x3d2 is the current codename for what is to become Xcompact3d version 2.0. This project is a rewrite of Xcompact3d leading to

• Improved serial and parallel performance.

• Support for NVIDIA GPU accelerators (using CUDA Fortran).

• Improved modularity and testability.

Contents:

• Report a bug
• Internals
  • Computing the spatial contribution to the momentum equation
  • Memory management
• Contributing to x3d2
  • Overview of the contribution process
  • Required tools
  • Setting up for development
  • Building and running the unit tests
  • Contributing guidelines
  • Fortran style guide
  • Contributing documentation
• API reference