[AgentX] offload Muse Spark Open H200 to HiCache / [AgentX] Muse Spark Open H200 全部并发启用 HiCache 卸载

inferencex-e2e/benchmarks/single_node/srt-slurm-recipes/musespark/sglang/h200-fp4-mtp/agentic.yaml

@@ -58,12 +58,18 @@ base:
       AGENTIC_WARMUP_GRACE_PERIOD: '3600'
 
 
-zip_override_tp8:
+zip_override_tp8_hicache:
   roles:
     agg:
       args:
         max-running-requests: [2, 4, 8, 16, 32, 48, 64, 96, 128]
         cuda-graph-max-bs-decode: [2, 4, 8, 16, 32, 48, 64, 96, 128]
+        # The host pool keeps evicted agentic prefixes for reuse across turns.
+        # 160 GB per rank plus the ~30 GB DFlash draft pool exceeded the
+        # ~1.39 TiB cgroup across 8 ranks; 120 GB keeps the total near 1.2 TB.
+        enable-hierarchical-cache: true
+        hicache-size: 120
   benchmark:
     env:
       CONC: ['1', '2', '4', '8', '16', '24', '32', '48', '64']
+      KV_OFFLOADING: dram

inferencex-e2e/configs/nvidia-master.yaml

@@ -8376,8 +8376,9 @@ musespark-fp4-h200-sglang-agentic-dflash:
   srt-recipe-dir: musespark/sglang/h200-fp4-mtp
   scenarios:
     agentic-coding:
-    - search-space:
-      - { tp: 8, ep: 8, kv-offloading: none, spec-decoding: draft_model, conc-list: [1, 2, 4, 8, 16, 24, 32, 48, 64], srt-recipe: agentic.yaml }
+    - dram-utilization: 0.80
+      search-space:
+      - { tp: 8, ep: 8, kv-offloading: dram, kv-offload-backend: { name: hicache }, spec-decoding: draft_model, conc-list: [1, 2, 4, 8, 16, 24, 32, 48, 64], srt-recipe: agentic.yaml }
 
 dsv41flash-fp4-b300-vllm-agentic-dspark:
   image: vllm/vllm-openai:nightly-dev-x86_64-cu130-ac9126e58aa7

inferencex-e2e/perf-changelog.yaml

@@ -10043,3 +10043,12 @@
     - "Enable SGLang HiCache CPU DRAM KV offloading above conc 8: the TP4/EP4 and TP8/EP8 conc 12 and 16 points add enable-hierarchical-cache with hicache-size 180 GB per rank and write_through (720 GB at TP4, 1.44 TB at TP8; dram-utilization 0.85), matching the GLM-5.2 MXFP4 MI355X SGLang sibling."
     - "Conc 1, 4 and 8 stay GPU-resident with no KV offloading. Image, serve flags, DFlash2 draft and golden AL are unchanged."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/4201
+
+- config-keys:
+    - musespark-fp4-h200-sglang-agentic-dflash
+  scenario-type:
+    - agentic-coding
+  description:
+    - "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