[Klaud Cold] Update minimaxm3-fp8-h200-vllm-agentic-mtp vLLM image to nightly-d9105ea8001e0a6d77a96327d17515bb5791fb36 (SemiAnalysisAI#2875)

Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>

Author: 2 people

configs/nvidia-master.yaml

@@ -7529,7 +7529,7 @@ minimaxm3-fp8-h100-vllm-agentic-mtp:
       - { tp: 8, spec-decoding: mtp, kv-offloading: dram, kv-offload-backend: { name: mooncake, version: "0.3.11.post1" }, conc-list: [6, 8] }
 
 minimaxm3-fp8-h200-vllm-agentic-mtp:
-  image: vllm/vllm-openai:v0.27.1
+  image: vllm/vllm-openai:nightly-d9105ea8001e0a6d77a96327d17515bb5791fb36
   model: MiniMaxAI/MiniMax-M3-MXFP8
   model-prefix: minimaxm3
   runner: cluster:h200-dgxc

perf-changelog.yaml

@@ -6930,3 +6930,11 @@
   description:
     - "Update SGLang image to lmsysorg/sglang:nightly-dev-cu13-20260907-30705c00 (2026-09-07 cu13 dev nightly, digest sha256:19b8fa1223cc339c1eae7a5b703f1a8c2543b5b119155bf3d7efaef18f77f007, tag commit sgl-project/sglang@30705c00; Docker Hub last pushed 2026-09-07T01:43:42Z) for both H200 Qwen3.5 FP8 SGLang AgentX MTP recipes: qwen3.5-fp8-h200-sglang-agentic-mtp from lmsysorg/sglang:v0.5.16-cu130 (v0.5.16 release) and qwen3.5-fp8-h200-sglang-agentic-hicache-mtp from lmsysorg/sglang:nightly-dev-cu13-20260815-a5ba081f (2026-08-15 nightly). Both route to benchmarks/single_node/agentic/qwen3.5_fp8_h200_mtp.sh, which is unchanged: SGLANG_ENABLE_SPEC_V2 EAGLE MTP at 3 steps, golden acceptance length 3.39, flashinfer attention with allreduce fusion, fp8 quantization and fp8_e4m3 KV, HiCache kernel IO / page_first layout. Concurrency grids unchanged. Same tag the B200 Qwen3.5 FP8/FP4 SGLang AgentX recipes moved to in #2861/#2862."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2868
+
+- config-keys:
+    - minimaxm3-fp8-h200-vllm-agentic-mtp
+  scenario-type:
+    - agentic-coding
+  description:
+    - "Update vLLM image from vllm/vllm-openai:v0.27.1 (v0.27.1 release) to vllm/vllm-openai:nightly-d9105ea8001e0a6d77a96327d17515bb5791fb36 (2026-09-07 upstream nightly, digest sha256:254eebf919e8b7b0d530d97fccc606c36f380ff6724d64bc190951bec1aee838, tag commit vllm-project/vllm@d9105ea8; Docker Hub last pushed 2026-09-07T06:16:01Z), the same tag the B200 MiniMax-M3 AgentX recipe moved to in #2860 and the ROCm counterpart of the MI325X/MI300X bumps in #2872/#2873. benchmarks/single_node/agentic/minimaxm3_fp8_h200_mtp.sh is unchanged: TRITON_ATTN attention, fp8 KV, EAGLE3 with the Inferact MiniMax-M3 EAGLE3-GQA draft pinned to FLASH_ATTN and the committed golden synthetic acceptance length 2.78, Mooncake 0.3.11.post1 DRAM offload on the host-tier arm; the resident TP8 c1/c2/c4/c6/c8/c10 and Mooncake DRAM offload c12/c14 grid is unchanged."
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2875