Merge branch 'develop' into lwpck-471

library/include/ck/library/utility/convolution_parameter.hpp

@@ -10,6 +10,8 @@
 
 #include "ck/ck.hpp"
 
+#include "ck/library/utility/numeric.hpp"
+
 namespace ck {
 namespace utils {
 namespace conv {
@@ -55,10 +57,8 @@ struct ConvParam
         // sizeof(InDataType) * (G * N * C * <input spatial lengths product>) +
         return sizeof(InDataType) *
                (G_ * N_ * C_ *
-                std::accumulate(std::begin(input_spatial_lengths_),
-                                std::begin(input_spatial_lengths_) + num_dim_spatial_,
-                                static_cast<std::size_t>(1),
-                                std::multiplies<std::size_t>()));
+                ck::accumulate_n<std::size_t>(
+                    std::begin(input_spatial_lengths_), num_dim_spatial_, 1, std::multiplies<>()));
     }
 
     template <typename WeiDataType>
@@ -67,10 +67,8 @@ struct ConvParam
         // sizeof(WeiDataType) * (G * K * C * <filter spatial lengths product>) +
         return sizeof(WeiDataType) *
                (G_ * K_ * C_ *
-                std::accumulate(std::begin(filter_spatial_lengths_),
-                                std::begin(filter_spatial_lengths_) + num_dim_spatial_,
-                                static_cast<std::size_t>(1),
-                                std::multiplies<std::size_t>()));
+                ck::accumulate_n<std::size_t>(
+                    std::begin(filter_spatial_lengths_), num_dim_spatial_, 1, std::multiplies<>()));
     }
 
     template <typename OutDataType>

library/include/ck/library/utility/fill.hpp

@@ -30,9 +30,10 @@ struct FillUniformDistribution
     }
 
     template <typename ForwardRange>
-    auto operator()(ForwardRange&& range) -> std::void_t<decltype(
-        std::declval<FillUniformDistribution>()(std::begin(std::forward<ForwardRange>(range)),
-                                                std::end(std::forward<ForwardRange>(range))))>
+    auto operator()(ForwardRange&& range) const
+        -> std::void_t<decltype(std::declval<const FillUniformDistribution&>()(
+            std::begin(std::forward<ForwardRange>(range)),
+            std::end(std::forward<ForwardRange>(range))))>
     {
         (*this)(std::begin(std::forward<ForwardRange>(range)),
                 std::end(std::forward<ForwardRange>(range)));
@@ -72,6 +73,16 @@ struct FillUniformDistributionIntegerValue
         std::generate(
             first, last, [&dis, &gen]() { return ck::type_convert<T>(std::round(dis(gen))); });
     }
+
+    template <typename ForwardRange>
+    auto operator()(ForwardRange&& range) const
+        -> std::void_t<decltype(std::declval<const FillUniformDistributionIntegerValue&>()(
+            std::begin(std::forward<ForwardRange>(range)),
+            std::end(std::forward<ForwardRange>(range))))>
+    {
+        (*this)(std::begin(std::forward<ForwardRange>(range)),
+                std::end(std::forward<ForwardRange>(range)));
+    }
 };
 
 template <typename T>

library/include/ck/library/utility/host_common_util.hpp

@@ -4,9 +4,11 @@
 #pragma once
 
 #include <vector>
+#include <array>
 #include <iostream>
 #include <fstream>
 #include <string>
+#include <algorithm>
 
 #include "ck/ck.hpp"
 
@@ -72,5 +74,63 @@ static inline std::vector<T> getTypeValuesFromString(const char* cstr_values)
     return (values);
 }
 
+template <int NDim>
+static inline std::vector<std::array<index_t, NDim>>
+get_index_set(const std::array<index_t, NDim>& dim_lengths)
+{
+    static_assert(NDim >= 1, "NDim >= 1 is required to use this function!");
+
+    if constexpr(NDim == 1)
+    {
+        std::vector<std::array<index_t, NDim>> index_set;
+
+        for(int i = 0; i < dim_lengths[0]; i++)
+        {
+            std::array<index_t, 1> index{i};
+
+            index_set.push_back(index);
+        };
+
+        return index_set;
+    }
+    else
+    {
+        std::vector<std::array<index_t, NDim>> index_set;
+        std::array<index_t, NDim - 1> partial_dim_lengths;
+
+        std::copy(dim_lengths.begin() + 1, dim_lengths.end(), partial_dim_lengths.begin());
+
+        std::vector<std::array<index_t, NDim - 1>> partial_index_set;
+
+        partial_index_set = get_index_set<NDim - 1>(partial_dim_lengths);
+
+        for(index_t i = 0; i < dim_lengths[0]; i++)
+            for(const auto& partial_index : partial_index_set)
+            {
+                std::array<index_t, NDim> index;
+
+                index[0] = i;
+
+                std::copy(partial_index.begin(), partial_index.end(), index.begin() + 1);
+
+                index_set.push_back(index);
+            };
+
+        return index_set;
+    };
+};
+
+template <int NDim>
+static inline size_t get_offset_from_index(const std::array<index_t, NDim>& strides,
+                                           const std::array<index_t, NDim>& index)
+{
+    size_t offset = 0;
+
+    for(int i = 0; i < NDim; i++)
+        offset += index[i] * strides[i];
+
+    return (offset);
+};
+
 } // namespace host_common
 } // namespace ck

library/include/ck/library/utility/host_tensor.hpp

@@ -14,6 +14,9 @@
 #include "ck/utility/data_type.hpp"
 #include "ck/utility/span.hpp"
 
+#include "ck/library/utility/algorithm.hpp"
+#include "ck/library/utility/ranges.hpp"
+
 template <typename Range>
 std::ostream& LogRange(std::ostream& os, Range&& range, std::string delim)
 {
@@ -84,10 +87,10 @@ struct HostTensorDescriptor
         this->CalculateStrides();
     }
 
-    template <typename Range,
+    template <typename Lengths,
               typename = std::enable_if_t<
-                  std::is_convertible_v<decltype(*std::begin(std::declval<Range>())), std::size_t>>>
-    HostTensorDescriptor(const Range& lens) : mLens(lens.begin(), lens.end())
+                  std::is_convertible_v<ck::ranges::range_value_t<Lengths>, std::size_t>>>
+    HostTensorDescriptor(const Lengths& lens) : mLens(lens.begin(), lens.end())
     {
         this->CalculateStrides();
     }
@@ -102,13 +105,12 @@ struct HostTensorDescriptor
     {
     }
 
-    template <
-        typename Range1,
-        typename Range2,
-        typename = std::enable_if_t<
-            std::is_convertible_v<decltype(*std::begin(std::declval<Range1>())), std::size_t> &&
-            std::is_convertible_v<decltype(*std::begin(std::declval<Range2>())), std::size_t>>>
-    HostTensorDescriptor(const Range1& lens, const Range2& strides)
+    template <typename Lengths,
+              typename Strides,
+              typename = std::enable_if_t<
+                  std::is_convertible_v<ck::ranges::range_value_t<Lengths>, std::size_t> &&
+                  std::is_convertible_v<ck::ranges::range_value_t<Strides>, std::size_t>>>
+    HostTensorDescriptor(const Lengths& lens, const Strides& strides)
         : mLens(lens.begin(), lens.end()), mStrides(strides.begin(), strides.end())
     {
     }
@@ -244,14 +246,20 @@ struct Tensor
     {
     }
 
-    template <typename X>
-    Tensor(std::vector<X> lens) : mDesc(lens), mData(mDesc.GetElementSpaceSize())
+    template <typename X, typename Y>
+    Tensor(std::initializer_list<X> lens, std::initializer_list<Y> strides)
+        : mDesc(lens, strides), mData(mDesc.GetElementSpaceSize())
     {
     }
 
-    template <typename X, typename Y>
-    Tensor(std::vector<X> lens, std::vector<Y> strides)
-        : mDesc(lens, strides), mData(mDesc.GetElementSpaceSize())
+    template <typename Lengths>
+    Tensor(const Lengths& lens) : mDesc(lens), mData(mDesc.GetElementSpaceSize())
+    {
+    }
+
+    template <typename Lengths, typename Strides>
+    Tensor(const Lengths& lens, const Strides& strides)
+        : mDesc(lens, strides), mData(GetElementSpaceSize())
     {
     }
 
@@ -261,10 +269,10 @@ struct Tensor
     Tensor<OutT> CopyAsType() const
     {
         Tensor<OutT> ret(mDesc);
-        for(size_t i = 0; i < mData.size(); i++)
-        {
-            ret.mData[i] = ck::type_convert<OutT>(mData[i]);
-        }
+
+        ck::ranges::transform(
+            mData, ret.mData.begin(), [](auto value) { return ck::type_convert<OutT>(value); });
+
         return ret;
     }
 
@@ -294,13 +302,7 @@ struct Tensor
 
     std::size_t GetElementSpaceSizeInBytes() const { return sizeof(T) * GetElementSpaceSize(); }
 
-    void SetZero()
-    {
-        for(auto& v : mData)
-        {
-            v = T{0};
-        }
-    }
+    void SetZero() { ck::ranges::fill<T>(mData, 0); }
 
     template <typename F>
     void ForEach_impl(F&& f, std::vector<size_t>& idx, size_t rank)

library/include/ck/library/utility/iterator.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include <iterator>
+#include <utility>
+
+#include "ck/utility/type.hpp"
+
+namespace ck {
+
+template <typename T>
+using iter_value_t = typename std::iterator_traits<remove_cvref_t<T>>::value_type;
+
+template <typename T>
+using iter_reference_t = decltype(*std::declval<T&>());
+
+template <typename T>
+using iter_difference_t = typename std::iterator_traits<remove_cvref_t<T>>::difference_type;
+
+} // namespace ck

library/include/ck/library/utility/numeric.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include <iterator>
+#include <numeric>
+
+namespace ck {
+template <typename T, typename ForwardIterator, typename Size, typename BinaryOperation>
+auto accumulate_n(ForwardIterator first, Size count, T init, BinaryOperation op)
+    -> decltype(std::accumulate(first, std::next(first, count), init, op))
+{
+    return std::accumulate(first, std::next(first, count), init, op);
+}
+} // namespace ck

library/include/ck/library/utility/ranges.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include <iterator>
+#include <type_traits>
+#include <utility>
+
+#include "ck/library/utility/iterator.hpp"
+
+namespace ck {
+namespace ranges {
+
+template <typename R>
+using iterator_t = decltype(std::begin(std::declval<R&>()));
+
+template <typename R>
+using sentinel_t = decltype(std::end(std::declval<R&>()));
+
+template <typename R>
+using range_size_t = decltype(std::size(std::declval<R&>()));
+
+template <typename R>
+using range_difference_t = ck::iter_difference_t<ranges::iterator_t<R>>;
+
+template <typename R>
+using range_value_t = iter_value_t<ranges::iterator_t<R>>;
+
+template <typename R>
+using range_reference_t = iter_reference_t<ranges::iterator_t<R>>;
+
+template <typename T, typename = void>
+struct is_range : std::false_type
+{
+};
+
+template <typename T>
+struct is_range<
+    T,
+    std::void_t<decltype(std::begin(std::declval<T&>())), decltype(std::end(std::declval<T&>()))>>
+    : std::true_type
+{
+};
+
+template <typename T>
+inline constexpr bool is_range_v = is_range<T>::value;
+
+template <typename T, typename = void>
+struct is_sized_range : std::false_type
+{
+};
+
+template <typename T>
+struct is_sized_range<T, std::void_t<decltype(std::size(std::declval<T&>()))>>
+    : std::bool_constant<is_range_v<T>>
+{
+};
+} // namespace ranges
+} // namespace ck

library/src/tensor_operation_instance/gpu/batched_gemm/device_batched_gemm_xdl_f16_f16_f16_gkm_gkn_gmn_instance.cpp

@@ -28,18 +28,43 @@ using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 // Compilation parameters for a[k, m] * b[k, n] = c[m, n]
 using device_batched_gemm_xdl_f16_f16_f16_gkm_gkn_gmn_instances = std::tuple<
     // clang-format off
-        //##########|        AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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|
-        //##########|         Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  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|
-        //##########|             |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>
+        //##########|        AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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| NumGemmK|          LoopScheduler|                    Pipeline|
+        //##########|         Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar| Prefetch|                       |                            |
+        //##########|             |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector| Stage   |                       |                            |
+        //##########|             |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |         |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES
+        // pipeline v1, 2 waves
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES
+        // pipeline v2, 1 wave
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
     // clang-format on
     >;
 

library/src/tensor_operation_instance/gpu/batched_gemm/device_batched_gemm_xdl_f16_f16_f16_gkm_gnk_gmn_instance.cpp

@@ -28,18 +28,43 @@ using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 // Compilation parameters for a[k, m] * b[n, k] = c[m, n]
 using device_batched_gemm_xdl_f16_f16_f16_gkm_gnk_gmn_instances = std::tuple<
     // clang-format off
-        //##########|        AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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|
-        //##########|         Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  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|
-        //##########|             |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>
+        //##########|        AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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| NumGemmK|          LoopScheduler|                    Pipeline|
+        //##########|         Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar| Prefetch|                       |                            |
+        //##########|             |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector| Stage   |                       |                            |
+        //##########|             |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |         |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES
+        // pipeline v1, 2 waves
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES
+        // pipeline v2, 1 wave
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>
+#endif
     // clang-format on
     >;
 

library/src/tensor_operation_instance/gpu/batched_gemm/device_batched_gemm_xdl_f16_f16_f16_gmk_gkn_gmn_instance.cpp

@@ -28,27 +28,70 @@ using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 // Compilation parameters for a[m, k] * b[k, n] = c[m, n]
 using device_batched_gemm_xdl_f16_f16_f16_gmk_gkn_gmn_instances = std::tuple<
     // clang-format off
-        //#################| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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|
-        //#################|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  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|
-        //#################|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   256,     4,  8,   32,   32,    1,    4,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,    64,     4,  8,   32,   32,    1,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    64,    32,    32,     4,  8,   32,   32,    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,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   256,     4,  8,   16,   16,    1,    8,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   128,     4,  8,   16,   16,    1,    4,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    64,     4,  8,   16,   16,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    32,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    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>
+        //#################| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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| NumGemmK|          LoopScheduler|                    Pipeline|
+        //#################|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar| Prefetch|                       |                            |
+        //#################|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector| Stage   |                       |                            |
+        //#################|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |         |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   256,     4,  8,   32,   32,    1,    4,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,    64,     4,  8,   32,   32,    1,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    64,    32,    32,     4,  8,   32,   32,    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,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   256,     4,  8,   16,   16,    1,    8,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   128,     4,  8,   16,   16,    1,    4,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    64,     4,  8,   16,   16,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    32,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    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
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   256,     4,  8,   32,   32,    1,    4,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,    64,     4,  8,   32,   32,    1,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    64,    32,    32,     4,  8,   32,   32,    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,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   256,     4,  8,   16,   16,    1,    8,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   128,     4,  8,   16,   16,    1,    4,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    64,     4,  8,   16,   16,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    32,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    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
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   256,     4,  8,   32,   32,    1,    4,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    32,    64,     4,  8,   32,   32,    1,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    64,    32,    32,     4,  8,   32,   32,    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,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   256,     4,  8,   16,   16,    1,    8,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,   128,     4,  8,   16,   16,    1,    4,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    64,     4,  8,   16,   16,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    16,    32,     4,  8,   16,   16,    1,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,      Row,    Row, PassThrough, PassThrough, PassThrough,    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
     >;
 

library/src/tensor_operation_instance/gpu/batched_gemm/device_batched_gemm_xdl_f16_f16_f16_gmk_gnk_gmn_instance.cpp

@@ -28,23 +28,58 @@ using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 // Compilation parameters for a[m, k] * b[n, k] = c[m, n]
 using device_batched_gemm_xdl_f16_f16_f16_gmk_gnk_gmn_instances = std::tuple<
     // clang-format off
-        //#################| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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|
-        //#################|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  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|
-        //#################|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  8,   32,   32,    2,    2,     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>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    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>,
-        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  8,   32,   32,    1,    2,     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>
+        //#################| AData| BData| CData| AccData| ALayout| BLayout| CLayout|           A|           B|           C| 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| NumGemmK|          LoopScheduler|                    Pipeline|
+        //#################|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar| Prefetch|                       |                            |
+        //#################|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector| Stage   |                       |                            |
+        //#################|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |         |                       |                            |
+        // pipeline v1, 1 wave
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  8,   32,   32,    2,    2,     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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  8,   32,   32,    1,    2,     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>
+#if CK_EXPERIMENTAL_INTER_WAVE_INSTANCES        
+        // pipeline v1, 2 waves
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  8,   32,   32,    2,    2,     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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Interwave,        PipelineVersion::v1>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  8,   32,   32,    1,    2,     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>
+#endif
+#if CK_EXPERIMENTAL_PIPELINE_V2_INSTANCES         
+        // pipeline v2, 1 wave
+        ,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  8,   32,   32,    2,    2,     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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,      true,               7,               1,        1, LoopScheduler::Default,        PipelineVersion::v2>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    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>,
+        DeviceBatchedGemmXdl<  F16,   F16,   F16,     F32,     Row,     Col,     Row, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  8,   32,   32,    1,    2,     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>
+#endif
     // clang-format on
     >;
 

library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm/device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp

@@ -66,9 +66,7 @@ using device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_
         //#######################################|        |         |         |        |      |       |       |      |        |         |            |            |            |            |            |               |         |      |       |      |      |      |      |    |    |     |     |     |  Wave|  Wave|  Wave|                |               |               |               |               |               |          |                 |                |                |                |                |                |           |                 |                |                |                |                |                |           |            |            |                             |                |           |
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    256,   128,    40,    64,    32,   4,   4,    2,   32,   32,     2,     4,     2,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>,
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    256,   128,    40,   128,    32,   4,   4,    2,   32,   32,     2,     4,     4,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>,
-#if CK_WORKAROUND_DISABLE_BROKEN_ATTN_KERNEL_INSTANCE == 0
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    128,   256,    40,    64,    32,   4,   4,    2,   32,   32,     1,     8,     2,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>,
-#endif
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    128,   256,    40,   128,    32,   4,   4,    2,   32,   32,     1,     8,     4,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>,
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    128,   128,    40,    64,    32,   4,   4,    2,   32,   32,     1,     4,     2,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>,
         DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle<     Row,      Col,      Row,     Row,   F16,    F16,    F16,   F16,     F32,      F16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded,        1,   256,    128,   128,    40,   128,    32,   4,   4,    2,   32,   32,     1,     4,     4,     S<2,128, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,     false,      S<2,128, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               4,               4,      false,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8,   Masking>

library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm_permute/CMakeLists.txt

 add_instance_library(device_batched_gemm_softmax_gemm_permute_instance
     device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
+    device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instance.cpp
 )
 

library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm_permute/device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_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_batched_gemm_softmax_gemm_permute_xdl_cshuffle.hpp"
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using BF16 = ck::bhalf_t;
+using F32  = float;
+
+using Row = ck::tensor_layout::gemm::RowMajor;
+using Col = ck::tensor_layout::gemm::ColumnMajor;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using Scale       = ck::tensor_operation::element_wise::Scale;
+
+static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto GemmPadded  = ck::tensor_operation::device::GemmSpecialization::MNKOPadding;
+
+static constexpr auto TensorDefault = ck::tensor_operation::device::TensorSpecialization::Default;
+
+// c[g, m, n] = a[g, m, k] * b[g, n, k]
+template <index_t NumDimG,
+          index_t NumDimM,
+          index_t NumDimN,
+          index_t NumDimK,
+          index_t NumDimO,
+          MaskingSpecialization MaskingSpec>
+using device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instances =
+    std::tuple<
+        // clang-format off
+        // #############################################|  NumDimG| NumDimM| NumDimN| NumDimK| NumDimO|  AData|  B0Data|  B1Data|  CData| Acc0BiasData| Acc1BiasData| AccData| CShuffle|           A|          B0|        Acc0|          B1|           C|           GEMM|   ATensorSpec|  B0TensorSpec|  B1TensorSpec|   CTensorSpec| NumGemmK| Block| Gemm01| Gemm0| Gemm0| Gemm1| Gemm1| AK1| BK1| B1K1| MPer| NPer| Gemm0| Gemm0| Gemm1|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockLds|  B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer| MaskingSpec|
+        // #############################################|         |        |        |        |        |   Type|    Type|    Type|   Type|         Type|         Type|    Type| DataType| Elementwise| Elementwise| Elementwise| Elementwise| Elementwise| Specialization|              |              |              |              | Prefetch|  Size|   MPer|  NPer|  KPer|  NPer|  KPer|    |    |     |  XDL|  XDL|  MXdl|  NXdl|  NXdl|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|    ThreadCluster|   ThreadCluster|  SrcAccessOrder|    SrcVectorDim|       SrcScalar|       DstScalar|  AddExtraN|    ThreadCluster|   ThreadCluster|  SrcAccessOrder|    SrcVectorDim|       SrcScalar|       DstScalar|  AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector|            |
+        // #############################################|         |        |        |        |        |       |        |        |       |             |             |        |         |   Operation|   Operation|   Operation|   Operation|   Operation|               |              |              |              |              |    Stage|      |  Block| Block| Block| Block| Block|    |    |     |     |     |   Per|   Per|   Per| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          |  Lengths_K0_N_K1|    ArrangeOrder|                |                |       PerVector|    PerVector_K1|           |  Lengths_K0_N_K1|    ArrangeOrder|                |                |       PerVector|    PerVector_K1|           |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|            |
+        // #############################################|         |        |        |        |        |       |        |        |       |             |             |        |         |            |            |            |            |            |               |              |              |              |              |         |      |       |      |      |      |      |    |    |     |     |     |  Wave|  Wave|  Wave|                |               |               |               |               |               |          |                 |                |                |                |                |                |           |                 |                |                |                |                |                |           |            |            |                             |                |            |
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    256,   128,    32,    64,    32,   8,   8,    2,   32,   32,     2,     4,     2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    256,   128,    32,   128,    32,   8,   8,    2,   32,   32,     2,     4,     4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   256,    32,    64,    32,   8,   8,    2,   32,   32,     1,     8,     2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   256,    32,   128,    32,   8,   8,    2,   32,   32,     1,     8,     4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   128,    64,    64,    32,   8,   8,    2,   32,   32,     1,     4,     2,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,     false,      S<8, 32, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,      false,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   128,    32,    64,    32,   8,   8,    2,   32,   32,     1,     4,     2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   128,    64,   128,    32,   8,   8,    2,   32,   32,     1,     4,     4,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,     false,      S<8, 32, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,      false,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   128,    32,   128,    32,   8,   8,    2,   32,   32,     1,     4,     4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,     64,   256,    32,   128,    32,   8,   8,    2,   16,   16,     1,    16,     8,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           8,               S<1, 16, 1,16>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,     64,   256,    32,    64,    32,   8,   8,    2,   16,   16,     1,    16,     4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           4,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,     64,   256,    64,   128,    32,   8,   8,    2,   16,   16,     1,    16,     8,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<8, 32, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           8,               S<1, 16, 1,16>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,    GemmDefault, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,     64,   256,    64,    64,    32,   8,   8,    2,   16,   16,     1,    16,     4,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<8, 32, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S<16, 16, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           4,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        // Padded fallback kernel
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,   128,    64,   128,    32,   8,   8,    2,   32,   32,     1,     4,     4,     S<8, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,     false,      S<8, 32, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,      false,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>,
+        DeviceBatchedGemmSoftmaxGemmPermute_Xdl_CShuffle<  NumDimG, NumDimM, NumDimN, NumDimK, NumDimO,   BF16,    BF16,    BF16,   BF16,  ck::Tuple<>,  ck::Tuple<>,     F32,     BF16, PassThrough, PassThrough,       Scale, PassThrough, PassThrough,     GemmPadded, TensorDefault, TensorDefault, TensorDefault, TensorDefault,        1,   256,    128,    64,    32,   128,    32,   8,   8,    2,   32,   32,     1,     2,     4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      true,      S<4, 64, 1>,      S<1, 0, 2>,      S<1, 0, 2>,               2,               8,               8,       true,     S< 8, 32, 1>,      S<0, 2, 1>,      S<0, 2, 1>,               1,               4,               2,      false,           1,           2,               S<1, 32, 1, 8>,               8, MaskingSpec>
+        // clang-format on
+        >;
+
+void add_device_batched_gemm_masking_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instances(
+    std::vector<std::unique_ptr<
+        DeviceBatchedGemmSoftmaxGemmPermute<2,
+                                            1,
+                                            1,
+                                            1,
+                                            1,
+                                            BF16,
+                                            BF16,
+                                            BF16,
+                                            BF16,
+                                            ck::Tuple<>,
+                                            ck::Tuple<>,
+                                            PassThrough,
+                                            PassThrough,
+                                            Scale,
+                                            PassThrough,
+                                            PassThrough,
+                                            MaskingSpecialization::MaskOutUpperTriangle>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances,
+        device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instances<
+            2,
+            1,
+            1,
+            1,
+            1,
+            MaskingSpecialization::MaskOutUpperTriangle>{});
+}
+
+void add_device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instances(
+    std::vector<
+        std::unique_ptr<DeviceBatchedGemmSoftmaxGemmPermute<2,
+                                                            1,
+                                                            1,
+                                                            1,
+                                                            1,
+                                                            BF16,
+                                                            BF16,
+                                                            BF16,
+                                                            BF16,
+                                                            ck::Tuple<>,
+                                                            ck::Tuple<>,
+                                                            PassThrough,
+                                                            PassThrough,
+                                                            Scale,
+                                                            PassThrough,
+                                                            PassThrough,
+                                                            MaskingSpecialization::MaskDisabled>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances,
+        device_batched_gemm_softmax_gemm_permute_xdl_cshuffle_bf16_bf16_bf16_bf16_gmk_gnk_gno_gmo_instances<
+            2,
+            1,
+            1,
+            1,
+            1,
+            MaskingSpecialization::MaskDisabled>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/batchnorm/CMakeLists.txt

+add_instance_library(device_batchnorm_instance
+    device_batchnorm_forward_f16_instance.cpp
+    device_batchnorm_forward_f32_instance.cpp
+    device_batchnorm_forward_bf16_instance.cpp
+    device_batchnorm_forward_f64_instance.cpp
+    device_batchnorm_backward_f16_instance.cpp
+    device_batchnorm_backward_f32_instance.cpp
+    device_batchnorm_backward_bf16_instance.cpp
+    device_batchnorm_backward_f64_instance.cpp
+)

library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_backward_bf16_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_batchnorm_backward_impl.hpp"
+#include "ck/utility/data_type.hpp"
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using BF16 = ck::bhalf_t;
+using F32  = float;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_bf16_blockwise_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, DscaleDbiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcVectorSize, DscaleDbiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_bf16_multiblock_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcDstVectorSize, BiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<BF16, F32, F32, F32, BF16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+void add_device_batchnorm_backward_rank_4_3_bf16_instances(
+    std::vector<std::unique_ptr<
+        DeviceBatchNormBwd<BF16, F32, F32, F32, BF16, F32, F32, PassThrough, 4, 3>>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_bf16_blockwise_instances<4, 3, PassThrough>{});
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_bf16_multiblock_instances<4, 3, PassThrough>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_backward_f16_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_batchnorm_backward_impl.hpp"
+#include "ck/utility/data_type.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 PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f16_blockwise_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, DscaleDbiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcVectorSize, DscaleDbiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,  
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f16_multiblock_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcDstVectorSize, BiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F16, F32, F32, F32, F16, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+void add_device_batchnorm_backward_rank_4_3_f16_instances(
+    std::vector<
+        std::unique_ptr<DeviceBatchNormBwd<F16, F32, F32, F32, F16, F32, F32, PassThrough, 4, 3>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f16_blockwise_instances<4, 3, PassThrough>{});
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f16_multiblock_instances<4, 3, PassThrough>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_backward_f32_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_batchnorm_backward_impl.hpp"
+#include "ck/utility/data_type.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 PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f32_blockwise_instances = std::tuple<
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, DscaleDbiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcVectorSize, DscaleDbiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+    >;
+// clang-format on
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f32_multiblock_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcDstVectorSize, BiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F32, F32, F32, F32, F32, F32, F32, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+void add_device_batchnorm_backward_rank_4_3_f32_instances(
+    std::vector<
+        std::unique_ptr<DeviceBatchNormBwd<F32, F32, F32, F32, F32, F32, F32, PassThrough, 4, 3>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f32_blockwise_instances<4, 3, PassThrough>{});
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f32_multiblock_instances<4, 3, PassThrough>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_backward_f64_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_batchnorm_backward_impl.hpp"
+#include "ck/utility/data_type.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 PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f64_blockwise_instances = std::tuple<
+      // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, DscaleDbiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcVectorSize, DscaleDbiasDstVectorSize, MeanVarSrcVectorSize 
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+      DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+    >;
+// clang-format on
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename DyElementwiseOp>
+using device_batchnorm_backward_f64_multiblock_instances =
+     std::tuple <
+        // XDataType, DxDataType, DyDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, DyElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XDyDxVectorDim, XSrcVectorSize, DySrcVectorSize, DxDstVectorSize, ScaleSrcDstVectorSize, BiasDstVectorSize, MeanVarSrcVectorSize 
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,  2,  2,    1,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  1,  1,  1,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    0,  2,  2,  2,    1,  1,  1>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    2,  2,  2>,
+        DeviceBatchNormBwdImpl<F64, F64, F64, F64, F64, F64, F64, DyElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,  2,  2,    1,  1,  1,  1,    1,  1,  1>
+     >;
+// clang-format on
+
+void add_device_batchnorm_backward_rank_4_3_f64_instances(
+    std::vector<
+        std::unique_ptr<DeviceBatchNormBwd<F64, F64, F64, F64, F64, F64, F64, PassThrough, 4, 3>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f64_blockwise_instances<4, 3, PassThrough>{});
+    add_device_operation_instances(
+        instances, device_batchnorm_backward_f64_multiblock_instances<4, 3, PassThrough>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

library/src/tensor_operation_instance/gpu/batchnorm/device_batchnorm_forward_bf16_instance.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "ck/ck.hpp"
+#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_batchnorm_forward_impl.hpp"
+#include "ck/utility/data_type.hpp"
+
+#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace instance {
+
+using BF16 = ck::bhalf_t;
+using F32  = float;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename YElementwiseOp>
+using device_batchnorm_forward_bf16_blockwise_instances =
+     std::tuple <
+        // XDataType, YDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, YElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcYDstVectorDim, XSrcVectorSize, YDstVectorSize, ScaleSrcVectorSize, BiasSrcVectorSize, MeanVarSrcDstVectorSize 
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 128, 2,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 64,  4,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 32,  8,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 16, 16,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 8,  32,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 4,  64,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 2, 128,    2,    2,    1,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    0,    2,    2,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    0,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    0,    1,    1,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    0,    2,    2,    1,    1,    1>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    1,    1,    1,    2,    2,    2>,  
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, false, 256, 1, 256,    2,    2,    1,    1,    1,    1,    1,    1>
+     >;
+// clang-format on
+
+// clang-format off
+template <index_t Rank, index_t NumReduceDim, typename YElementwiseOp>
+using device_batchnorm_forward_bf16_multiblock_instances =
+     std::tuple <
+        // XDataType, YDataType, AccDataType, ScaleDataType, BiasDataType, MeanVarDataType, YElementwiseOp, Rank, NumReduceDim, UseMultiBlockInK, BLockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, XSrcYDstVectorDim, XSrcVectorSize, YDstVectorSize, ScaleSrcVectorSize, BiasSrcVectorSize, MeanVarSrcDstVectorSize
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 128, 2,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 64,  4,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 32,  8,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 16, 16,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 8,  32,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 4,  64,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 2, 128,    2,    2,    1,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    0,    2,    2,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    0,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    0,    1,    1,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    0,    2,    2,    1,    1,    1>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    1,    1,    1,    2,    2,    2>,
+        DeviceBatchNormFwdImpl<BF16, BF16, F32, BF16, BF16, F32, YElementwiseOp, Rank, NumReduceDim, true, 256, 1, 256,    2,    2,    1,    1,    1,    1,    1,    1>
+     >;
+// clang-format on
+
+void add_device_batchnorm_forward_rank_4_3_bf16_instances(
+    std::vector<
+        std::unique_ptr<DeviceBatchNormFwd<BF16, BF16, F32, BF16, BF16, F32, PassThrough, 4, 3>>>&
+        instances)
+{
+    add_device_operation_instances(
+        instances, device_batchnorm_forward_bf16_blockwise_instances<4, 3, PassThrough>{});
+    add_device_operation_instances(
+        instances, device_batchnorm_forward_bf16_multiblock_instances<4, 3, PassThrough>{});
+}
+
+} // namespace instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck