diff --git a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
index a6b18a9d06..0875a2a640 100644
--- a/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
+++ b/include/ck/tensor_operation/gpu/device/device_gemm_multiple_d.hpp
@@ -147,7 +147,7 @@ struct DeviceGemmMultipleDSplitKBPreShuffle : public BaseOperator
     virtual int GetPreShuffleParameters() = 0;
 };
 
-#define  ShufflePadded   256
+#define ShufflePadded 256
 } // namespace device
 } // namespace tensor_operation
 } // namespace ck
diff --git a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d_b_preshuffle.hpp b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d_b_preshuffle.hpp
index 62867f3590..49a2f47539 100644
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d_b_preshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdl_cshuffle_v3_multi_d_b_preshuffle.hpp
@@ -936,6 +936,11 @@ struct GridwiseGemmMultiD_xdl_cshuffle_v3_b_preshuffle
             return false;
         }
 
+        if(karg.N % NPerXdl != 0)
+        {
+            return false;
+        }
+
         if constexpr(!(GemmSpec == tensor_operation::device::GemmSpecialization::MPadding ||
                        GemmSpec == tensor_operation::device::GemmSpecialization::MNPadding ||
                        GemmSpec == tensor_operation::device::GemmSpecialization::MKPadding ||