diff --git a/example/67_gemm_microscaling/moe_gemm2_xdl_mx_fp4.cpp b/example/67_gemm_microscaling/moe_gemm2_xdl_mx_fp4.cpp
index e3b1498bab..95d0ab763a 100644
--- a/example/67_gemm_microscaling/moe_gemm2_xdl_mx_fp4.cpp
+++ b/example/67_gemm_microscaling/moe_gemm2_xdl_mx_fp4.cpp
@@ -164,12 +164,12 @@ using DeviceOpInstance                     = ck::tensor_operation::device::Devic
     A0DataType,  A1DataType,  B0DataType,  B1DataType,  DsDataType, EDataType, AccDataType, CShuffleDataType,
     AElementOp,  BElementOp, CDEElementOp, GemmSpec,   
     ScaleBlockSize,      256,   
-    MPerBlock,   128,    256,
+    MPerBlock,   128,    128,
     32,   32,
     16,   16,
     8,    2,
-    S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 32, 32, 0,
-    S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 32, 32, 0,
+    S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 32, 32, 0,
+    S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 32, 32, 0,
     2,    2,   S<1, 32, 1, 8>, S<2, 1, 1, 1>,
     ck::BlockGemmPipelineScheduler::Intrawave, ck::BlockGemmPipelineVersion::v1, 0, false, false, MulRoutedWeight, false, ck::index_t, A0DataType>;
 // clang-format on
@@ -306,13 +306,13 @@ int main(int argc, char* argv[])
     {
     case 0: break;
     case 1:
-        a0_t_k_k.GenerateTensorValue(GeneratorTensor_2<A0DataType>{-2, 2});
-        b0_e_n_k.GenerateTensorValue(GeneratorTensor_2<B0DataType>{-2, 2});
-        a1_t_k_k.GenerateTensorValue(GeneratorTensor_3<A1DataType>{0, 1.0});
-        b1_e_n_k.GenerateTensorValue(GeneratorTensor_3<B1DataType>{0, 1.0});
+        a0_t_k_k.GenerateTensorValue(GeneratorTensor_2<A0DataType>{-1, 1});
+        b0_e_n_k.GenerateTensorValue(GeneratorTensor_2<B0DataType>{-1, 1});
+        a1_t_k_k.GenerateTensorValue(GeneratorTensor_2<A1DataType>{0, 1});
+        b1_e_n_k.GenerateTensorValue(GeneratorTensor_2<B1DataType>{0, 1});
         d0_t_n.GenerateTensorValue(GeneratorTensor_1<D0DataType>{}); // will to remove
         d1_e_n.GenerateTensorValue(GeneratorTensor_1<D1DataType>{}); // will to remove
-        d2_e_n.GenerateTensorValue(GeneratorTensor_2<D2DataType>{-2, 2});
+        d2_e_n.GenerateTensorValue(GeneratorTensor_2<D2DataType>{-1, 1});
         break;
     case 2:
         a0_t_k_k.GenerateTensorValue(GeneratorTensor_1<A0DataType>{});
diff --git a/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_selector_tmp.hpp b/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_moe_selector.hpp
similarity index 98%
rename from include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_selector_tmp.hpp
rename to include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_moe_selector.hpp
index 16965c9cec..4eeb9ccc14 100644
--- a/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_selector_tmp.hpp
+++ b/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_moe_selector.hpp
@@ -3,7 +3,7 @@
 
 #pragma once
 
-#include "ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_mx_tmp.hpp"
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_moe_mx.hpp"
 
 namespace ck {
 
@@ -75,7 +75,7 @@ constexpr auto BlockGemmMXBPreshufflePipeline_Selector()
         }
         else
         {
-            return BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_mx_tmp<
+            return BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_moe_mx<
                 BlkGemmPipeSche,
                 ThreadBlockSize,
                 ScaleBlockSize,
diff --git a/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_mx_tmp.hpp b/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_moe_mx.hpp
similarity index 99%
rename from include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_mx_tmp.hpp
rename to include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_moe_mx.hpp
index 9e12b0420d..773a1ab1fe 100644
--- a/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_mx_tmp.hpp
+++ b/include/ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuflle_v1_moe_mx.hpp
@@ -34,7 +34,7 @@ template <BlockGemmPipelineScheduler BlkGemmPipelineVer,
           index_t MRepeat, // MXdlPerWave
           index_t NRepeat, // NXdlPerWave
           index_t KPack>
-struct BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_mx_tmp
+struct BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_moe_mx
 {
 };
 
@@ -58,7 +58,7 @@ template <index_t ThreadBlockSize,
           index_t MRepeat, // MXdlPerWave
           index_t NRepeat, // NXdlPerWave
           index_t KPack>
-struct BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_mx_tmp<BlockGemmPipelineScheduler::Intrawave,
+struct BlockwiseGemmXdlops_pipeline_bpreshuffle_v1_moe_mx<BlockGemmPipelineScheduler::Intrawave,
                                                           ThreadBlockSize,
                                                           ScaleBlockSize,
                                                           ADataType,
diff --git a/include/ck/tensor_operation/gpu/grid/gridwise_moe_mx_gemm.hpp b/include/ck/tensor_operation/gpu/grid/gridwise_moe_mx_gemm.hpp
index abcb3e5227..14e7e19d70 100644
--- a/include/ck/tensor_operation/gpu/grid/gridwise_moe_mx_gemm.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_moe_mx_gemm.hpp
@@ -10,7 +10,7 @@
 #include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
 // #include
 // "ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_selector.hpp"
-#include "ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_selector_tmp.hpp" // for verification temporary, will to remove
+#include "ck/tensor_operation/gpu/block/blockwise_gemm_pipeline_xdlops_b_preshuffle_mx_moe_selector.hpp"
 #include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v4r1_gather.hpp"
 #include "ck/tensor_operation/gpu/block/thread_group_tensor_slice_transfer_v6r1.hpp"
 #include "ck/tensor_operation/gpu/thread/threadwise_tensor_slice_transfer.hpp"