diff --git a/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x.sh b/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x.sh
index b036e060e1..87809f54b6 100644
--- a/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x.sh
+++ b/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x.sh
@@ -20,6 +20,7 @@ if [[ "$MODEL" != /* ]]; then hf download "$MODEL"; fi
 export SGLANG_USE_AITER=1
 export SGLANG_USE_AITER_UNIFIED_ATTN=1
 export AITER_FLYDSL_FORCE=1
+export SGLANG_MAMBA_SSM_DTYPE=bfloat16
 
 SERVER_LOG=/workspace/server.log
 MEM_FRAC_STATIC=${MEM_FRAC_STATIC:-0.8}
diff --git a/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x_mtp.sh b/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x_mtp.sh
index 8081b824e1..16e86e5443 100755
--- a/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x_mtp.sh
+++ b/benchmarks/single_node/fixed_seq_len/qwen3.5_fp4_mi355x_mtp.sh
@@ -20,6 +20,7 @@ hf download "$MODEL"
 export SGLANG_USE_AITER=1
 export SGLANG_USE_AITER_UNIFIED_ATTN=1
 export AITER_FLYDSL_FORCE=1
+export SGLANG_MAMBA_SSM_DTYPE=bfloat16
 
 SERVER_LOG=/workspace/server.log
 MEM_FRAC_STATIC=${MEM_FRAC_STATIC:-0.8}
diff --git a/configs/amd-master.yaml b/configs/amd-master.yaml
index 4e1d9a64d7..3e3170dac6 100644
--- a/configs/amd-master.yaml
+++ b/configs/amd-master.yaml
@@ -470,7 +470,7 @@ qwen3.5-fp8-mi355x-sglang-disagg:
           - "DECODE_MTP_SIZE=0"
 
 qwen3.5-fp4-mi355x-sglang:
-  image: lmsysorg/sglang-rocm:v0.5.13-rocm720-mi35x-20260612
+  image: lmsysorg/sglang-rocm:v0.5.15-rocm720-mi35x-20260713
   model: amd/Qwen3.5-397B-A17B-MXFP4
   model-prefix: qwen3.5
   runner: mi355x
@@ -512,7 +512,7 @@ qwen3.5-fp4-mi355x-atom:
       - { tp: 4, conc-start: 4, conc-end: 16 }
 
 qwen3.5-fp4-mi355x-sglang-mtp:
-  image: lmsysorg/sglang-rocm:v0.5.13-rocm720-mi35x-20260612
+  image: lmsysorg/sglang-rocm:v0.5.15-rocm720-mi35x-20260713
   model: amd/Qwen3.5-397B-A17B-MXFP4
   model-prefix: qwen3.5
   runner: mi355x
diff --git a/perf-changelog.yaml b/perf-changelog.yaml
index e8e9078c2e..fe36101866 100644
--- a/perf-changelog.yaml
+++ b/perf-changelog.yaml
@@ -4833,7 +4833,6 @@
     - "Add GLM-5.1 FP8 GB200 disaggregated dynamo-sglang MTP config covering 8k1k, high-throughput and low-latency topologies"
     - "Image: lmsysorg/sglang:v0.5.13.post1-cu130"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2123
-  
 
 - config-keys:
     - dsv4-fp8-mi325x-vllm
@@ -4860,3 +4859,11 @@
     - "Add GB300 DeepSeek-V4 dynamo-sglang agentic recipes (agg-tp4 + dep8 disagg pareto)"
   pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2157
   
+- config-keys:
+    - qwen3.5-fp4-mi355x-sglang
+    - qwen3.5-fp4-mi355x-sglang-mtp
+  description:
+    - "Bump image to lmsysorg/sglang-rocm:v0.5.15-rocm720-mi35x-20260713"
+    - "Add SGLANG_MAMBA_SSM_DTYPE=bfloat16 in both non-MTP and MTP benchmark scripts"
+  pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2201  
+