[DNM] upgrade oneapi 2026.0#321
Open
jikunshang wants to merge 4 commits into
Open
Conversation
Signed-off-by: Kunshang Ji <jikunshang95@gmail.com>
Signed-off-by: Kunshang Ji <jikunshang95@gmail.com>
Contributor
There was a problem hiding this comment.
Pull request overview
Upgrades the XPU build/CI environment to target oneAPI 2026.0 and a newer (nightly) Torch XPU stack, adjusting the Docker image, dependency pins, SYCL AOT device list, and CI workflow image tagging.
Changes:
- Update Torch and Triton-XPU dependencies to pinned nightly wheel URLs.
- Bump the XPU Docker base image to intel/deep-learning-essentials:2026.0.0 and adjust GPU runtime package installs / pip index settings.
- Update SYCL AOT device defaults and CI workflow docker image tagging.
Reviewed changes
Copilot reviewed 4 out of 5 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
| requirements.txt | Pins Torch/Triton-XPU to specific nightly wheel URLs. |
| pyproject.toml | Updates build-system Torch requirement to a direct nightly wheel URL. |
| Dockerfile.xpu | Bumps oneAPI base image to 2026.0.0 and updates package/index configuration. |
| CMakeLists.txt | Adds cri to default SYCL AOT device list. |
| .github/workflows/ut.yaml | Switches CI docker tagging and container image reference to test-213. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
76
to
78
| container: | ||
| image: localhost:5000/xpu-kernel-ci-image:latest | ||
| image: localhost:5000/xpu-kernel-ci-image:test-213 | ||
| options: --device /dev/dri -v /dev/dri/by-path:/dev/dri/by-path --privileged -v ccache:/root/.ccache -e CCACHE_DIR=/root/.ccache |
Comment on lines
5
to
7
| ARG PYTHON_VERSION=3.12 | ||
| ARG PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/xpu" | ||
| ARG PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/test/xpu" | ||
|
|
Comment on lines
29
to
37
| RUN mkdir neo && cd neo && \ | ||
| wget https://github.com/intel/intel-graphics-compiler/releases/download/v2.22.2/intel-igc-core-2_2.22.2+20121_amd64.deb && \ | ||
| wget https://github.com/intel/intel-graphics-compiler/releases/download/v2.22.2/intel-igc-opencl-2_2.22.2+20121_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.44.36015.8/intel-ocloc_25.44.36015.8-0_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.44.36015.8/intel-opencl-icd_25.44.36015.8-0_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.44.36015.8/libigdgmm12_22.8.2_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.44.36015.8/libze-intel-gpu1_25.44.36015.8-0_amd64.deb && \ | ||
| wget https://github.com/intel/intel-graphics-compiler/releases/download/v2.11.7/intel-igc-core-2_2.11.7+19146_amd64.deb && \ | ||
| wget https://github.com/intel/intel-graphics-compiler/releases/download/v2.11.7/intel-igc-opencl-2_2.11.7+19146_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.18.33578.6/intel-ocloc_25.18.33578.6-0_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.18.33578.6/intel-opencl-icd_25.18.33578.6-0_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.18.33578.6/libigdgmm12_22.7.0_amd64.deb && \ | ||
| wget https://github.com/intel/compute-runtime/releases/download/25.18.33578.6/libze-intel-gpu1_25.18.33578.6-0_amd64.deb && \ | ||
| wget https://github.com/oneapi-src/level-zero/releases/download/v1.21.9/level-zero_1.21.9+u24.04_amd64.deb && \ | ||
| dpkg -i *.deb && cd .. && rm -rf neo |
| # VLLM_XPU_AOT_DEVICES="pvc,bmg-g21-a0" export | ||
| # VLLM_XPU_XE2_AOT_DEVICES="pvc,bmg-g31-a0" | ||
| set(AOT_DEVICES "pvc,bmg,bmg-g21-a0,bmg-g31-a0") | ||
| set(AOT_DEVICES "pvc,bmg,bmg-g21-a0,bmg-g31-a0,cri") |
Comment on lines
8
to
10
| "setuptools-scm>=8.0", | ||
| "torch == 2.11.0+xpu", | ||
| "torch @ https://download-r2.pytorch.org/whl/nightly/xpu/torch-2.13.0.dev20260502%2Bxpu-cp312-cp312-linux_x86_64.whl", | ||
| "wheel", |
Comment on lines
+46
to
+48
| docker build -t xpu-kernel-ci-image:test-213 -f Dockerfile.xpu . | ||
| docker tag xpu-kernel-ci-image:test-213 ${{ env.REGISTRY }}/xpu-kernel-ci-image:test-213 | ||
| docker push ${{ env.REGISTRY }}/xpu-kernel-ci-image:test-213 |
Signed-off-by: Kunshang Ji <jikunshang95@gmail.com>
Collaborator
Author
|
pending on torch upgrade to 2026.0 toolchain, then we can use latest nightly torch for further test. pytorch/pytorch#182003 while build seems pass now. |
zufangzhu
reviewed
May 6, 2026
| "setuptools>=77.0.3,<80.0.0", | ||
| "setuptools-scm>=8.0", | ||
| "torch == 2.11.0+xpu", | ||
| "torch @ https://download-r2.pytorch.org/whl/nightly/xpu/torch-2.13.0.dev20260502%2Bxpu-cp312-cp312-linux_x86_64.whl", |
Collaborator
There was a problem hiding this comment.
Why skip 2.12 and upgrade directly to 2.13...
Collaborator
There was a problem hiding this comment.
Is it because only this version supports the new OneAPI?
Collaborator
Author
There was a problem hiding this comment.
torch 2.12 will use 2025.3
torch 2.13 will use 2026.0, but it's WIP, not merged yet.
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
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.
Essential Elements of an Effective PR Description Checklist
supported_models.mdandexamplesfor a new model.PLEASE FILL IN THE PR DESCRIPTION HERE ENSURING ALL CHECKLIST ITEMS ABOVE HAVE BEEN CONSIDERED.
Purpose
Test Plan
Test Result
(Optional) Documentation Update
BEFORE SUBMITTING, PLEASE READ https://docs.vllm.ai/en/latest/contributing (anything written below this line will be removed by GitHub Actions)