CyRSoXS: GPU enabled RSoXS simulation

• CyRSoXS is a GPU-accelerated codebase that calculates resonant X-ray scattering in the Born Approximation. It takes a voxel-based model and optical constants as input, and returns simulated X-ray scattering patterns. These models can be derived from real space measurements (AFM, TEM, 4DSTEM), or synthetically generated using procedural morphology generators. Currently, it supports both 32 bit and 64 bit float data types.

• The code can be executed through Python (enabled through Pybind11) or directly through the executable.

Version : 1.2.0.0

The developer version of the CyRSoXS code is available at https://bitbucket.org/baskargroup/cy-rsoxs. All stable future releases will be merged with the current usnistgov repository.

Dependencies

Hardware

• CUDA GPU with Compute Capability >=7
  • CUDA is a parallel computing platform and programming model creaded by NVIDIA

Required Dependencies for building from source

• A C++ compiler with C++17 support is required.
• gcc compatible with your CUDA toolkit
• CUDA Toolkit with nvcc
• HDF5 (optional at the system level for Python/uv builds: CMake can compile a bundled static HDF5 when CYRSOXS_FETCH_HDF5 is enabled, which is the default for SKBUILD installs)
  • Hierarchical Data Format (version 5) is a set of file formats designed to store and organize large amounts of data. CyRSoXS uses it to store models and simulated scattering patterns
• OpenMP
• libconfig==1.7.2

Additional dependencies for building with Pybind

• Python >= 3.10 (for the Pybind11 extension build via uv or pip)

Optional Dependencies

• Doxygen - for building the documentation.
• Docker - if you want to run CyRSoXS in a Docker container.

Installation

• conda-forge (recommended pre-built GPU binaries): conda install cyrsoxs -c conda-forge
• uv / pip from source: this repository ships a pyproject.toml using scikit-build-core. With CUDA and OpenMP on the system, run uv sync --all-groups (or uv sync --all-extras) or uv build. HDF5 is built automatically when not installed system-wide (CYRSOXS_FETCH_HDF5, on by default for Python builds). Tag-triggered PyPI uploads currently publish sdists only (binary manylinux CUDA wheels would need a dedicated cibuildwheel image policy); conda-forge remains the primary channel for curated GPU stacks.
• CMake-only flows for the CLI, libconfig, and detailed HDF5 setup are in docs/INSTALL.md.

Publishing to PyPI on annotated tags is handled by .github/workflows/release.yml using trusted publishing; configure the cyrsoxs project on PyPI to trust this GitHub repository and workflow before relying on automated uploads.

Simulation Componenets

Running CyRSoXS

Visualization

Known Issues

Documentation

CyRSoXS documentation is available at https://cyrsoxs.readthedocs.io, and instructions to build the documentation are in the Installation Instructions.

Running CyRSoXS with Docker

The Docker image for CyRSoXS is available at Docker Hub. This comes installed with Nvidia toolkit and other dependencies. See docs/DOCKER.md for detailed instructions on how to build CyRSoXS with the Docker image.

Contributors

This software was developed at Iowa State University in collaboration with NIST. The Iowa State team provided expertise in high performance computing, and the NIST team provided scientific expertise on the RSoXS technique.

Iowa State University

• Kumar Saurabh, Github, ORCiD, Email
• Adarsh Krishnamurthy, Webpage, ORCiD, Email
• Baskar Ganapathysubramanian, Webpage, ORCiD, Email

NIST

• Eliot Gann, Github, ORCiD, Email
• Dean M. DeLongchamp, Webpage, ORCiD, Email
• Peter J. Dudenas, Github, ORCiD, Email
• Tyler B. Martin, Github, ORCiD, Email
• Peter Beaucage, Github, ORCiD, Email

Acknowledgement

We thank the Office of Naval Research Multidisciplinary University Research Initiative (ONR MURI) Center for Self-Assembled Organic Electronics for providing support for this work.

Frequently Asked Questions

See docs/FAQ.md for some of the known issues or Frequently asked questions.

Contact

Questions and comments should be sent to Dr. Baskar Ganapathysubramanian at baskarg@iastate.edu, Dr. Adarsh Krishnamurthy adarsh@iastate.edu, or Dr. Dean DeLongchamp at dean.delongchamp@nist.gov

NIST Disclaimer

Any mention of commercial products is for information only; it does not imply recommendation or endorsement by NIST.