diff --git a/runs/frontier-fidelity-envelope-v1/jobs_qwen235_collective_profile.toml b/runs/frontier-fidelity-envelope-v1/jobs_qwen235_collective_profile.toml
index ce6ed9f..65d369a 100644
--- a/runs/frontier-fidelity-envelope-v1/jobs_qwen235_collective_profile.toml
+++ b/runs/frontier-fidelity-envelope-v1/jobs_qwen235_collective_profile.toml
@@ -43,3 +43,25 @@ MODEL = "/home/admin/cpfs/wjh/models/Qwen/Qwen3-235B-A22B-FP8"
 OUTPUT_ROOT = "/home/admin/cpfs/wjh/aituner/gpu-fleet-fidelity-envelope-v1/artifacts/qwen235-serving-allreduce-tp8-20260719-v1"
 VENV_ROOT = "/tmp/wjh/venvs/vllm-0.20.0-cu129-profiler-v1"
 VLLM_SOURCE = "/home/admin/cpfs/wjh/agentic-kv/third_party/vllm_v20_build"
+
+[[jobs]]
+name = "qwen235-vllm020-serving-allreduce-tp8-20260719-v2"
+gpus = 8
+gpu_model = "H20"
+hosts = ["dash0"]
+command = "cd /home/admin/cpfs/wjh/aituner/aituner-q235-collective-20260719/runs/frontier-qwen30-vllm020-profile-v1 && timeout --signal=TERM --kill-after=30s 1020 bash run_allreduce_profile.sh"
+artifacts = ["artifacts/qwen235-serving-allreduce-tp8-20260719-v2"]
+
+[jobs.env]
+XDG_CACHE_HOME = "/tmp/wjh/.cache"
+VLLM_CACHE_ROOT = "/tmp/wjh/.cache/vllm"
+TP = "8"
+NUM_TOKENS = "1 2 4 8 16 24 32 40 48 56 63 64 65 72 80 88 96 104 112 120 128 136 144 152 160 168 176 184 192 200 208 216 224 232 240 248 255 256 257 512 1024 2048 4096 8192"
+HIDDEN_DIM = "4096"
+COLLECTIVE_CONTRACT = "qwen235-serving-projected"
+TRIALS = "3"
+REPEATS = "20"
+MODEL = "/home/admin/cpfs/wjh/models/Qwen/Qwen3-235B-A22B-FP8"
+OUTPUT_ROOT = "/home/admin/cpfs/wjh/aituner/gpu-fleet-fidelity-envelope-v1/artifacts/qwen235-serving-allreduce-tp8-20260719-v2"
+VENV_ROOT = "/tmp/wjh/venvs/vllm-0.20.0-cu129-profiler-v1"
+VLLM_SOURCE = "/home/admin/cpfs/wjh/agentic-kv/third_party/vllm_v20_build"
diff --git a/runs/frontier-qwen30-vllm020-profile-v1/profile_vllm020_allreduce.py b/runs/frontier-qwen30-vllm020-profile-v1/profile_vllm020_allreduce.py
index 63309a1..04564b8 100644
--- a/runs/frontier-qwen30-vllm020-profile-v1/profile_vllm020_allreduce.py
+++ b/runs/frontier-qwen30-vllm020-profile-v1/profile_vllm020_allreduce.py
@@ -147,9 +147,28 @@ def main() -> None:
                     f"no FlashInfer fusion limit for SM{capability}, TP{world_size}"
                 ) from error
             if input_tensor.nbytes <= fusion_limit_bytes:
+                from vllm.distributed.device_communicators.flashinfer_all_reduce import (
+                    get_fi_ar_workspace,
+                )
+
                 fi_comm = communicator.fi_ar_comm
                 if fi_comm is None or fi_comm.disabled:
                     raise SystemExit("FlashInfer all-reduce communicator is unavailable")
+                max_token_num = fusion_limit_bytes // (
+                    args.hidden_dim * input_tensor.element_size()
+                )
+                workspace = get_fi_ar_workspace(
+                    world_size=world_size,
+                    rank=rank,
+                    max_token_num=max_token_num,
+                    hidden_dim=args.hidden_dim,
+                    dtype=input_tensor.dtype,
+                    group=get_tp_group().device_group,
+                )
+                if workspace is None:
+                    raise SystemExit(
+                        "FlashInfer rejected the serving-matched integer workspace"
+                    )
                 if not fi_comm.should_use_fi_ar(input_tensor):
                     raise SystemExit("FlashInfer rejected a fusion-eligible payload")
                 return (