Skip to content

Enable building to target AMD GPUs #1731

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ndgrigorian merged 6 commits into from
Nov 24, 2024
Merged

Enable building to target AMD GPUs #1731

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
ab7cf09
Enable CMake options for building for AMD
ndgrigorian Jul 13, 2024
9fa6054
Refactor cmake for setting AMD targets and fix incorrect logic
ndgrigorian Oct 17, 2024
42f8ace
Remove use of `std::log1p` in `math_utils.hpp`
ndgrigorian Oct 17, 2024
a559d10
Change DPCTL_TARGET_AMD to DPCTL_TARGET_HIP
ndgrigorian Oct 17, 2024
ac6cc7b
Implement HIP backend
ndgrigorian Oct 24, 2024
c316380
Add documentation on building for AMD devices to dpctl installation g…
ndgrigorian Nov 22, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Add documentation on building for AMD devices to dpctl installation g… 
…uide
  • Loading branch information
@ndgrigorian
ndgrigorian committed Nov 22, 2024
commit c316380cdd2210f6d028d127b2e67d747a0487b6
18 changes: 17 additions & 1 deletion docs/doc_sources/beginners_guides/installation.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -158,12 +158,28 @@ The following plugins from CodePlay are supported:
.. _codeplay_nv_plugin: https://developer.codeplay.com/products/oneapi/nvidia/
.. _codeplay_amd_plugin: https://developer.codeplay.com/products/oneapi/amd/

Build ``dpctl`` as follows:
``dpctl`` can be built for CUDA devices as follows:

.. code-block:: bash

python scripts/build_locally.py --verbose --cmake-opts="-DDPCTL_TARGET_CUDA=ON"

And for AMD devices

.. code-block:: bash

python scripts/build_locally.py --verbose --cmake-opts="-DDPCTL_TARGET_HIP=gfx1030"

Note that the `oneAPI for AMD GPUs` plugin requires the architecture be specified and only
one architecture can be specified at a time.

It is, however, possible to build for Intel devices, CUDA devices, and an AMD device
architecture all at once:

.. code-block:: bash

python scripts/build_locally.py --verbose --cmake-opts="-DDPCTL_TARGET_CUDA=ON \
-DDPCTL_TARGET_HIP=gfx1030"

Running Examples and Tests
==========================
Expand Down
3 changes: 3 additions & 0 deletions docs/doc_sources/beginners_guides/managing_devices.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -123,6 +123,9 @@ of valid settings are:
* - ``cuda:*``
- All devices only from CUDA backend are available

* - ``hip:*``
- All devices only from HIP backend are available

* - ``level_zero:0,1``
- Two specific devices from Level-Zero backend are available

Expand Down