diff --git a/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh b/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh
index c1d3d4da9..2cdcaa8cf 100644
--- a/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh
+++ b/benchmarks/single_node/agentic/qwen3.5_fp4_mi355x_sglang_mtp.sh
@@ -11,7 +11,8 @@ source "$(dirname "$0")/../../benchmark_lib.sh"
 export EVAL_FRAMEWORK="lm-eval"
 
 check_env_vars \
-    MODEL TP CONC EP_SIZE RESULT_DIR DURATION
+    MODEL TP CONC EP_SIZE KV_OFFLOADING \
+    TOTAL_CPU_DRAM_GB RESULT_DIR DURATION
 
 SCHEDULER_RECV_INTERVAL=${SCHEDULER_RECV_INTERVAL:-30}
 
@@ -53,6 +54,22 @@ trap cleanup_agentic_services EXIT
 trap 'exit 130' INT
 trap 'exit 143' TERM
 
+CACHE_ARGS=()
+if require_agentic_kv_offload_backend hicache; then
+    HICACHE_RATIO="${HICACHE_RATIO:-1.5}"
+    HICACHE_WRITE_POLICY="${HICACHE_WRITE_POLICY:-write_through}"
+    HICACHE_IO_BACKEND="${HICACHE_IO_BACKEND:-direct}"
+    HICACHE_MEM_LAYOUT="${HICACHE_MEM_LAYOUT:-page_first_direct}"
+    echo "HiCache CPU tier: ratio=$HICACHE_RATIO, write_policy=$HICACHE_WRITE_POLICY, io_backend=$HICACHE_IO_BACKEND, mem_layout=$HICACHE_MEM_LAYOUT, dram_budget=${TOTAL_CPU_DRAM_GB} GB, tp=$TP"
+    CACHE_ARGS=(
+        --enable-hierarchical-cache
+        --hicache-ratio "$HICACHE_RATIO"
+        --hicache-write-policy "$HICACHE_WRITE_POLICY"
+        --hicache-io-backend "$HICACHE_IO_BACKEND"
+        --hicache-mem-layout "$HICACHE_MEM_LAYOUT"
+    )
+fi
+
 PARALLEL_ARGS=(
     --tp "$TP"
     --dp 1
@@ -111,6 +128,7 @@ SGLANG_CMD=(
     --speculative-num-draft-tokens 4
     --enable-metrics
     --enable-cache-report
+    "${CACHE_ARGS[@]}"
 )
 
 printf '%q ' "${SGLANG_CMD[@]}" | tee "$RESULT_DIR/sglang_command.txt"
diff --git a/configs/amd-master.yaml b/configs/amd-master.yaml
index 67a060539..74b9701ca 100644
--- a/configs/amd-master.yaml
+++ b/configs/amd-master.yaml
@@ -372,8 +372,10 @@ qwen3.5-fp4-mi355x-sglang-agentic-mtp:
     agentic-coding:
     - dram-utilization: 0.80
       search-space:
-      - { tp: 2, ep: 2, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20] }
       - { tp: 4, ep: 1, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20, 24, 28, 32, 40] }
+      - { tp: 2, ep: 1, spec-decoding: mtp, kv-offloading: none, conc-list: [1, 4, 8, 12, 16, 20] }
+      - { tp: 4, ep: 1, spec-decoding: mtp, kv-offloading: dram, kv-offload-backend: { name: hicache }, conc-list: [40, 48, 56, 64] }
+      - { tp: 2, ep: 1, spec-decoding: mtp, kv-offloading: dram, kv-offload-backend: { name: hicache }, conc-list: [20, 24, 28, 32] }
 
 qwen3.5-fp4-mi355x-sglang-disagg:
   image: lmsysorg/sglang-rocm:v0.5.12.post1-rocm720-mi35x-20260523
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index d85a0f87d..28179dbbf 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -6288,3 +6288,11 @@
     - "Increase prefill mem_fraction_static from 0.72 to 0.85 and reduce HICACHE_RATIO from 4 to 3."
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2643
   
+- config-keys:
+    - qwen3.5-fp4-mi355x-sglang-agentic-mtp
+  scenario-type:
+    - agentic-coding
+  description:
+    - "Replace the TP2/EP2 arm with TP2/EP1 so the 2-GPU point matches qwen3.5-fp4-b200-sglang-agentic-mtp."
+    - "Add HiCache host-DRAM KV tier arms at TP4 concurrency 40, 48, 56, and 64 and TP2 concurrency 20, 24, 28, and 32, using hicache ratio 1.5 with write_through, direct io, and page_first_direct layout."
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2693