EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[Perf] move quant before allgather in Allgather EP (#3420)

Browse Source 
### What this PR does / why we need it?
move quant before allgather in Allgather EP, rely on
https://github.com/vllm-project/vllm-ascend/pull/3334

Deepseek R1 W8A8 performance on A2 with
`HCCL_ALGO="level0:NA;level1:pipeline"`:
| Seq length | Mean TTFT (ms) main | Mean TTFT (ms)  this PR |
|----------|----------|----------|
| 4k   |  375.21  | 364.99   |
| 16k  | 1465.23   | 1421.75  |
### Does this PR introduce _any_ user-facing change?

### How was this patch tested?


- vLLM version: v0.11.0
- vLLM main:
83f478bb19

---------

Signed-off-by: realliujiaxu <realliujiaxu@163.com>
This commit is contained in:
realliujiaxu
2025-11-04 16:49:58 +08:00
committed by GitHub
parent 44b58b8665
commit bedf223771
10 changed files with 160 additions and 66 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
vllm_ascend/quantization/w4a8_dynamic.py
View File

@@ -386,7 +386,6 @@ class AscendW4A8DynamicFusedMoEMethod:
w2_scale_bias=layer.w2_scale_bias,
topk_weights=topk_weights,
topk_ids=topk_ids,
use_int4_w4a8=True,
expert_map=expert_map,
log2phy=log2phy,
global_redundant_expert_num=global_redundant_expert_num,