Repository navigation
Conversation
Adds support for the AMD ROCm imaging stack as a dedicated pip extra:
- **pyproject.toml**: new `rocm = ["amd-hipcim>=26.6.0; platform_system == 'Linux'"]`
extra. `amd-hipcim` ships the `cucim` Python namespace so all existing
`optional_import("cucim…")` call sites work on ROCm without code changes.
Omitted from `all` (requires AMD's package index, not PyPI). The `>=26.6.0`
floor skips the public PyPI placeholder stub (0.0.2).
- **docs/source/installation.md**: new `rocm` section documenting the extra,
the AMD index requirement, the cp312/x86_64-only wheel availability, and
the `[all]` incompatibility on ROCm. Fixes the developer-setup recipe
(`print_dependencies.py all testing` instead of `\*`, which now also
includes the index-restricted `rocm` extra).
- **monai/config/deviceconfig.py**, **monai/config/check_env.py**: report
`HIP version` instead of `CUDA version: None` on ROCm PyTorch builds, where
`torch.version.cuda` is `None` and the toolkit version is `torch.version.hip`.
No existing consumer of the `get_gpu_info()` dict keys is affected.
- **Dockerfile.rocm**: align `amd-hipcim>=26.6.0` floor with pyproject.toml.
- **tests/config/test_print_info.py**: two mock-based tests covering the
HIP/CUDA version branch in `get_gpu_info()` under both ROCm and CUDA builds.
Signed-off-by: Patel, Nilaykumar K <NilaykumarKantibhai.Patel@amd.com>
prateek9623
changed the title
prateek9623
force-pushed
the
rocm-extra-and-docs
branch
from
October 8, 2026 10:44
dae09ca to
effbe1c
Compare
3 of 7 tasks
prateek9623
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds support for the AMD ROCm imaging stack as a dedicated pip extra:
pyproject.toml: new
rocm = ["amd-hipcim>=26.6.0; platform_system == 'Linux'"]extra.amd-hipcimships thecucimPython namespace so all existingoptional_import("cucim…")call sites work on ROCm without code changes. Omitted fromall(requires AMD's package index, not PyPI). The>=26.6.0floor skips the public PyPI placeholder stub (0.0.2).docs/source/installation.md: new
rocmsection documenting the extra, the AMD index requirement, the cp312/x86_64-only wheel availability, and the[all]incompatibility on ROCm. Fixes the developer-setup recipe (print_dependencies.py all testinginstead of\*, which now also includes the index-restrictedrocmextra).monai/config/deviceconfig.py, monai/config/check_env.py: report
HIP versioninstead ofCUDA version: Noneon ROCm PyTorch builds, wheretorch.version.cudaisNoneand the toolkit version istorch.version.hip. No existing consumer of theget_gpu_info()dict keys is affected.Dockerfile.rocm: align
amd-hipcim>=26.6.0floor with pyproject.toml.tests/config/test_print_info.py: two mock-based tests covering the HIP/CUDA version branch in
get_gpu_info()under both ROCm and CUDA builds.