Skip to content

Add Kimi K2.5 INT4 single-node MI325X vLLM benchmark (TP8)#857

Merged
cquil11 merged 3 commits intomainfrom
claude/issue-856-20260303-0249
Mar 9, 2026
Merged

Add Kimi K2.5 INT4 single-node MI325X vLLM benchmark (TP8)#857
cquil11 merged 3 commits intomainfrom
claude/issue-856-20260303-0249

Conversation

@functionstackx
Copy link
Contributor

@functionstackx functionstackx commented Mar 3, 2026

following AMD andy's recipe https://x.com/linluo77/status/2017024513595301985

Add Kimi K2.5 INT4 single-node MI325X vLLM benchmark (TP8) using vLLM ROCm v0.16.0, based on MI355X INT4 recipe with AMD Andy Luo's recipe comment.

Closes #856

Generated with Claude Code

- Add benchmark script benchmarks/single_node/kimik2.5_int4_mi325x.sh
  based on MI355X INT4 recipe with AMD Andy Luo's recipe comment
- Add kimik2.5-int4-mi325x-vllm config to amd-master.yaml using
  vllm/vllm-openai-rocm:v0.16.0 image
- Update perf-changelog.yaml

Closes #856

Co-authored-by: functionstackx <functionstackx@users.noreply.github.com>
Copy link
Collaborator

@chunfangamd chunfangamd left a comment

Choose a reason for hiding this comment

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

lgtm

@cquil11 cquil11 merged commit f7135ac into main Mar 9, 2026
91 of 100 checks passed
@cquil11 cquil11 deleted the claude/issue-856-20260303-0249 branch March 9, 2026 17:57
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

Development

Successfully merging this pull request may close these issues.

vllm 0.16 single node mi325 kimi k2.5 vllm tp8

3 participants