Shamrock is a general-purpose HPC hydrodynamics simulation code focused on astrophysical applications. The entire codebase is written in C++17, with all accelerated components implemented using SYCL, which can be compiled directly to native CUDA, ROCm, OpenMP, OpenCL, and more. Shamrock also supports multiple GPUs and heterogeneous clusters using MPI.
Comprehensive user documentation and guides
Python bindings documentation and examples
Browse through practical examples and use cases
Source code, issues, and contributions
Complete C++ API documentation and reference
Complete Python API documentation and reference
Scale seamlessly from laptops to exascale supercomputers with native multi-GPU and heterogeneous cluster support using MPI.
Support for multiple hydrodynamics solvers including both grid-based and particle-based methods for comprehensive astrophysical simulations.
Specialized for astrophysical applications with support for SPH, AMR, and other advanced hydrodynamics methods.
All accelerated components use SYCL, compiled directly to native CUDA, ROCm, OpenMP, OpenCL, and more for maximum performance.
Built with modern C++17 standards, providing robust performance and maintainable code for scientific computing.
Follow our comprehensive quickstart guide to install Shamrock and run your first hydrodynamics simulation. The guide covers installation, device selection, and basic usage patterns.
View Quickstart GuideMultiple installation methods: from source, Spack, Homebrew, or Docker
Choose between GPU and CPU devices with our device management system
Use Shamrock as an Python interpreter, Python package, or in Jupyter notebooks
Contribute by reporting issues, suggesting features, or contributing code
Contribute to Shamrock