[Hardware][XPU] Register matmul and linear batch-invariant kernels for XPU (#49209)

Signed-off-by: tzielinski-habana <tomasz.zielinski@intel.com>
Signed-off-by: Tomasz Zielinski <85164140+tzielinski-habana@users.noreply.github.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Tomasz Zielinski
2026-09-02 13:31:38 +00:00
committed by GitHub
co-authored by mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> Claude Opus 5 Harry Mellor
parent 7894394b07
commit ffe3bb3c72
7 changed files with 272 additions and 33 deletions
+23 -1
View File
@@ -17,7 +17,29 @@ Batch invariance is crucial for several use cases:
## Hardware Requirements
Batch invariance requires NVIDIA GPUs with compute capability 8.0 or higher.
Batch invariance is supported on the following platforms:
- NVIDIA GPUs with compute capability 8.0 or higher.
- Intel XPUs with Triton support.
### Attention Backend Selection for XPU
On XPU, Triton Attention backend is required for batch invariance.
Select this backend using Qwen/Qwen3-1.7B model as an example:
```python
llm = LLM(
model="Qwen/Qwen3-1.7B",
attention_config={"backend": "TRITON_ATTN"},
)
```
Or via the CLI:
```bash
VLLM_BATCH_INVARIANT=1 vllm serve Qwen/Qwen3-1.7B \
--attention-config.backend TRITON_ATTN
```
## Enabling Batch Invariance