diff --git a/example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8_bpreshuffle.cpp b/example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8_bpreshuffle.cpp
index 492d976411..fe1eca51b0 100644
--- a/example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8_bpreshuffle.cpp
+++ b/example/65_gemm_multiply_multiply/gemm_multiply_multiply_xdl_fp8_bpreshuffle.cpp
@@ -139,13 +139,13 @@ using DeviceOpInstance = ck::tensor_operation::device::DeviceGemmMultiD_Xdl_CShu
     // clang-format off
     <   Row, Col, DsLayout, ELayout, A0DataType, B0DataType, DsDataType, EDataType, AccDataType, CShuffleDataType,
         AElementOp,  BElementOp, CDEElementOp, GemmSpec, 256,
-        128,   128,    128,
+        256,   256,    128,
         16,   16,
         16,   16,
-        8,    2,
+        16,    4,
         S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0,
         S<8, 32, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 16, 16, 0,
-        1,    1,   S<1, 32, 1, 8>, S<8, 8, 1>,
+        2,    1,   S<1, 32, 1, 8>, S<8, 8, 1>,
         ck::BlockGemmPipelineScheduler::Intrawave, ck::BlockGemmPipelineVersion::v3, FP8>;
 // clang-format on
 
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 d0ff50a9c4..b988c9e39b 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
@@ -168,9 +168,7 @@ struct GridwiseGemmMultiD_xdl_cshuffle_v3_b_preshuffle
     static constexpr bool is_single_rate_mfma =
         (((is_same<ComputeTypeA, half_t>::value || is_same<ComputeTypeA, bhalf_t>::value) &&
           lcm_AK1_BK1 <= 4) ||
-         (is_same<ComputeTypeA, int8_t>::value && lcm_AK1_BK1 <= 8) ||
-         ((is_same<ComputeTypeA, f8_t>::value || is_same<ComputeTypeA, bf8_t>::value) &&
-          lcm_AK1_BK1 < 32))
+         (is_same<ComputeTypeA, int8_t>::value && lcm_AK1_BK1 <= 8))
             ? true
             : false;
     static constexpr auto is_scale_mfma    = false;