Merge branch 'develop' into aosewski/wavelet_omniperf

dc0bae32 · Adam Osewski · 68474822 · ba40c2ce · dc0bae32 · dc0bae32
Commit dc0bae32 authored Feb 01, 2023 by Adam Osewski
20 changed files
--- a/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f16_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f16_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/utility/tuple.hpp"
+#include "ck/utility/data_type.hpp"
+
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp"
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+using Normalize = ck::tensor_operation::element_wise::NormalizeInInfer;
+
+// clang-format off
+template <index_t Rank>
+using device_batchnorm_infer_f16_instances =
+     std::tuple <
+        // Tuple<XDataType, MeanDataType, VarDataType, ScaleDataType, BiasDataType>, Tuple<YDataType>, NormalizeOp, Rank, MPerThread, Sequence<XVectorSize, MeanDataType, VarDataType, ScaleVectorSize, BiasVectorSize>, Sequence<YVectorSize> 
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 1, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 2, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 2, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 2, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 2, Sequence<2, 2, 2, 2, 2>, Sequence<2> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<2, 2, 2, 2, 2>, Sequence<2> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<4, 1, 1, 1, 1>, Sequence<4> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<1, 4, 4, 4, 4>, Sequence<1> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<4, 2, 2, 2, 2>, Sequence<4> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<2, 4, 4, 4, 4>, Sequence<2> >,  
+        DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, Rank, 4, Sequence<4, 4, 4, 4, 4>, Sequence<4> >  
+     >;
+// clang-format on
+
+void add_device_batchnorm_infer_rank_4_f16_instances(
+    std::vector<std::unique_ptr<
+        DeviceElementwise<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 4>>>& instances)
+{
+    add_device_operation_instances(instances, device_batchnorm_infer_f16_instances<4>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f32_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f32_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/utility/tuple.hpp"
+
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp"
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using F32 = float;
+
+using Normalize = ck::tensor_operation::element_wise::NormalizeInInfer;
+
+// clang-format off
+template <index_t Rank>
+using device_batchnorm_infer_f32_instances =
+     std::tuple <
+        // Tuple<XDataType, MeanDataType, VarDataType, ScaleDataType, BiasDataType>, Tuple<YDataType>, NormalizeOp, Rank, MPerThread, Sequence<XVectorSize, MeanDataType, VarDataType, ScaleVectorSize, BiasVectorSize>, Sequence<YVectorSize> 
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 1, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 2, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 2, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 2, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 2, Sequence<2, 2, 2, 2, 2>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<2, 2, 2, 2, 2>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<4, 1, 1, 1, 1>, Sequence<4> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<1, 4, 4, 4, 4>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<4, 2, 2, 2, 2>, Sequence<4> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<2, 4, 4, 4, 4>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, Rank, 4, Sequence<4, 4, 4, 4, 4>, Sequence<4> >
+     >;
+// clang-format on
+
+void add_device_batchnorm_infer_rank_4_f32_instances(
+    std::vector<std::unique_ptr<
+        DeviceElementwise<Tuple<F32, F32, F32, F32, F32>, Tuple<F32>, Normalize, 4>>>& instances)
+{
+    add_device_operation_instances(instances, device_batchnorm_infer_f32_instances<4>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f64_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_infer_f64_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/utility/tuple.hpp"
+
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp"
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using F64 = double;
+
+using Normalize = ck::tensor_operation::element_wise::NormalizeInInfer;
+
+// clang-format off
+template <index_t Rank>
+using device_batchnorm_infer_f64_instances =
+     std::tuple <
+        // Tuple<XDataType, MeanDataType, VarDataType, ScaleDataType, BiasDataType>, Tuple<YDataType>, NormalizeOp, Rank, MPerThread, Sequence<XVectorSize, MeanDataType, VarDataType, ScaleVectorSize, BiasVectorSize>, Sequence<YVectorSize> 
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 1, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 2, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 2, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 2, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 2, Sequence<2, 2, 2, 2, 2>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 4, Sequence<1, 1, 1, 1, 1>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 4, Sequence<2, 1, 1, 1, 1>, Sequence<2> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 4, Sequence<1, 2, 2, 2, 2>, Sequence<1> >,
+        DeviceElementwiseImpl<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, Rank, 4, Sequence<2, 2, 2, 2, 2>, Sequence<2> >
+     >;
+// clang-format on
+
+void add_device_batchnorm_infer_rank_4_f64_instances(
+    std::vector<std::unique_ptr<
+        DeviceElementwise<Tuple<F64, F64, F64, F64, F64>, Tuple<F64>, Normalize, 4>>>& instances)
+{
+    add_device_operation_instances(instances, device_batchnorm_infer_f64_instances<4>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/elementwise/device_normalize_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/elementwise/device_normalize_instance.cpp
@@ -6,7 +6,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_elementwise.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_elementwise_impl.hpp"
 #include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"

 namespace ck {
@@ -28,15 +28,15 @@ using Normalize = ck::tensor_operation::element_wise::Normalize;
 using device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances = std::tuple<
    // clang-format off
    //###################|<in, mean, square_mean, gamma, beta>| <out>|  functor| NDim| MPerThread| <in, mean, square_mean, gamma, beta ScalarPerVector>| <out ScalarPerVector>|
-    DeviceElementwise<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   8,       Sequence<8, 1, 1, 8, 8>,      Sequence<8>                >,
-    DeviceElementwise<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   4,       Sequence<4, 1, 1, 4, 4>,      Sequence<4>                >,
-    DeviceElementwise<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   2,       Sequence<2, 1, 1, 2, 2>,      Sequence<2>                >,
-    DeviceElementwise<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   1,       Sequence<1, 1, 1, 1, 1>,      Sequence<1>                >
+    DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   8,       Sequence<8, 1, 1, 8, 8>,      Sequence<8>                >,
+    DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   4,       Sequence<4, 1, 1, 4, 4>,      Sequence<4>                >,
+    DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   2,       Sequence<2, 1, 1, 2, 2>,      Sequence<2>                >,
+    DeviceElementwiseImpl<Tuple<F16, F32, F32, F16, F16>,  Tuple<F16>,  Normalize,  2,   1,       Sequence<1, 1, 1, 1, 1>,      Sequence<1>                >
    // clang-format on
    >;

 void add_device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances(
-    std::vector<DeviceElementwiseBasePtr<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2>>&
+    std::vector<DeviceElementwisePtr<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2>>&
        instances)
 {
    add_device_operation_instances(

--- a/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_km_kn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_km_kn_mn_instance.cpp
@@ -25,7 +25,8 @@ using S = ck::Sequence<Is...>;

 using PassThrough = ck::tensor_operation::element_wise::PassThrough;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault   = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNPadding = ck::tensor_operation::device::GemmSpecialization::MNPadding;

 // Compilation parameters for a[k, m] * b[k, n] = c[m, n]
 using device_gemm_xdl_f16_f16_f16_km_kn_mn_instances =
@@ -71,12 +72,36 @@ using device_gemm_xdl_f16_f16_f16_km_kn_mn_instances =
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_xdl_f16_f16_f16_km_kn_mn_irregular_tile_instances = std::tuple<
+    // clang-format off
+        //###########| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C|          GEMM| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds| CThreadTransfer| CThreadTransfer| NumPrefetch|          LoopScheduler|                     Pipeline|
+        //###########|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|              |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |            |            |            |              |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |            |                       |                             |
+        // pipeline v1, 1 wave
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,     Row,     Row, PassThrough, PassThrough, PassThrough,  GemmMNPadding,   64,     16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,     Row,     Row, PassThrough, PassThrough, PassThrough,  GemmMNPadding,   64,     16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,     Row,     Row, PassThrough, PassThrough, PassThrough,  GemmMNPadding,   64,     16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
 void add_device_gemm_xdl_f16_f16_f16_km_kn_mn_instances(
    std::vector<std::unique_ptr<
        DeviceGemm<Col, Row, Row, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
        instances)
 {
    add_device_operation_instances(instances, device_gemm_xdl_f16_f16_f16_km_kn_mn_instances{});
+    add_device_operation_instances(instances,
+                                   device_gemm_xdl_f16_f16_f16_km_kn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_km_nk_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_km_nk_mn_instance.cpp
@@ -25,7 +25,8 @@ using S = ck::Sequence<Is...>;

 using PassThrough = ck::tensor_operation::element_wise::PassThrough;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault   = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNPadding = ck::tensor_operation::device::GemmSpecialization::MNPadding;

 // Compilation parameters for a[k, m] * b[n, k] = c[m, n]
 using device_gemm_xdl_f16_f16_f16_km_nk_mn_instances =
@@ -71,12 +72,36 @@ using device_gemm_xdl_f16_f16_f16_km_nk_mn_instances =
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_xdl_f16_f16_f16_km_nk_mn_irregular_tile_instances = std::tuple<
+    // clang-format off
+        //###########| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C|          GEMM| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds| CThreadTransfer| CThreadTransfer| NumPrefetch|          LoopScheduler|                     Pipeline|
+        //###########|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|              |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |            |            |            |              |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |            |                       |                             |
+        // pipeline v1, 1 wave
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   GemmMNPadding,   64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   GemmMNPadding,   64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   GemmMNPadding,   64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
 void add_device_gemm_xdl_f16_f16_f16_km_nk_mn_instances(
    std::vector<std::unique_ptr<
        DeviceGemm<Col, Col, Row, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
        instances)
 {
    add_device_operation_instances(instances, device_gemm_xdl_f16_f16_f16_km_nk_mn_instances{});
+    add_device_operation_instances(instances,
+                                   device_gemm_xdl_f16_f16_f16_km_nk_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_mk_kn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_mk_kn_mn_instance.cpp
@@ -25,7 +25,8 @@ using S = ck::Sequence<Is...>;

 using PassThrough = ck::tensor_operation::element_wise::PassThrough;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault   = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNPadding = ck::tensor_operation::device::GemmSpecialization::MNPadding;

 // Compilation parameters for a[m, k] * b[k, n] = c[m, n]
 using device_gemm_xdl_f16_f16_f16_mk_kn_mn_instances =
@@ -98,12 +99,36 @@ using device_gemm_xdl_f16_f16_f16_mk_kn_mn_instances =
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_xdl_f16_f16_f16_mk_kn_mn_irregular_tile_instances = std::tuple<
+    // clang-format off
+        //###########| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C|          GEMM| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds| CThreadTransfer| CThreadTransfer| NumPrefetch|          LoopScheduler|                     Pipeline|
+        //###########|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|              |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|            |                       |                             |
+        //###########|      |      |      |        |        |        |        |            |            |            |              |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |            |                       |                             |
+        // pipeline v1, 1 wave
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    16,    16,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
 void add_device_gemm_xdl_f16_f16_f16_mk_kn_mn_instances(
    std::vector<std::unique_ptr<
        DeviceGemm<Row, Row, Row, F16, F16, F16, PassThrough, PassThrough, PassThrough>>>&
        instances)
 {
    add_device_operation_instances(instances, device_gemm_xdl_f16_f16_f16_mk_kn_mn_instances{});
+    add_device_operation_instances(instances,
+                                   device_gemm_xdl_f16_f16_f16_mk_kn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_mk_nk_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm/device_gemm_xdl_f16_f16_f16_mk_nk_mn_instance.cpp
@@ -94,17 +94,20 @@ using device_gemm_xdl_f16_f16_f16_mk_nk_mn_irregular_tile_instances = std::tuple
        //###########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|              |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|            |                       |                             |
        //###########|      |      |      |        |        |        |        |            |            |            |              |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |            |                       |                             |
        // pipeline v1, 1 wave
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     8,  8,   16,   16,    2,    9,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<8,  8, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     4,  8,   16,   16,    2,    9,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v1>
 #if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
        // pipeline v1, 2 waves
        ,
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     8,  8,   16,   16,    2,    9,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<8,  8, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     4,  8,   16,   16,    2,    9,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Interwave,        PipelineVersion::v1>
 #endif
 #if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
        // pipeline v2, 1 wave
        ,
+        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     8,  8,   16,   16,    2,    9,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<8,  8, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>,
        DeviceGemmXdl<   F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough, GemmMNPadding,   256,   128,   144,     4,  8,   16,   16,    2,    9,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 16, 4>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              2,              2,      true,               7,               1,           1,  LoopScheduler::Default,        PipelineVersion::v2>
 #endif

--- a/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instance.cpp
@@ -30,7 +30,8 @@ using S = ck::Sequence<Is...>;
 using PassThrough    = ck::tensor_operation::element_wise::PassThrough;
 using AddAddFastGelu = ck::tensor_operation::element_wise::AddAddFastGelu;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;

 // e = elementwise((a * b), d0, d1)
 // outout: e[m, n]
@@ -102,6 +103,29 @@ using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
 void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instances(
    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
                                                    Row,
@@ -118,6 +142,9 @@ void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn
    add_device_operation_instances(
        instances,
        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instance.cpp
@@ -30,7 +30,8 @@ using S = ck::Sequence<Is...>;
 using PassThrough    = ck::tensor_operation::element_wise::PassThrough;
 using AddAddFastGelu = ck::tensor_operation::element_wise::AddAddFastGelu;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;

 // e = elementwise((a * b), d0, d1)
 // outout: e[m, n]
@@ -102,6 +103,29 @@ using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,      PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
 void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instances(
    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
                                                    Col,
@@ -118,6 +142,9 @@ void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn
    add_device_operation_instances(
        instances,
        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_instance.cpp
@@ -30,7 +30,8 @@ using S = ck::Sequence<Is...>;
 using PassThrough    = ck::tensor_operation::element_wise::PassThrough;
 using AddAddFastGelu = ck::tensor_operation::element_wise::AddAddFastGelu;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;

 // e = elementwise((a * b), d0, d1)
 // outout: e[m, n]
@@ -102,6 +103,29 @@ using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
 void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_instances(
    std::vector<std::unique_ptr<DeviceGemmMultipleD<Row,
                                                    Row,
@@ -118,6 +142,9 @@ void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn
    add_device_operation_instances(
        instances,
        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_add_fastgelu/device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_instance.cpp
@@ -30,7 +30,8 @@ using S = ck::Sequence<Is...>;
 using PassThrough    = ck::tensor_operation::element_wise::PassThrough;
 using AddAddFastGelu = ck::tensor_operation::element_wise::AddAddFastGelu;

-static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;

 // e = elementwise((a * b), d0, d1)
 // outout: e[m, n]
@@ -93,6 +94,29 @@ using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn
        // clang-format on
        >;

+// irregular tile size
+using device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_F16_Tuple,   F16, PassThrough, PassThrough, AddAddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
 void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_instances(
    std::vector<std::unique_ptr<DeviceGemmMultipleD<Row,
                                                    Col,
@@ -109,6 +133,9 @@ void add_device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn
    add_device_operation_instances(
        instances,
        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_irregular_tile_instances{});
 }

 } // namespace instance

--- a/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/CMakeLists.txt
+add_instance_library(device_gemm_add_fastgelu_instance
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instance.cpp
+   device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
+)
--- a/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/utility/sequence.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+// e = elementwise((a * b), d0)
+// outout: e[m, n]
+// input: a[k, m], b[k, n], d0[m, n]
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instances = std::tuple<
+    // clang-format off
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|         CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise| Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|   Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |            |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   2,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   2,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   2,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   2,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   2,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   2,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   2,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   2,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   2,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
+// irregular tile size
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,      PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
+void add_device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
+                                                    Row,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddFastGelu>>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_kn_mn_mn_irregular_tile_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/utility/sequence.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+// e = elementwise((a * b), d0, d1)
+// outout: e[m, n]
+// input: a[k, m], b[n, k], d0[m, n], d1[m, n]
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instances = std::tuple<
+    // clang-format off
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   8,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   8,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   8,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   8,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   8,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   8,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   2,   8,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   2,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   2,   8,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   2,   8,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
+// irregular tile size
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,      PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
+void add_device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
+                                                    Col,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddFastGelu>>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_km_nk_mn_mn_irregular_tile_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/utility/sequence.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+// e = elementwise((a * b), d0, d1)
+// outout: e[m, n]
+// input: a[m, k], b[k, n], d0[m, n], d1[m, n]
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instances = std::tuple<
+    // clang-format off
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   2,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   2,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   2,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   2,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   2,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   2,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   2,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   2,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   2,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
+// irregular tile size
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Row,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
+void add_device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Row,
+                                                    Row,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddFastGelu>>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_kn_mn_mn_irregular_tile_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_fastgelu/device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+#include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/utility/sequence.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+// e = elementwise((a * b), d0, d1)
+// outout: e[m, n]
+// input: a[m, k], b[n, k], d0[m, n], d1[m ,n]
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instances = std::tuple<
+    // clang-format off
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    32,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    32,    64,    32,   8,   8,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES         
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    32,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    32,    64,    32,   8,   8,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,   128,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,   128,    32,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    64,    32,    32,   8,   8,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F32, F16_Tuple,   F16, PassThrough, PassThrough, AddFastGelu,    GemmDefault,        1,    64,    32,    64,    32,   8,   8,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 4>,               8, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+    // clang-format on
+    >;
+
+// irregular tile size
+using device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_irregular_tile_instances =
+    std::tuple<
+        // clang-format off
+        //##############################|      A|      B|            Ds|      E| AData| BData| AccData| CShuffle|        DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|          LoopScheduler|                    Pipeline|
+        //##############################| Layout| Layout|        Layout| Layout|  Type|  Type|    Type| DataType|          Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|                       |                            |
+        //##############################|       |       |              |       |      |      |        |         |              |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES        
+        // pipeline v2, 1 wave
+        ,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Row,    Col,     Row_Tuple,    Row,   F16,   F16,     F32,      F32,     F16_Tuple,   F16, PassThrough, PassThrough,    AddFastGelu, GemmMNKPadding,        1,    64,    16,    16,    32,   8,   8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              8,         1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              8,         1,           1,           1,               S<1, 16, 1, 4>,               1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
+        // clang-format on
+        >;
+
+void add_device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Row,
+                                                    Col,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddFastGelu>>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_instances{});
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_fastgelu_xdl_c_shuffle_f16_f16_f16_f16_mk_nk_mn_mn_irregular_tile_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/gemm_add_multiply/CMakeLists.txt
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_multiply/CMakeLists.txt
+add_instance_library(device_gemm_add_multiply_instance
+   device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instance.cpp
+   device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instance.cpp
+   device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_kn_mn_mn_mn_instance.cpp
+   device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_mk_nk_mn_mn_mn_instance.cpp
+)
--- a/library/src/tensor_operation_instance/gpu/gemm_add_multiply/device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_multiply/device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using F16       = ck::half_t;
+using F32       = float;
+using F16_Tuple = ck::Tuple<F16, F16>;
+
+using Row       = ck::tensor_layout::gemm::RowMajor;
+using Col       = ck::tensor_layout::gemm::ColumnMajor;
+using Row_Tuple = ck::Tuple<Row, Row>;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using AddMultiply = ck::tensor_operation::element_wise::AddMultiply;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+using device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instances =
+    std::tuple<
+        // clang-format off
+        // no padding
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|           CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|   Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|     Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |              |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   256,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   256,    32,   2,   2,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,    64,    32,   2,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,    64,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,    64,    32,   2,   2,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,    64,   128,    32,   2,   2,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+
+        // M/N/K Padding
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|           CDE|            GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|   Elementwise|  Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|     Operation|                |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |              |                |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   256,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   256,    32,   2,   2,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,   128,    32,   2,   2,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,    64,    32,   2,   2,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,    64,   128,    32,   2,   2,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,    64,    32,   2,   2,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,    64,   128,    32,   2,   2,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              2,         0,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Row, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>
+        // clang-format on
+        >;
+
+void add_device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
+                                                    Row,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddMultiply>>>& instances)
+{
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_kn_mn_mn_mn_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck
--- a/library/src/tensor_operation_instance/gpu/gemm_add_multiply/device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instance.cpp
+++ b/library/src/tensor_operation_instance/gpu/gemm_add_multiply/device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instance.cpp
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include <cstdlib>
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
+#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using F16       = ck::half_t;
+using F32       = float;
+using F16_Tuple = ck::Tuple<F16, F16>;
+
+using Row       = ck::tensor_layout::gemm::RowMajor;
+using Col       = ck::tensor_layout::gemm::ColumnMajor;
+using Row_Tuple = ck::Tuple<Row, Row>;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using AddMultiply = ck::tensor_operation::element_wise::AddMultiply;
+
+static constexpr auto GemmDefault    = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmMNKPadding = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
+
+using device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instances =
+    std::tuple<
+        // clang-format off
+        // no padding
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   256,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   256,    32,   2,   8,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,    64,    32,   2,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,    64,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,    64,    32,   2,   8,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,    64,   128,    32,   2,   8,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply,    GemmDefault,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+
+        // M/N/K Padding
+        //##############################|      A|      B|        Ds|      E| AData| BData| AccData| CShuffle|    DsData| EData|           A|           B|            CDE|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################| Layout| Layout|    Layout| Layout|  Type|  Type|    Type| DataType|      Type|  Type| Elementwise| Elementwise|    Elementwise| Specialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|
+        //##############################|       |       |          |       |      |      |        |         |          |      |   Operation|   Operation|      Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################|       |       |          |       |      |      |        |         |          |      |            |            |               |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   256,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   256,    32,   2,   8,   32,   32,    2,    4,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   256,    32,   8,   8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,   128,    32,   2,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,    64,    32,   2,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,   128,    64,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 4>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,    64,   128,    32,   2,   8,   32,   32,    2,    2,     S<8, 16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   128,    64,   128,    32,   8,   8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 16, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,    64,    32,   2,   8,   32,   32,    2,    1,     S<8, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,   128,    64,    32,   8,   8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,    64,   128,    32,   2,   8,   32,   32,    1,    2,     S<16,16, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              2,         0,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>,
+        DeviceGemmMultipleD_Xdl_CShuffle<    Col,    Col, Row_Tuple,    Row,   F16,   F16,     F32,      F16, F16_Tuple,   F16, PassThrough, PassThrough,    AddMultiply, GemmMNKPadding,        1,   256,    64,   128,    32,   8,   8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8>
+        // clang-format on
+        >;
+
+void add_device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instances(
+    std::vector<std::unique_ptr<DeviceGemmMultipleD<Col,
+                                                    Col,
+                                                    Row_Tuple,
+                                                    Row,
+                                                    F16,
+                                                    F16,
+                                                    F16_Tuple,
+                                                    F16,
+                                                    PassThrough,
+                                                    PassThrough,
+                                                    AddMultiply>>>& instances)
+{
+    add_device_operation_instances(
+        instances,
+        device_gemm_add_multiply_xdl_c_shuffle_f16_f16_f16_f16_f16_km_nk_mn_mn_mn_instances{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck