Skip to content

GitLab

Unverified Commit f5fc52ff authored by Peter Eastman's avatar Peter Eastman Committed by GitHub
Browse files

Updated installation instructions (#4693)

parent ac32ce76
Hide whitespace changes
Inline Side-by-side
Showing with 12 additions and 11 deletions
docs-source/usersguide/application/01_getting_started.rst
View file @ f5fc52ff
...@@ -56,12 +56,14 @@ Anaconda or Miniconda. ...@@ -56,12 +56,14 @@ Anaconda or Miniconda.
modern drivers from your vendor. modern drivers from your vendor.
* If you have an Nvidia GPU, download the latest drivers from * If you have an Nvidia GPU, download the latest drivers from
https://www.nvidia.com/Download/index.aspx. CUDA itself will be provided by https://www.nvidia.com/Download/index.aspx. CUDA itself will be installed
the :code:`cudatoolkit` package when you install :code:`openmm` in the next steps. automatically when you install :code:`openmm` in the next steps.
* If you have an AMD GPU and are using Linux or Windows, download the latest * If you have an AMD GPU and are using Linux or Windows, download the latest
version of the drivers from https://support.amd.com. On OS X, OpenCL version of the drivers from https://support.amd.com. To use the HIP
is included with the operating system and is supported on OS X 10.10.3 or platform (recommended), you also need to install HIP/ROCm by following the
later. instructions at https://rocm.docs.amd.com.
* On macOS, OpenCL is included with the operating system and is supported on
macOS 10.10.3 or later.
3. Open a command line terminal and type the following command 3. Open a command line terminal and type the following command
:: ::
...@@ -74,13 +76,12 @@ Alternatively you can request a version that is compiled for a specific CUDA ...@@ -74,13 +76,12 @@ Alternatively you can request a version that is compiled for a specific CUDA
version with the command version with the command
:: ::
conda install -c conda-forge openmm cudatoolkit=10.0 conda install -c conda-forge openmm cuda-version=12
where :code:`10.0` should be replaced with the particular CUDA version where :code:`12` should be replaced with the particular CUDA version
you want to target. We build packages for CUDA 9.2 and above on Linux, you want to target. We build packages for CUDA 11 and above. Because different
and CUDA 10.0 and above on Windows. Because different CUDA releases are CUDA releases are not binary compatible with each other, OpenMM can only work
not binary compatible with each other, OpenMM can only work with the particular with the particular CUDA version it was compiled with.
CUDA version it was compiled with.
.. note:: .. note::
......
Markdown is supported
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