharing.

Methodology:

Stores and queries full data provenance in a relational database; provides a workflow language/engine with automation features including error handling and conditional execution; exposes a plugin interface for integration with simulation codes and HPC systems; supports automated tagging and deposition of computed properties to TCOD with accompanying provenance graphs.