[torchair]remove aicpu op #2640
Merged
Conversation
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
Signed-off-by: zhangdepeng <zhangdepeng2@huawei.com>
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces a targeted optimization within the Ascend attention backend. The change modifies the calculation of block_size used for KV cache indexing by adding a zero-value NPU tensor. This technique is likely intended to influence the operator dispatching mechanism, ensuring that the subsequent integer division and modulo operations are executed on NPU kernels rather than falling back to slower AICPU operations. The change is localized and appears to be a correct, albeit subtle, platform-specific performance enhancement. I have not identified any critical or high-severity issues with this implementation.
|
👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:
If CI fails, you can run linting and testing checks locally according Contributing and Testing. |
|
wangxiyuan
approved these changes
Aug 30, 2025
845473182
pushed a commit
to raindaywhu/vllm-ascend
that referenced
this pull request
Sep 1, 2025
…into main_829 * 'main_829' of https://github.com/raindaywhu/vllm-ascend: [torchair]remove aicpu op (vllm-project#2640) bugfix for torchair graph (vllm-project#2639) [CI] fix UT error. (vllm-project#2644) [3/N][Feat][Graph] Support `all-to-all` and quantized models with ACL Graph (vllm-project#2614) [Bugfix] Fix mc2 operator error in aclgraph + ep<16 scenario (vllm-project#2609)
wenba0
pushed a commit
to wenba0/vllm-ascend
that referenced
this pull request
Sep 5, 2025
### What this PR does / why we need it? remove aicpu op for torchair mode ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? vLLM version: v0.10.1.1 vLLM main: vllm-project/vllm@05d839c - vLLM version: v0.10.1.1 - vLLM main: vllm-project/vllm@67c1490 Signed-off-by: zhangdepeng <zhangdepeng2@huawei.com> Co-authored-by: zhangdepeng <zhangdepeng2@huawei.com> Signed-off-by: lijiaojiao <lijiaojiao990304@163.com>
wangxiaoteng888
pushed a commit
to LCAIZJ/vllm-ascend
that referenced
this pull request
Sep 25, 2025
### What this PR does / why we need it? remove aicpu op for torchair mode ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? vLLM version: v0.10.1.1 vLLM main: vllm-project/vllm@05d839c - vLLM version: v0.10.1.1 - vLLM main: vllm-project/vllm@67c1490 Signed-off-by: zhangdepeng <zhangdepeng2@huawei.com> Co-authored-by: zhangdepeng <zhangdepeng2@huawei.com>
chopper0126
pushed a commit
to chopper0126/vllm-ascend
that referenced
this pull request
Sep 26, 2025
### What this PR does / why we need it? remove aicpu op for torchair mode ### Does this PR introduce _any_ user-facing change? No ### How was this patch tested? vLLM version: v0.10.1.1 vLLM main: vllm-project/vllm@05d839c - vLLM version: v0.10.1.1 - vLLM main: vllm-project/vllm@67c1490 Signed-off-by: zhangdepeng <zhangdepeng2@huawei.com> Co-authored-by: zhangdepeng <zhangdepeng2@huawei.com>
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.
What this PR does / why we need it?
remove aicpu op
Does this PR introduce any user-facing change?
No
How was this patch tested?
vLLM version: v0.10.1.1
vLLM main: vllm-project/vllm@05d839c