Skip to content
10 differentiable physical simulators built with Taichi differentiable programming (DiffTaichi, ICLR 2020)
Branch: master
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
examples Fixed smoke_taichi_gpu with taichi v0.3.21 Jan 13, 2020
.gitignore upgrade taichi Jan 3, 2020 Update Dec 31, 2019
requirements.txt upgrade taichi Jan 3, 2020

DiffTaichi: Differentiable Programming for Physical Simulation (ICLR 2020)

Yuanming Hu, Luke Anderson, Tzu-Mao Li, Qi Sun, Nathan Carr, Jonathan Ragan-Kelley, Frédo Durand

[Paper] [Video] (with instructions to reproduce every demo)

Differentiable programming in Taichi allows you to optimize neural network controllers efficiently with brute-force gradient descent, instead of using reinforcement learning.

The DiffTaichi differentiable programming framework is now officially part of Taichi. This repo only contains examples.

DiffTaichi significantly boosts the performance and productivity of differentiable physical simulators. For example, the differentiable elastic object simulator (ChainQueen) in DiffTaichi is 188x faster than an implementation in TensorFlow. The DiffTaichi version also runs as fast as the CUDA implementation, with the code being 4.2x shorter.

Most of the 10 differentiable simulators can be implemented within 2-3 hours (if you don't consider the coding speed of Yuanming Hu as an outlier...)

Questions regarding the simulators/autodiff compiler go to Yuanming Hu (yuanming at or Issues.

How to run

Step 1: Install Taichi with pip: (Most examples do not need a GPU to run.)

# CPU only. No GPU/CUDA needed. (Linux, OS X and Windows)
python3 -m pip install taichi-nightly

# With GPU (CUDA 10.0) support (Linux only)
python3 -m pip install taichi-nightly-cuda-10-0

# With GPU (CUDA 10.1) support (Linux only)
python3 -m pip install taichi-nightly-cuda-10-1

Step 2: Run example scripts in the examples folder: (Please wait for all GIFs to load :-)

Differentiable Elastic Object Simulator [python3]

Gardient descent iteration 0 and gradient descent iteration 80:

Differentiable 3D Elastic Object Simulator [python3]

Gradient descent iteration 40:

Differentiable 3D Fliud Simulator [python3]

Gradient descent iteration 450:

Differentiable Height Field Water Simulator [python3]

Gradient descent iteration 180:

Differentiable (Adversarial) Water Renderer [python3]

Differentiable water simulation + differentiable water rendering + (differentiable) CNN

Optimization goal: find an initial water height field, so that after simulation and shading, VGG16 thinks the squirrel image is a goldfish. Input image: VGG16=fox squirrel (42.21%)

Left: center activation . Right: An activation that fools VGG (VGG16=goldfish (99.91%))

Differentiable Rigid Body Simulator [python3 [1/2] train]

2048 time steps. Gardient descent iteration 20:

Differentiable Mass-Spring Simulator [python3 [1/2/3] train]

682 time steps. Gardient descent iteration 20:

Differentiable Billiard Simulator [python3]

Gardient descent iteration 0 and gradient descent iteration 100:

See the video for the remaining two simulators.


  title={DiffTaichi: Differentiable Programming for Physical Simulation},
  author={Hu, Yuanming and Anderson, Luke and Li, Tzu-Mao and Sun, Qi and Carr, Nathan and Ragan-Kelley, Jonathan and Durand, Fr{\'e}do},
  journal={arXiv preprint arXiv:1910.00935},
You can’t perform that action at this time.