DiffTaichi: Differentiable Programming for Physical Simulation

10/01/2019
by   Yuanming Hu, et al.
29

We study the problem of learning and optimizing through physical simulations via differentiable programming. We present DiffTaichi, a new differentiable programming language tailored for building high-performance differentiable physical simulations. We demonstrate the performance and productivity of our language in gradient-based learning and optimization tasks on 10 different physical simulators. For example, a differentiable elastic object simulator written in our language is 4.2x faster than the hand-engineered CUDA version yet runs as fast, and is 188x faster than TensorFlow. Using our differentiable programs, neural network controllers are typically optimized within only tens of iterations. Finally, we share the lessons learned from our experience developing these simulators, that is, differentiating physical simulators does not always yield useful gradients of the physical system being simulated. We systematically study the underlying reasons and propose solutions to improve gradient quality.

READ FULL TEXT

page 6

page 7

page 17

research
12/15/2020

AsyncTaichi: Whole-Program Optimizations for Megakernel Sparse Computation and Differentiable Programming

We present a whole-program optimization framework for the Taichi program...
research
09/14/2021

Differentiable Physics: A Position Piece

Differentiable physics provides a new approach for modeling and understa...
research
08/22/2018

k-meansNet: When k-means Meets Differentiable Programming

In this paper, we study how to make clustering benefiting from different...
research
10/05/2020

ξ-torch: differentiable scientific computing library

Physics-informed learning has shown to have a better generalization than...
research
04/07/2021

PlasticineLab: A Soft-Body Manipulation Benchmark with Differentiable Physics

Simulated virtual environments serve as one of the main driving forces b...
research
12/07/2020

Using Differentiable Programming for Flexible Statistical Modeling

Differentiable programming has recently received much interest as a para...
research
03/18/2022

Half-Inverse Gradients for Physical Deep Learning

Recent works in deep learning have shown that integrating differentiable...

Please sign up or login with your details

Forgot password? Click here to reset