atomate2

Official Resources

• Homepage: https://materialsproject.github.io/atomate2/
• Documentation: https://materialsproject.github.io/atomate2/
• Source Repository: https://github.com/materialsproject/atomate2
• License: BSD 3-Clause License

Overview

Atomate2 is the next-generation workflow library for computational materials science, succeeding Atomate. It is built on top of jobflow (instead of FireWorks directly) and Pymatgen. It offers a more modern, flexible, and easier-to-use API for defining workflows. It supports VASP, CP2K, ForceField (via LAMMPS), and other codes, with a focus on modularity and dynamic workflow generation.

Scientific domain: Workflow automation, high-throughput materials science
Target user community: VASP/CP2K users, next-gen Materials Project contributors

Capabilities (CRITICAL)

• Modern Workflow Engine: Powered by jobflow, allowing local execution (without a database) or remote execution (via FireWorks).
• Dynamic Flows: Easier creation of dynamic workflows (e.g., loops, conditionals) compared to Atomate 1.
• Codes Supported: VASP, CP2K, LAMMPS, Quantum ESPRESSO (via plugins).
• Output Handling: Better schema for output documents (JobStore).
• Interoperability: Can easily switch between different workflow runners (local, FireWorks, etc.).

Sources: Atomate2 documentation

Inputs & Outputs

• Input formats: Pymatgen objects, Jobflow Jobs/Flows
• Output data types: JSON/BSON documents (locally or in MongoDB)

Interfaces & Ecosystem

• jobflow: The underlying workflow library
• pymatgen: Core materials analysis
• FireWorks: Optional backend for remote execution
• Custodian: Error handling

Workflow and Usage

1. Define a job: job = relax_job(structure)
2. Run locally (for testing): run_locally(job)
3. Or run via FireWorks: flow = Flow([job]); flow.submit(lpad)

Performance Characteristics

• improved serialization compared to Atomate 1
• Faster workflow construction
• Flexible execution models (local vs distributed)

Application Areas

• High-throughput materials discovery
• Complex simulation pipelines
• Rapid prototyping of workflows

Community and Support

• Developed by Materials Project team (LBNL)
• Active development on GitHub
• Emerging standard for MP workflows

Verification & Sources

Primary sources:

1. Homepage: https://materialsproject.github.io/atomate2/
2. GitHub: https://github.com/materialsproject/atomate2

Confidence: VERIFIED

Verification status: ✅ VERIFIED

• Website: ACTIVE
• Documentation: COMPREHENSIVE
• Source: OPEN (GitHub)
• Development: ACTIVE (Materials Project)
• Applications: Next-gen workflows, VASP, CP2K