AiiDA

Official Resources

• Homepage: https://www.aiida.net/
• Documentation: https://aiida.readthedocs.io/
• Source Repository: https://github.com/aiidateam/aiida-core
• License: MIT License

Overview

AiiDA is a sophisticated open-source workflow management system designed for high-throughput computational science. It focuses on reproducibility, provenance tracking, and automation. AiiDA manages the entire lifecycle of a calculation: preparing inputs, submitting jobs to supercomputers (via SLURM, PBS, etc.), parsing outputs, and storing the entire provenance graph (inputs, outputs, and the code that produced them) in a database.

Scientific domain: Workflow management, high-throughput computing, data provenance
Target user community: Computational materials scientists, chemists, physicists

Capabilities (CRITICAL)

• Provenance Tracking: Automatically records the full history of every calculation (inputs, codes, outputs) in a directed acyclic graph (DAG).
• Workflow Engine: Robust engine for running complex, interdependent calculations (WorkChains) with error handling.
• Plugin System: Massive ecosystem of plugins for DFT codes (VASP, QE, CP2K, etc.), schedulers, and transport.
• HPC Interface: Abstraction layer for interacting with remote clusters and job schedulers.
• Database: efficient storage and query of calculation data (PostgreSQL backend).
• Querying: Powerful QueryBuilder for searching the provenance graph.

Sources: AiiDA website, Comp. Mater. Sci. 111, 218 (2016)

Inputs & Outputs

• Input formats: Python script defining Process or WorkChain inputs (Nodes).
• Output data types: Database entries (Nodes), files (retrieved from remote), provenance graph.

Interfaces & Ecosystem

• Plugins: aiida-quantumespresso, aiida-vasp, aiida-cp2k, aiida-siesta, aiida-wannier90, etc.
• Materials Cloud: AiiDA archives can be published directly to Materials Cloud.
• AiiDA Lab: Web-based interface (Jupyter) for running AiiDA workflows.

Workflow and Usage

1. Setup: Configure computer (remote cluster) and code (e.g., pw.x).
2. Launch: Submit a calculation or workflow via Python script or shell (verdi run script.py).

   from aiida.engine import submit
   from aiida.plugins import CalculationFactory
   PwCalculation = CalculationFactory('quantumespresso.pw')
   submit(PwCalculation, **inputs)
   

3. Monitor: Check status with verdi process list.
4. Analyze: Query results with QueryBuilder or inspect nodes.

Performance Characteristics

• High throughput: Can manage thousands of concurrent calculations.
• Daemon-based: Runs in the background, managing job submission and retrieval.
• Database performance depends on size and tuning (PostgreSQL).

Application Areas

• High-throughput materials screening
• Database generation (e.g., Materials Cloud)
• Complex multi-step simulations (e.g., phonon dispersions, GW)
• Reproducible research studies

Community and Support

• Developed by EPFL (Theos group) and wider community
• Large, active ecosystem
• Regular tutorials and hackathons
• Discourse forum for support

Verification & Sources

Primary sources:

1. Homepage: https://www.aiida.net/
2. GitHub: https://github.com/aiidateam/aiida-core
3. Publication: G. Pizzi et al., Comp. Mater. Sci. 111, 218 (2016); S. P. Huber et al., Sci. Data 7, 300 (2020)

Confidence: VERIFIED

Verification status: ✅ VERIFIED

• Website: ACTIVE
• Documentation: COMPREHENSIVE
• Source: OPEN (GitHub)
• Development: ACTIVE (EPFL/AiiDA Team)
• Applications: Workflow management, provenance, high-throughput