Skip to content
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

refactor CUDA versions in dependencies.yaml #14733

Merged
merged 6 commits into from
Jan 11, 2024
Merged

refactor CUDA versions in dependencies.yaml #14733

Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
73414b6
refactor CUDA versions in dependencies.yaml
jameslamb Jan 10, 2024
1b306cf
remove fallback, alphabetize lists
jameslamb Jan 10, 2024
10bf3ba
add cuda runtime back to java env
jameslamb Jan 10, 2024
c4677f8
Merge branch 'branch-24.02' into rework-dependencies
jameslamb Jan 10, 2024
1b21663
empty commit to re-trigger CI
jameslamb Jan 11, 2024
cfd3edf
Merge branch 'branch-24.02' into rework-dependencies
jameslamb Jan 11, 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
69 changes: 37 additions & 32 deletions dependencies.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,8 @@ files:
- build_wheels
- build_python_common
- build_python_cudf
- cudatoolkit
- cuda
- cuda_version
- develop
- docs
- libarrow_build
Expand All @@ -29,31 +30,32 @@ files:
test_cpp:
output: none
includes:
- cudatoolkit
- test_cpp
- cuda_version
- libarrow_run
- test_cpp
test_python:
output: none
includes:
- cudatoolkit
- cuda_version
- py_version
- pyarrow_run
- test_python_common
- test_python_cudf
- test_python_dask_cudf
- pyarrow_run
test_java:
output: none
includes:
- build_all
- cuda
- cuda_version
jameslamb marked this conversation as resolved.
Show resolved Hide resolved
- libarrow_run
- cudatoolkit
- test_java
test_notebooks:
output: none
includes:
- cuda_version
- notebooks
- py_version
- notebook_cuda_version
checks:
output: none
includes:
Expand All @@ -62,7 +64,8 @@ files:
docs:
output: none
includes:
- cudatoolkit
- cuda
- cuda_version
- docs
- libarrow_run
- py_version
Expand Down Expand Up @@ -333,37 +336,51 @@ dependencies:
# Allow runtime version to float up to minor version
# Disallow pyarrow 14.0.0 due to a CVE
- pyarrow>=14.0.1,<15.0.0a0
cudatoolkit:
cuda_version:
specific:
- output_types: conda
matrices:
- matrix:
cuda: "12.*"
cuda: "11.2"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we drop 11.2 from this file? We still support 11.2 via CEC, but we don't need to be able to build envs with it any more do we?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Perhaps not really in scope for this PR. Something we can come back to later.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If we are supporting it, wouldn't we want a way to construct an environment with this CUDA version for testing, debugging, etc.?

Copy link
Contributor

@bdice bdice Jan 11, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

11.2 has been dropped from our CI matrix and Docker images. I think it’s OK to remove this but I agree it’s probably better to keep the scope narrow here. At some point we will drop CUDA 11 entirely and clean up a lot of stuff.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Something we can come back to later.

I think it should come later, yeah.

packages:
- cuda-cudart-dev
- cuda-nvrtc-dev
- cuda-nvtx-dev
- libcurand-dev
- matrix: # Fallback for CUDA 11 or no matrix
- cuda-version=11.2
- matrix:
cuda: "11.4"
packages:
- cudatoolkit
- output_types: conda
matrices:
- cuda-version=11.4
- matrix:
cuda: "11.5"
packages:
- cuda-version=11.5
- matrix:
cuda: "11.8"
packages:
- cuda-version=11.8
- matrix:
cuda: "12.0"
packages:
- cuda-version=12.0
cuda:
specific:
- output_types: conda
matrices:
- matrix:
cuda: "12.*"
packages:
- cuda-cudart-dev
- cuda-nvrtc-dev
- cuda-nvtx-dev
- libcurand-dev
- matrix:
cuda: "11.8"
packages:
- cuda-version=11.8
- cudatoolkit
- cuda-nvtx=11.8
- libcurand-dev=10.3.0.86
- libcurand=10.3.0.86
- matrix:
cuda: "11.5"
packages:
- cuda-version=11.5
- cudatoolkit
- cuda-nvtx=11.5
# Can't hard pin the version since 11.x is missing many
Expand All @@ -373,15 +390,13 @@ dependencies:
- matrix:
cuda: "11.4"
packages:
- cuda-version=11.4
- cudatoolkit
- &cudanvtx114 cuda-nvtx=11.4
- &libcurand_dev114 libcurand-dev>=10.2.5.43,<=10.2.5.120
- &libcurand114 libcurand>=10.2.5.43,<=10.2.5.120
- matrix:
cuda: "11.2"
packages:
- cuda-version=11.2
- cudatoolkit
# The NVIDIA channel doesn't publish pkgs older than 11.4 for
# these libs, so 11.2 uses 11.4 packages (the oldest
Expand Down Expand Up @@ -753,13 +768,3 @@ dependencies:
packages:
- ipython
- openpyxl
notebook_cuda_version:
specific:
- output_types: conda
matrices:
- matrix: {cuda: "12.0"}
packages:
- cuda-version=12.0
- matrix: {cuda: "11.8"}
packages:
- cuda-version=11.8
Loading