Repository navigation
Conversation
Documents the qualified B300 single-node TP8 + DCP8 DSpark lane with an embedded Mooncake DRAM tier from InferenceX vllm-project#4381 as the b300_agentx_agg preset, with per-concurrency points for 1, 4, 8, 12, 14, 16, 24 and 28. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com> Signed-off-by: hshrivastava-droid <hshrivastava@nvidia.com>
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
Contributor
wzhao18
approved these changes
Oct 30, 2026
Contributor
|
/ci run |
|
✅ Triggered Buildkite CI #98562 for commit |
Merged
6 tasks done
esmeetu
temporarily deployed
to
recipes-preview
October 30, 2026 13:15 — with 
GitHub Actions
Inactive
GitHub Actions
Inactive
vllm-recipes-publisher Bot
pushed a commit
to vllm-project/recipes
that referenced
this pull request
Oct 30, 2026
vllm-project/vllm@main:recipes d14f72c7fe13 → 32b30dfe46d7 [latest.vllm] vllm-project/vllm@d14f72c...32b30df Commits: - vllm-project/vllm@32b30dfe46d7 [Recipe] Kimi-K3: add B300 AgentX aggregated Mooncake preset (vllm-project/vllm#64617) Recipes: - moonshotai/Kimi-K3 (modified): https://recipes.vllm.ai/moonshotai/Kimi-K3 Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
github-actions Bot
added a commit
to Yogaraj-Alamenda/recipes
that referenced
this pull request
Oct 30, 2026
…mni@1c8d26d03fc2 vllm-project/vllm@main:recipes a20280f30e31 → 6a95c89c3719 [latest.vllm] vllm-project/vllm@a20280f...6a95c89 Commits: - vllm-project/vllm@32b30dfe46d7 [Recipe] Kimi-K3: add B300 AgentX aggregated Mooncake preset (vllm-project/vllm#64617) Recipes: - moonshotai/Kimi-K3 (modified): https://recipes.vllm.ai/moonshotai/Kimi-K3 vllm-project/vllm-omni@main:recipes b69d1238ecae → 1c8d26d03fc2 [latest.vllm-omni] vllm-project/vllm-omni@b69d123...1c8d26d Commits: - vllm-project/vllm-omni@b5a6fd836185 [Model] MiniMax-H3: Continue a previous video from cached latents across requests (vllm-project/vllm-omni#9342) Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
github-actions Bot
added a commit
to yiminghub2024/recipes
that referenced
this pull request
Oct 30, 2026
…mni@1c8d26d03fc2 vllm-project/vllm@main:recipes 617135eed34b → ca5c2d7843a5 [latest.vllm] vllm-project/vllm@617135e...ca5c2d7 Commits: - vllm-project/vllm@32b30dfe46d7 [Recipe] Kimi-K3: add B300 AgentX aggregated Mooncake preset (vllm-project/vllm#64617) Recipes: - moonshotai/Kimi-K3 (modified): https://recipes.vllm.ai/moonshotai/Kimi-K3 vllm-project/vllm-omni@main:recipes 5d365c0ceae5 → 1c8d26d03fc2 [latest.vllm-omni] vllm-project/vllm-omni@5d365c0...1c8d26d Commits: - vllm-project/vllm-omni@b5a6fd836185 [Model] MiniMax-H3: Continue a previous video from cached latents across requests (vllm-project/vllm-omni#9342) Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
github-actions Bot
added a commit
to monkeyLoveding/recipes
that referenced
this pull request
Oct 30, 2026
vllm-project/vllm@main:recipes 24420b7a341a → e811e502efb2 [latest.vllm] vllm-project/vllm@24420b7...e811e50 Commits: - vllm-project/vllm@32b30dfe46d7 [Recipe] Kimi-K3: add B300 AgentX aggregated Mooncake preset (vllm-project/vllm#64617) Recipes: - moonshotai/Kimi-K3 (modified): https://recipes.vllm.ai/moonshotai/Kimi-K3 Signed-off-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
This branch was previously deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Purpose
Adds the
b300_agentx_aggpreset torecipes/models/moonshotai/Kimi-K3.yaml, documenting the qualified InferenceX AgentX lanekimik3-fp4-b300-vllm-agentic-dsparkfrom SemiAnalysisAI/InferenceX#4381 (pinned to809f704). It serves one 8-GPU B300 node with TP8 + DCP8, the Inferact DSpark draft, and an embedded RDMA Mooncake DRAM tier. B300 previously had only the generic Blackwell baseline.b300), strategysingle_node_tp. Opt-in (notdefault), because the lane carries benchmark-only settings.--max-num-seqs(2x concurrency) and its CUDA graph capture sizes.vllm/vllm-openai@sha256:8c1434bfee7d7626eef78ef70600434ffba4a1b8813cb1d1c8707ad25f47035e.--disable-uvicorn-access-log,PYTHONNOUSERSITEand the benchmark client variables (harness/logging only);RDMAV_DRIVERSandWITH_NVIDIA_PEERMEM(specific to the lane's cluster).--disable-flashinfer-autotunefrom the Blackwell baseline is removed in favour of the lane's--enable-flashinfer-autotune.Test Plan
From a
vllm-project/recipescheckout withsources.local.yamlpointing at this branch'srecipes/:node scripts/build-recipes-api.mjs,node scripts/validate-source.mjs --source vllm --dir <this>/recipes,pnpm test:schema,pnpm test:presets,pnpm test:guide-html.vllm_commandof all 8 points inpublic/moonshotai/Kimi-K3/presets/b300_agentx_agg.jsonwith the lane's effective args: 0 mismatches. The only extra flag is--max-model-len 1048576, which the Blackwell baseline emits and equals the model's native context.pnpm devUI preview.Test Result
All checks above pass.
Before/after of
presets/b300_agentx_agg.json: the file is new (no prior preset for B300).hw/b300/strategies/single_node_tp.jsonis unchanged because the preset is notdefault.AI assistance disclosure
Prepared with Claude Sonnet 5.5 (
claude-sonnet-5-5) under the submitter's direction; the submitter reviewed the diff and the rendered commands.🤖 Generated with Claude Code