feat: trim glm5.2 b300 DEP arm to conc 48 only / 将 glm5.2 b300 DEP 分支收敛为仅 conc 48

The sweep (run 29717273540) showed DEP conc 48 is the frontier peak
(163 tok/s/gpu out, ttft 2.8s, GPU radix hit 0.93) while conc 64 overflows
the HBM+host radix cache (hit collapses to 0.57), thrashing on re-prefill
down to 41 tok/s/gpu at 56s ttft. conc 64 is strictly dominated by conc 48
on both throughput and interactivity and cannot be recovered without more
host-pool capacity than the 3TB node can give (ratio > 0.75 OOMs, per
#2279). Drop it: DEP arm conc-list is now [48]. The agentic eval also moves
to conc 48 (off the conc-64 point whose eval hung).

中文：扫描(run 29717273540)显示 DEP conc 48 是前沿峰值(输出 163 tok/s/gpu，
ttft 2.8s，GPU radix 命中 0.93)，而 conc 64 会使 HBM+host radix 缓存溢出(命中
率跌至 0.57)，陷入 re-prefill 抖动，降至 41 tok/s/gpu、ttft 56s。conc 64 在吞吐
与交互性两个维度上都被 conc 48 严格支配，且在 3TB 节点上无法通过增大 host pool
恢复(ratio > 0.75 会 OOM，见 #2279)。故移除：DEP 分支 conc-list 现为 [48]。
agentic 评估点也随之移到 conc 48(避开评估卡住的 conc 64 点)。

Author: Oseltamivir

configs/nvidia-master.yaml

@@ -7875,5 +7875,8 @@ glm5.2-fp4-b300-sglang-agentic:
     - dram-utilization: 0.80
       search-space:
       - { tp: 8, kv-offloading: dram, kv-offload-backend: { name: hicache }, conc-list: [1, 2, 4, 8, 16, 32] }
-      # DEP (attention-DP + EP8 expert-parallel MoE) throughput arm, capped at conc 64.
-      - { tp: 8, ep: 8, dp-attn: true, kv-offloading: dram, kv-offload-backend: { name: hicache }, conc-list: [48, 64], router: { name: sglang-router, version: "0.3.2" } }
+      # DEP (attention-DP + EP8 expert-parallel MoE) throughput arm at conc 48 — the
+      # frontier peak (163 tok/s/gpu out, ttft 2.8s). conc >=64 overflows the HBM+host
+      # radix cache (GPU hit 0.93->0.57 at 48->64) and thrashes on re-prefill, so it is
+      # strictly dominated by conc 48 on both throughput and interactivity.
+      - { tp: 8, ep: 8, dp-attn: true, kv-offloading: dram, kv-offload-backend: { name: hicache }, conc-list: [48], router: { name: sglang-router, version: "0.3.2" } }

perf-changelog.yaml

@@ -4971,6 +4971,6 @@
     - glm5.2-fp4-b300-sglang-agentic
   description:
     - "Convert the high-throughput arm to DEP: add ep: 8 to the TP8/DP8 attention-DP arm so the MoE layers run expert-parallel across all 8 ranks (--ep-size 8) instead of TP-sharded (the arm previously left ep unset -> EP_SIZE=1). Attention stays data-parallel behind sglang-router consistent hashing; the recipe already passes --ep-size EP_SIZE unconditionally, so only the master-config search space changes"
-    - "Cap the DEP arm concurrency at 64: conc-list [48, 64] (was [48, 64, 96, 128, 192, 256, 512])"
+    - "DEP arm runs conc-list [48] only (was [48, 64, 96, 128, 192, 256, 512]): conc 48 is the frontier peak (163 tok/s/gpu out, ttft 2.8s); conc >=64 overflows the HBM+host radix cache (GPU hit 0.93->0.57 at 48->64) and thrashes on re-prefill, so it is strictly dominated"
     - "Low-latency TP8 arm (conc [1, 2, 4, 8, 16, 32], fp8 KV + cutedsl bf16 GEMM) unchanged; DEP mirrors the proven dsv4-fp4-b300-sglang tp8/ep8/dp-attn DEP arm"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2281