taichimd
所属分类:GPU/显卡
开发工具:Python
文件大小:27916KB
下载次数:0
上传日期:2022-05-26 21:49:17
上 传 者:
sh-1993
说明: 使用太极编程语言的交互式GPU加速分子动力学
(Interactive, GPU-accelerated Molecular Dynamics using the Taichi programming language)
文件列表:
LICENSE (35149, 2022-05-27)
benckmarks (0, 2022-05-27)
benckmarks\calculate_hessian.pdf (63738, 2022-05-27)
benckmarks\fig1.png (393983, 2022-05-27)
benckmarks\fig2.png (104219, 2022-05-27)
benckmarks\fig3.png (685618, 2022-05-27)
benckmarks\fig4.png (613372, 2022-05-27)
benckmarks\integrator_benchmark.py (2195, 2022-05-27)
benckmarks\movie.gif (5774924, 2022-05-27)
benckmarks\plots.ipynb (5336, 2022-05-27)
benckmarks\result_chain.json (201174, 2022-05-27)
benckmarks\result_lj.json (331686, 2022-05-27)
mpm28.py (1120, 2022-05-27)
mpm48.py (2698, 2022-05-27)
mpm99.gif (3222353, 2022-05-27)
preview.gif (9405086, 2022-05-27)
propane.gif (8181656, 2022-05-27)
run_examples.py (2654, 2022-05-27)
taichimd (0, 2022-05-27)
taichimd\__init__.py (181, 2022-05-27)
taichimd\analyzer.py (1225, 2022-05-27)
taichimd\common.py (1214, 2022-05-27)
taichimd\consts.py (397, 2022-05-27)
taichimd\examples.py (3722, 2022-05-27)
taichimd\forcefield.py (10710, 2022-05-27)
taichimd\graphics.py (17291, 2022-05-27)
taichimd\grid.py (9946, 2022-05-27)
taichimd\integrator.py (9430, 2022-05-27)
taichimd\interaction.py (6096, 2022-05-27)
taichimd\molecule.py (801, 2022-05-27)
taichimd\system.py (12246, 2022-05-27)
taichimd\t3mini.py (12246, 2022-05-27)
taichimd\ui.py (8689, 2022-05-27)
# TaichiMD
Interactive, GPU-accelerated Molecular (& Macroscopic) Dynamics using the Taichi programming language
(05/26/2022) TaichiMD is now made compatible with taichi v1.0.
![](https://github.com/victoriacity/taichimd/blob/master/propane.gif)![](https://github.com/victoriacity/taichimd/blob/master/mpm99.gif)
## Goals
* To extend capabilities of the Taichi programming language in computer graphics to molecular simulation education and research
* To achieve interactive, real-time particle-based simulations accelerated by GPUs
* To provide a platform for rapid implementation of novel simulation algorithms and machine-learned simulations
## Examples
### Molecular simulation examples
`python run_examples.py [-h] example ensemble`
positional arguments:
```
example [lj | mixlj | biglj | ho | chain | pr]
lj: Lenneard-Jones system with 4096 molecules, in reduced units;
biglj: Lenneard-Jones system with 0.26 million molecules, in reduced units;
mixlj: 3-component Lenneard-Jones mixture with 6000 molecules, in reduced units;
ho: Harmonic oscillator around the center of the simulation box;
chain: 5 harmonic-bond chain molecules with 100 atoms each, bond bending and torsion not included, in real
units
pr: 512 propane molecules using the TraPPE-UA force field with a harmonic bond
stretching potential at 423 K in a 50*50*50 angstrom box
ensemble [NVE | NVT]
NVE ensemble with Verlet integration or NVT ensemble with Nose-Hoover thermostat
```
optional arguments: `-h, --help show help message and exit`
### Macroscopic simulation examples
`mpm28.py`: The MPM88 Taichi example (fluid simulation) in 3 dimensions.
`mpm48.py`: The MPM99 Taichi example (fluid, jelly and snow) in 3 dimensions.
## Graphics
TaichiMD ships with a miniature static version of [`taichi_three`](https://github.com/victoriacity/taichimd/blob/master/https://github.com/taichi-dev/taichi_three) for rendering graphics, which requires the `taichi_glsl` package. If the renderer was not imported correctly, the example simulations will use the Taichi GUI canvas for graphics.
## Possible extensions
### Microscopic forces
* Add support for torsional potentials as cosine series
* Add Coulomb forces between charged atoms
### Acceleration algorithms
* Implement Ewald summation and/or fast multipole method for long-range forces (Coulomb)
### Macroscopic forces
* Implement gravity and wall boundaries
* Incorporate macroscopic particle simulation algorithms (SPH, etc.)
* Incorporate agent-based simulation algorithms (crowd simulation, particle swarm) and optimization-based integrators
近期下载者:
相关文件:
收藏者: