Skip to content

[AMD][MI35X] Qwen3.5-fp4 SGLang single-node benchmark #1680

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
1am9trash wants to merge 4 commits into
base: main
Choose a base branch
from
+21 −4
Open

[AMD][MI35X] Qwen3.5-fp4 SGLang single-node benchmark #1680

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
08809b9
[AMD][MI355X] Qwen3.5-fp4: add aiter unified attn, allreduce fusion, …
yichiche Jun 8, 2026
60da2b8
Update config
1am9trash Jun 8, 2026
59ba6f5
Update change log
1am9trash Jun 8, 2026
c8804e2
Update change log and remove OOM config
1am9trash Jun 8, 2026
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
8 changes: 4 additions & 4 deletions .github/configs/amd-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -426,7 +426,7 @@ qwen3.5-fp8-mi355x-sglang-disagg:
- "DECODE_MTP_SIZE=0"

qwen3.5-fp4-mi355x-sglang:
image: lmsysorg/sglang:v0.5.12-rocm720-mi35x
image: lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260604
model: amd/Qwen3.5-397B-A17B-MXFP4
model-prefix: qwen3.5
runner: mi355x
Expand Down Expand Up @@ -468,7 +468,7 @@ qwen3.5-fp4-mi355x-atom:
- { tp: 4, conc-start: 4, conc-end: 16 }

qwen3.5-fp4-mi355x-sglang-mtp:
image: lmsysorg/sglang-rocm:v0.5.12-rocm720-mi35x-20260517
image: lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260604
model: amd/Qwen3.5-397B-A17B-MXFP4
model-prefix: qwen3.5
runner: mi355x
Expand All @@ -480,12 +480,12 @@ qwen3.5-fp4-mi355x-sglang-mtp:
- isl: 1024
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 256, spec-decoding: mtp }
- { tp: 2, conc-start: 4, conc-end: 128, spec-decoding: mtp }
- { tp: 4, conc-start: 4, conc-end: 16, spec-decoding: mtp }
- isl: 8192
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 256, spec-decoding: mtp }
- { tp: 2, conc-start: 4, conc-end: 128, spec-decoding: mtp }
- { tp: 4, conc-start: 4, conc-end: 16, spec-decoding: mtp }

qwen3.5-fp4-mi355x-sglang-disagg:
Expand Down
3 changes: 3 additions & 0 deletions benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ fi
if [[ "$MODEL" != /* ]]; then hf download "$MODEL"; fi

export SGLANG_USE_AITER=1
export SGLANG_USE_AITER_UNIFIED_ATTN=1

SERVER_LOG=/workspace/server.log
MEM_FRAC_STATIC=${MEM_FRAC_STATIC:-0.8}
Expand All @@ -38,6 +39,8 @@ python3 -m sglang.launch_server --model-path=$MODEL --trust-remote-code \
--model-loader-extra-config '{"enable_multithread_load": true}' \
--watchdog-timeout 1200 \
--disable-radix-cache \
--enable-aiter-allreduce-fusion --max-running-requests $CONC \
--page-size 16 \
Comment thread
1am9trash marked this conversation as resolved.
> $SERVER_LOG 2>&1 &

SERVER_PID=$!
Expand Down
3 changes: 3 additions & 0 deletions benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x_mtp.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,7 @@ fi
hf download "$MODEL"

export SGLANG_USE_AITER=1
export SGLANG_USE_AITER_UNIFIED_ATTN=1

SERVER_LOG=/workspace/server.log
MEM_FRAC_STATIC=${MEM_FRAC_STATIC:-0.8}
Expand All @@ -38,6 +39,8 @@ python3 -m sglang.launch_server --model-path=$MODEL --trust-remote-code \
--model-loader-extra-config '{"enable_multithread_load": true}' \
--watchdog-timeout 1200 \
--disable-radix-cache \
--enable-aiter-allreduce-fusion --max-running-requests $CONC \
Comment on lines 21 to +42

@functionstackx functionstackx Jun 8, 2026

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

thanks for the contributing @1am9trash can u add this too the cookbook?

Image

--page-size 16 \
--speculative-algorithm EAGLE \
--speculative-num-steps 3 \
--speculative-eagle-topk 1 \
Expand Down
11 changes: 11 additions & 0 deletions perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3523,3 +3523,14 @@
- "Aligned decode params with Weiliang config: swa-full-tokens-ratio=0.20, max-running-requests=18432, moe-dense-tp-size=1; added prefill enable-dp-lm-head and cuda-graph-max-bs=512"
- "Remove 4 dominated old configs (4p-dep16-8n, 8p-dep16-12n, 10p-dep16-14n, 12p-dep12-15n) superseded by wide-EP frontier"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1586

- config-keys:
- qwen3.5-fp4-mi355x-sglang
- qwen3.5-fp4-mi355x-sglang-mtp
description:
- "Bump image to lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260604."
- "Update script for aiter attention backend from triton."
- "Enable aiter unfied attention."
- "Enable aiter allreduce fusion."
- "Remove sweep config mtp+cc256+tp2, which may OOM."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1680