Skip to content

[PowerX] Qwen3.5 MI300X AgentX measured-power coverage #3716

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
edwingao28 wants to merge 4 commits into
base: main
Choose a base branch
from
Open

[PowerX] Qwen3.5 MI300X AgentX measured-power coverage #3716

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
2677603
feat: complete Qwen3.5 MI300X PowerX coverage
edwingao28 Oct 4, 2026
058e595
chore: link Qwen3.5 MI300X PowerX coverage PR
edwingao28 Oct 4, 2026
617ecf0
chore: merge main and re-append the PR #3716 changelog entry [skip-sw…
edwingao28 Oct 16, 2026
4b1c1ab
chore: keep one accurate changelog block for this PR [skip-sweep]
edwingao28 Oct 16, 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
4 changes: 2 additions & 2 deletions inferencex-e2e/benchmarks/single_node/srt-slurm-recipes/qwen3.5/sglang/mi300x-fp8/8k1k.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,7 +37,7 @@ zip_override_concurrency:
roles:
agg:
args:
cuda-graph-max-bs: [4, 8, 16, 32, 64]
cuda-graph-max-bs: [1, 2, 4, 8, 16, 32, 64]
benchmark:
env:
CONC: ['4', '8', '16', '32', '64']
CONC: ['1', '2', '4', '8', '16', '32', '64']
2 changes: 1 addition & 1 deletion inferencex-e2e/configs/amd-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -512,7 +512,7 @@ qwen3.5-fp8-mi300x-sglang:
osl: 1024
search-space:
- tp: 8
conc-start: 4
conc-start: 1
conc-end: 64
srt-recipe: 8k1k.yaml

Expand Down
8 changes: 8 additions & 0 deletions inferencex-e2e/perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10052,3 +10052,11 @@
- "Offload every Muse Spark Open H200 point to host DRAM with SGLang HiCache (120 GB per rank so the target and DFlash draft host pools fit the node cgroup, dram-utilization 0.80), keeping the concurrency 1-64 sweep."
- "Keep the GPU-resident settings unchanged: mem-fraction-static 0.85, matching running-request and decode-graph caps."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4199

- config-keys:
- qwen3.5-fp8-mi300x-sglang
scenario-type:
- fixed-seq-len
description:
- "Extend the Qwen3.5 FP8 MI300X SGLang 8k1k (ISL 8192 / OSL 1024, TP8) concurrency ladder from 4-64 to 1-64 in amd-master.yaml and the matching srt-slurm zip_override. The reused sweep measured and validated GPU power at every point. Leave the MTP arm unchanged."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/3716
Loading