Skip to content

Build vLLM nightly wheels for CUDA 13.0 #163239

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

Open
wants to merge 6 commits into
base: main
Choose a base branch
from
Open

Build vLLM nightly wheels for CUDA 13.0 #163239

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
82df8a8
Build vLLM nightly wheels for CUDA 13.0
huydhn Sep 18, 2025
1c13887
Try to lower max_jobs
huydhn Sep 18, 2025
fc0614d
Merge branch 'main' into vllm-wheel-cuda13
huydhn Sep 18, 2025
8b07999
Merge branch 'main' into vllm-wheel-cuda13
huydhn Oct 12, 2025
aa76098
Leave cu130 aarch64 build for later
huydhn Oct 12, 2025
72e1bea
Add more context to comment
huydhn Oct 12, 2025
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
2 changes: 1 addition & 1 deletion .github/actions/build-external-packages/action.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,7 +65,7 @@ runs:
cd .ci/lumen_cli
python3 -m pip install -e .
)
MAX_JOBS="$(nproc --ignore=6)"
MAX_JOBS="$(nproc --ignore=10)"
export MAX_JOBS
# Split the comma-separated list and build each target
Expand Down
30 changes: 25 additions & 5 deletions .github/workflows/build-vllm-wheel.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,9 +27,6 @@ jobs:
fail-fast: false
matrix:
python-version: [ '3.12' ]
# TODO (huydhn): Add cu130 after https://github.com/vllm-project/vllm/issues/24464 is resolved
platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ]
device: [ 'cu128', 'cu129' ]
include:
- platform: manylinux_2_28_x86_64
device: cu128
Expand All @@ -39,6 +36,10 @@ jobs:
device: cu129
manylinux-image: 'pytorch/manylinux2_28-builder:cuda12.9'
runner: linux.12xlarge.memory
- platform: manylinux_2_28_x86_64
device: cu130
manylinux-image: 'pytorch/manylinux2_28-builder:cuda13.0'
runner: linux.12xlarge.memory
- platform: manylinux_2_28_aarch64
device: cu128
manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.8'
Expand All @@ -47,6 +48,12 @@ jobs:
device: cu129
manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda12.9'
runner: linux.arm64.r7g.12xlarge.memory
# TODO (huydhn): Add cu130 aarch64 once PyTorch is on 2.9+ and
# xformers is update to support 13.0
# - platform: manylinux_2_28_aarch64
# device: cu130
# manylinux-image: 'pytorch/manylinuxaarch64-builder:cuda13.0'
# runner: linux.arm64.r7g.12xlarge.memory
name: "Build ${{ matrix.device }} vLLM wheel on ${{ matrix.platform }}"
runs-on: ${{ matrix.runner }}
timeout-minutes: 480
Expand Down Expand Up @@ -168,8 +175,21 @@ jobs:
strategy:
fail-fast: false
matrix:
platform: [ 'manylinux_2_28_x86_64', 'manylinux_2_28_aarch64' ]
device: [ 'cu128', 'cu129' ]
include:
- platform: manylinux_2_28_x86_64
device: cu128
- platform: manylinux_2_28_x86_64
device: cu129
- platform: manylinux_2_28_x86_64
device: cu130
- platform: manylinux_2_28_aarch64
device: cu128
- platform: manylinux_2_28_aarch64
device: cu129
# TODO (huydhn): Add cu130 aarch64 once PyTorch is on 2.9+ and
# xformers is update to support 13.0
# - platform: manylinux_2_28_aarch64
# device: cu130
env:
PLATFORM: ${{ matrix.platform }}
BUILD_DEVICE: ${{ matrix.device }}
Expand Down
Loading