Commit 057092bb authored by Ryan Gutenkunst's avatar Ryan Gutenkunst
Browse files

GPU installation instructions

parent bf038e35
......@@ -15,6 +15,13 @@ It is also recommended that you install
The easiest way to obtain all these dependencies to install the [Anaconda Python Distribution](https://www.anaconda.com/distribution/). And the easiest way to install dadi is then via `conda`, using the command `conda install -c conda-forge dadi`. dadi can also be installed via `pip`, using the command `python3 -m pip install dadi`.
### GPU computing
dadi can be sped up substantially by running on a CUDA-enabled Nvidia GPU.
To enable this functionality, you will need to install the [CUDA Toolkit](https://developer.nvidia.com/cuda-downloads).
After install the CUDA Toolkit, you will then need to install PyCUDA and scikit-cuda.
Both of these can be installed from the Python Package Index using pip, `python3 -m pip install pycuda` and `python3 -m pip install scikit-cuda`.
### Installing from source
dadi can be easily installed from [source code](https://bitbucket.org/gutenkunstlab/dadi/src/master/), as long as you have an appropriate C compiler installed. (On OS X, you'll need to install the Developer Tools to get gcc. On Windows, you'll need to install the Microsoft Visual Studio to get C/C++ builder.) To do so, first unpack the source code tarball, `unzip dadi-<version>.zip` In the `dadi-<version>` directory, run `python setup.py install`. This will compile the C modules dadi uses and install those plus all dadi Python files in your Python installation's `site-packages` directory. A (growing) series of tests can be run in the `tests` directory, via `python run_tests.py`.
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment