Skip to content

[AMD/ROCM] Update minimaxm2.5-fp8-mi355x-atom config #1194

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
seungrokj wants to merge 9 commits into
base: main
Choose a base branch
from
Open

[AMD/ROCM] Update minimaxm2.5-fp8-mi355x-atom config #1194

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
3ed70d6
atom minimax fp4 on mi355x
seungrokj Apr 16, 2026
ecce815
Update amd-master.yaml
seungrokj Apr 17, 2026
d2e4cb5
Update minimaxm2.5-fp4-mi355x-atom config and add minimaxm2.5-fp4-mi3…
seungrokj Apr 25, 2026
6f4696d
Merge branch 'main' into srok/atom_minimaxm2.5_fp4
seungrokj Apr 25, 2026
d613138
Fill in perf-changelog description for minimaxm2.5-fp4-mi355x-atom
seungrokj Apr 25, 2026
0ba47da
Update minimaxm2.5-fp8-mi355x-atom config and add perf-changelog entry
seungrokj Apr 27, 2026
ebc6424
Fill in PR link for minimaxm2.5-fp8-mi355x-atom perf-changelog entry
seungrokj Apr 27, 2026
9b0c6c5
Update minimaxm2.5-fp8-mi355x-atom: add gpu-memory-utilization, remov…
seungrokj Apr 27, 2026
6d4278b
Fill in perf-changelog description for minimaxm2.5-fp8-mi355x-atom
seungrokj Apr 27, 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
12 changes: 5 additions & 7 deletions .github/configs/amd-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -534,7 +534,7 @@ minimaxm2.5-fp8-mi355x-vllm:
- { tp: 8, ep: 8, conc-start: 2, conc-end: 2 }

minimaxm2.5-fp8-mi355x-atom:
image: rocm/atom:rocm7.2.1-ubuntu24.04-pytorch2.9.1-atom0.1.2
image: rocm/atom:rocm7.2.2_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom0.1.2.post
model: MiniMaxAI/MiniMax-M2.5
model-prefix: minimaxm2.5
runner: mi355x
Expand All @@ -545,15 +545,13 @@ minimaxm2.5-fp8-mi355x-atom:
- isl: 1024
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 128 }
- { tp: 4, conc-start: 4, conc-end: 128 }
- { tp: 8, ep: 8, conc-start: 32, conc-end: 256 }
- { tp: 2, conc-start: 4, conc-end: 256 }
- { tp: 4, conc-start: 4, conc-end: 256 }
- isl: 8192
osl: 1024
search-space:
- { tp: 2, conc-start: 4, conc-end: 128 }
- { tp: 4, conc-start: 4, conc-end: 128 }
- { tp: 8, ep: 8, conc-start: 32, conc-end: 256 }
- { tp: 2, conc-start: 4, conc-end: 256 }
- { tp: 4, conc-start: 4, conc-end: 256 }

minimaxm2.5-fp4-mi355x-vllm:
image: vllm/vllm-openai-rocm:v0.19.1
Expand Down
2 changes: 2 additions & 0 deletions benchmarks/single_node/minimaxm2.5_fp8_mi355x_atom.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,7 @@ fi

# Start GPU monitoring (power, temperature, clocks every second)
start_gpu_monitor
MEM_FRAC_STATIC=0.9

set -x

Expand All @@ -47,6 +48,7 @@ python3 -m atom.entrypoints.openai_server \
--server-port $PORT \
-tp $TP \
--kv_cache_dtype fp8 $CALCULATED_MAX_MODEL_LEN $EP \
--gpu-memory-utilization $MEM_FRAC_STATIC \
--trust-remote-code \
> $SERVER_LOG 2>&1 &

Expand Down
7 changes: 7 additions & 0 deletions perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1812,3 +1812,10 @@
- "Topologies: low-conc 1p1d-dep8-tep8 (4 nodes, mirrored from NVIDIA srt-slurm PR #71 with offload kept and numa-bind dropped); mid 1p1d-dep8-dep16 (6 nodes) and high 3p1d-dep8-dep16 (10 nodes) hand-rolled, structurally derived from the kimi-k2.5 1k/1k pattern"
- "Recipes stored under benchmarks/multi_node/srt-slurm-recipes/ and overlaid onto the upstream srt-slurm checkout at runtime"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1129

- config-keys:
- minimaxm2.5-fp8-mi355x-atom
description:
- "Update MiniMax-M2.5 FP8 MI355X Atom benchmark (rocm/atom:rocm7.2.2_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom0.1.2.post)"
- "Concurrency up to 256 on TP2/TP4"
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1194
Loading