taichi-fluid icon indicating copy to clipboard operation
taichi-fluid copied to clipboard

A collection of CFD related resources for Taichi developers.

banner

Github-widget Youtube-widget Twitter-widget Zhihu-widget Bilibili-widget

A collection of CFD related resources for Taichi developers.

Taichi is an open-source, imperative, parallel programming language for high-performance numerical computation. It is embedded in Python and uses just-in-time (JIT) compiler frameworks (e.g. LLVM) to offload compute-intensive Python code to the native GPU or CPU instructions.

Taichi provides several advantages over existing computational fluid dynamics tools:

  • Performance: Through the @ti.kernel decorator, Taichi's JIT compiler automatically compiles your Python functions into efficient GPU or CPU machine code for parallel execution.
  • Portability: Write your code once and run it everywhere. You can easily reproduce other's work without worrying about environment setup.
  • Simplicity: Data structure detached from computational logic. Tuning performance with only a few lines of change.

:eyes: All fluid simulation projects in Taichi are driven by and for the community. Please feel free to open up an issue to recommend any awesome fluid project you see or build.

Contents

  • Installation of Taichi
  • Learning Resources
  • CFD Projects in Taichi

Installation of Taichi

You can easily install Taichi with Python's package installer pip:

pip install taichi

After you have installed Taichi, running a Taichi program is as simple as:

python your_program.py

More information can be found in Taichi's Documentation.

Learning Resources

CFD Projects in Taichi

Incompressible Fluids

Compressible Fluids

Particle Based Methods

Computational Graphics

Fluid Engine

Artificial Intelligence