Welcome to RAMSES-GPU’s documentation!¶
- RAMSES-GPU is a general-purpose Hydrodynamics (HD) and Magneto-hydrodynamics (MHD) simulation code primarily written for astrophysics applications. It allows to run very large simulations (up to grids of size 4096^3) on the largest GPU cluster available today.
- RAMSES-GPU contains both a CPU and a GPU implementation (using NVIDIA CUDA programing model), on a uniform grid only, of the MHD algorithm that was originally developped at the Service d’Astrophysique of CEA/Saclay for the code RAMSES ([Teyssier2002], [Fromang2006]) and later implemented in HERACLES.
See also video of the GTC2014 talk
NVIDIA CUDA Teaching and Research center¶
- Parallel performance
- Gallery and Tests