Fusion Conv+Bias+ReLU(+Add) (#62)

* fix relu

* clean up

* clean up

* adding 1x1 conv

* adding 1x1 conv

* added 1x1 conv

* refactor

* refactor

* refactor

* added profiler for conv+bias+relu+add

* clean up

* adding conv+bias+relu

* adding conv+bias+relu

* added conv+bias+relu

* Update README.md

* update cpu verification

* adding c shuffle

* update static_tensor for dealing with invalid element

* adding c shuffle

* debugging

* fix bug

* convert to fp16 before shuffle

* shuffle more than one M/NRepeat

* clean up

* remove coordinate step hack from GridwiseGemm_k0mk1_k0nk1_mn_xdlops_v3r1

* clean up

* remove coordinate step hack from all gridwise gemm xdl

* clean up coordinate step hack

* clean up coordinate step hack

* ThreadwiseTensorSliceTransfer_v3r2 support pointwise op on both src and dst

* adding output shuffle in conv+bias+relu+add

* update

* added conv+bias+relu+add with c shuffle

* added conv+bias+relu+add with c shuffle

* fix forward_sweep bugs in threadwise copy

* clean up

* refactor

* clean up

* clean up

* added conv_c_shuffle+bias_relu

* clean up

* added conv+bias+relu+atomic_add

* clean up

* clean up

* clean up

* clean up

* clean up

* clean up

* misc fixes; add 1x1 specialization

* clean up

* delete unused device op

* clean up

* add support for odd C value

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v3r3.hpp

+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V3R3_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V3R3_HPP
+
+#include "common_header.hpp"
+#include "tensor_descriptor.hpp"
+#include "tensor_descriptor_helper.hpp"
+#include "static_tensor.hpp"
+
+namespace ck {
+
+namespace detail {
+// TODO: How to fix this? It uses an struct instead of lambda because lambda
+// doesn't have constructor
+template <index_t SrcVectorDim,
+          index_t SrcScalarPerVector,
+          index_t DstVectorDim,
+          index_t DstScalarPerVector>
+struct lambda_scalar_per_access_for_src_and_dst
+{
+    __host__ __device__ constexpr auto operator()(index_t i) const
+    {
+        if(i == SrcVectorDim && i == DstVectorDim)
+        {
+            return math::lcm(SrcScalarPerVector, DstScalarPerVector);
+        }
+        else if(i == SrcVectorDim)
+        {
+            return SrcScalarPerVector;
+        }
+        else if(i == DstVectorDim)
+        {
+            return DstScalarPerVector;
+        }
+        else
+        {
+            return 1;
+        }
+    }
+};
+
+} // namespace detail
+
+// Assume:
+//   1. src_desc and dst_desc are not known at compile-time
+//   2. SrcBuffer and DstBuffer are DynamicBuffer
+//   3. src_slice_origin and dst_slice_origin are not known at compile-time,
+//   4. Use thread buffer
+template <typename SliceLengths,
+          typename SrcElementwiseOperation,
+          typename DstElementwiseOperation,
+          InMemoryDataOperationEnum_t DstInMemOp,
+          typename SrcData,
+          typename DstData,
+          typename SrcDesc,
+          typename DstDesc,
+          typename Dst0Desc,
+          typename Dst1Desc,
+          typename SrcDimAccessOrder,
+          typename DstDimAccessOrder,
+          index_t SrcVectorDim,
+          index_t DstVectorDim,
+          index_t SrcScalarPerVector,
+          index_t DstScalarPerVector,
+          index_t SrcScalarStrideInVector,
+          index_t DstScalarStrideInVector,
+          bool SrcResetCoordinateAfterRun, // control whether to move back src coordinate after each
+                                           // RunRead(),  will be fused with MoveSrcSliceWindow to
+                                           // save addr computation
+          bool DstResetCoordinateAfterRun> // control whether to move back dst coordinate after each
+                                           // RunWrite(),  will be fused with MoveDstSliceWindow to
+                                           // save addr computation
+struct ThreadwiseTensorSliceTransfer_v3r3
+{
+    static constexpr index_t nDim = SliceLengths::Size();
+    using Index                   = MultiIndex<nDim>;
+
+    using SrcCoord  = decltype(make_tensor_coordinate(SrcDesc{}, Index{}));
+    using DstCoord  = decltype(make_tensor_coordinate(DstDesc{}, Index{}));
+    using Dst0Coord = decltype(make_tensor_coordinate(Dst0Desc{}, Index{}));
+    using Dst1Coord = decltype(make_tensor_coordinate(Dst1Desc{}, Index{}));
+
+    using SrcCoordStep  = decltype(make_tensor_coordinate_step(SrcDesc{}, Index{}));
+    using DstCoordStep  = decltype(make_tensor_coordinate_step(DstDesc{}, Index{}));
+    using Dst0CoordStep = decltype(make_tensor_coordinate_step(Dst0Desc{}, Index{}));
+    using Dst1CoordStep = decltype(make_tensor_coordinate_step(Dst1Desc{}, Index{}));
+
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v3r3(
+        const SrcDesc& src_desc,
+        const Index& src_slice_origin,
+        const SrcElementwiseOperation& src_element_op,
+        const DstDesc& dst_desc,
+        const Dst0Desc& dst0_desc,
+        const Dst1Desc& dst1_desc,
+        const Index& dst_slice_origin,
+        const DstElementwiseOperation& dst_element_op)
+        : src_coord_(make_tensor_coordinate(src_desc, src_slice_origin)),
+          dst_coord_(make_tensor_coordinate(dst_desc, dst_slice_origin)),
+          dst0_coord_(make_tensor_coordinate(dst0_desc, dst_slice_origin)),
+          dst1_coord_(make_tensor_coordinate(dst1_desc, dst_slice_origin)),
+          src_element_op_(src_element_op),
+          dst_element_op_(dst_element_op)
+    {
+    }
+
+    __device__ void SetSrcSliceOrigin(const SrcDesc& src_desc, const Index& src_slice_origin_idx)
+    {
+        src_coord_ = make_tensor_coordinate(src_desc, src_slice_origin_idx);
+    }
+
+    __device__ void SetDstSliceOrigin(const DstDesc& dst_desc,
+                                      const Dst0Desc& dst0_desc,
+                                      const Dst1Desc& dst1_desc,
+                                      const Index& dst_slice_origin_idx)
+    {
+        dst_coord_  = make_tensor_coordinate(dst_desc, dst_slice_origin_idx);
+        dst0_coord_ = make_tensor_coordinate(dst0_desc, dst_slice_origin_idx);
+        dst1_coord_ = make_tensor_coordinate(dst1_desc, dst_slice_origin_idx);
+    }
+
+    template <typename SrcBuffer>
+    __device__ void RunRead(const SrcDesc& src_desc, const SrcBuffer& src_buf)
+    {
+        static_assert(SrcBuffer::GetAddressSpace() == AddressSpaceEnum_t::Global or
+                          SrcBuffer::GetAddressSpace() == AddressSpaceEnum_t::Lds,
+                      "wrong!");
+
+        static_assert(
+            is_same<remove_cvref_t<typename SrcBuffer::type>, remove_cvref_t<SrcData>>::value,
+            "wrong! SrcBuffer and SrcData data type are inconsistent");
+
+        constexpr auto I0 = Number<0>{};
+        constexpr auto I1 = Number<1>{};
+
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto src_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<SrcVectorDim, SrcScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto src_access_lengths = SliceLengths{} / src_scalar_per_access;
+
+        constexpr auto src_dim_access_order = SrcDimAccessOrder{};
+
+        constexpr auto ordered_src_access_lengths =
+            container_reorder_given_new2old(src_access_lengths, src_dim_access_order);
+
+        // make forward steps
+        const auto src_forward_steps = generate_tuple(
+            [&](auto i) {
+                Index forward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    forward_step_idx(j) = (i.value == j.value) ? src_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(src_desc, forward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make backward steps
+        const auto src_backward_steps = generate_tuple(
+            [&](auto i) {
+                Index backward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    backward_step_idx(j) = (i.value == j.value) ? -src_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(src_desc, backward_step_idx);
+            },
+            Number<nDim>{});
+
+        // loop over tensor and copy
+        static_ford<decltype(ordered_src_access_lengths)>{}([&](auto ordered_src_access_idx) {
+            // judge move forward or move backward
+            constexpr auto forward_sweep = [&]() {
+                StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+                forward_sweep_(I0) = true;
+
+                static_for<1, nDim, 1>{}([&](auto i) {
+                    index_t tmp = ordered_src_access_idx[I0];
+
+                    static_for<1, i, 1>{}([&](auto j) {
+                        tmp = tmp * ordered_src_access_lengths[j] + ordered_src_access_idx[j];
+                    });
+
+                    forward_sweep_(i) = tmp % 2 == 0;
+                });
+
+                return forward_sweep_;
+            }();
+
+            // calculate src data index
+            constexpr auto src_data_idx = [&]() {
+                Index ordered_idx;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    ordered_idx(i) = forward_sweep[i] ? ordered_src_access_idx[i]
+                                                      : ordered_src_access_lengths[i] - 1 -
+                                                            ordered_src_access_idx[i];
+                });
+
+                return container_reorder_given_old2new(ordered_idx, src_dim_access_order) *
+                       src_scalar_per_access;
+            }();
+
+            constexpr auto src_data_idx_seq = generate_sequence_v2(
+                [&](auto i) { return Number<src_data_idx[i]>{}; }, Number<src_data_idx.Size()>{});
+
+            const bool is_src_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src_desc, src_coord_);
+
+            using src_vector_type = vector_type_maker_t<SrcData, SrcScalarPerVector>;
+            using src_vector_t    = typename src_vector_type::type;
+
+            // copy data from src_buf into src_vector_container
+            auto src_vector_container = src_vector_type{
+                src_buf.template Get<src_vector_t>(src_coord_.GetOffset(), is_src_valid)};
+
+            // apply SrcElementwiseOperation on src_vector_container
+            static_for<0, SrcScalarPerVector, 1>{}([&](auto i) {
+                src_vector_container.template AsType<SrcData>()(i) =
+                    src_element_op_(src_vector_container.template AsType<SrcData>()[i]);
+            });
+
+            // copy data from src_vector_container into src_thread_scratch_
+            src_thread_scratch_.template SetAsType<src_vector_t>(
+                src_data_idx_seq, src_vector_container.template AsType<src_vector_t>()[I0]);
+
+            constexpr auto move_on_dim = [&]() constexpr
+            {
+                StaticallyIndexedArray<bool, nDim> move_on_dim_;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    move_on_dim_(i) = ordered_src_access_idx[i] < ordered_src_access_lengths[i] - 1;
+
+                    static_for<i + 1, nDim, 1>{}([&](auto j) {
+                        move_on_dim_(i) &=
+                            ordered_src_access_idx[j] == ordered_src_access_lengths[j] - 1;
+                    });
+                });
+
+                return move_on_dim_;
+            }
+            ();
+
+            // move src coord
+            static_for<0, nDim, 1>{}([&](auto i) {
+                if constexpr(move_on_dim[i])
+                {
+                    if constexpr(forward_sweep[i])
+                    {
+                        move_tensor_coordinate(
+                            src_desc, src_coord_, src_forward_steps[src_dim_access_order[i]]);
+                    }
+                    else
+                    {
+                        move_tensor_coordinate(
+                            src_desc, src_coord_, src_backward_steps[src_dim_access_order[i]]);
+                    }
+                }
+            });
+        });
+
+        // move src coordinate back to slice origin (or not)
+        if constexpr(SrcResetCoordinateAfterRun)
+        {
+            const auto src_reset_step =
+                make_tensor_coordinate_step(src_desc, GetSrcCoordinateResetStep());
+
+            move_tensor_coordinate(src_desc, src_coord_, src_reset_step);
+        }
+    }
+
+    __device__ void TransferDataFromSrcThreadScratchToDstThreadScratch()
+    {
+#if !CK_EXPERIMENTAL_USE_IN_REGISTER_SUB_DWORD_TRANSPOSE
+        static_ford<SliceLengths>{}([&](auto idx) {
+            // convert from SrcData to DstData here
+            dst_thread_scratch_(idx) = type_convert<DstData>(src_thread_scratch_[idx]);
+        });
+#else
+        // sub-dword transpose between src_thread_scratch_ and dst_thread_scratch_
+        // TODO make this logic more generic for more sub-dword datatype
+        if constexpr(SrcVectorDim != DstVectorDim &&
+                     is_same<half_t, remove_cvref_t<SrcData>>::value &&
+                     is_same<half_t, remove_cvref_t<DstData>>::value &&
+                     SrcScalarPerVector % 2 == 0 && DstScalarPerVector % 2 == 0)
+        {
+            // each transpose does
+            // DstScalarPerVector # of src vectors in src_thread_scratch_
+            // SrcScalarPerVector # of dst vectors in dst_thread_scratch_
+            constexpr index_t num_src_vector = Number<DstScalarPerVector>{};
+            constexpr index_t num_dst_vector = Number<SrcScalarPerVector>{};
+
+            // Assume SrcVectorDim is not the same as DstVectorDim, so we do transpose
+            // TODO: make this logic generic for all scenario
+            static_assert(SrcVectorDim != DstVectorDim, "wrong");
+
+            constexpr auto src_scalar_step_in_vector = generate_sequence(
+                detail::lambda_scalar_step_in_vector<SrcVectorDim>{}, Number<nDim>{});
+
+            constexpr auto dst_scalar_step_in_vector = generate_sequence(
+                detail::lambda_scalar_step_in_vector<DstVectorDim>{}, Number<nDim>{});
+
+            constexpr auto scalar_per_access = generate_sequence(
+                detail::lambda_scalar_per_access_for_src_and_dst<SrcVectorDim,
+                                                                 SrcScalarPerVector,
+                                                                 DstVectorDim,
+                                                                 DstScalarPerVector>{},
+                Number<nDim>{});
+
+            constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+            static_ford<decltype(access_lengths)>{}([&](auto access_idx) {
+                constexpr auto data_idx = access_idx * scalar_per_access;
+
+                constexpr auto data_idx_seq = generate_sequence_v2(
+                    [&](auto i) { return Number<data_idx[i]>{}; }, Number<nDim>{});
+
+                // TODO type_convert is not used yet!!!!!
+                using src_vector_t = vector_type_maker_t<SrcData, SrcScalarPerVector>;
+                using dst_vector_t = vector_type_maker_t<DstData, DstScalarPerVector>;
+
+                // get DstScalarPerVector # of read-only references to src vectors from
+                // src_thread_scratch_
+                const auto src_vector_refs = generate_tie(
+                    [&](auto i) -> const src_vector_t& {
+                        // i increment corresponds to movement in DstVectorDim
+                        return src_thread_scratch_.GetVectorTypeReference(
+                            data_idx_seq + i * dst_scalar_step_in_vector);
+                    },
+                    Number<num_src_vector>{});
+
+                // get SrcScalarPerVector # of references to dst vectors from dst_thread_scratch_
+                auto dst_vector_refs = generate_tie(
+                    [&](auto i) -> dst_vector_t& {
+                        // i increment corresponds to movement in SrcVectorDim
+                        return dst_thread_scratch_.GetVectorTypeReference(
+                            data_idx_seq + i * src_scalar_step_in_vector);
+                    },
+                    Number<num_dst_vector>{});
+
+                // do data transpose
+                // TODO type_convert is not used yet!!!!!
+                transpose_vectors<SrcData, DstScalarPerVector, SrcScalarPerVector>{}(
+                    src_vector_refs, dst_vector_refs);
+            });
+        }
+        else
+        {
+            static_ford<SliceLengths>{}([&](auto idx) {
+                // convert from SrcData to DstData here
+                dst_thread_scratch_(idx) = type_convert<DstData>(src_thread_scratch_[idx]);
+            });
+        }
+#endif
+    }
+
+    template <typename DstBuffer, typename Dst0Buffer, typename Dst1Buffer>
+    __device__ void RunWrite(const DstDesc& dst_desc,
+                             DstBuffer& dst_buf,
+                             const Dst0Desc& dst0_desc,
+                             const Dst0Buffer& dst0_buf,
+                             const Dst1Desc& dst1_desc,
+                             const Dst1Buffer& dst1_buf)
+    {
+        // if there is transpose, it's done here
+        // TODO move this elsewhere
+        TransferDataFromSrcThreadScratchToDstThreadScratch();
+
+        static_assert(DstBuffer::GetAddressSpace() == AddressSpaceEnum_t::Global or
+                          DstBuffer::GetAddressSpace() == AddressSpaceEnum_t::Lds,
+                      "wrong!");
+
+        static_assert(
+            is_same<remove_cvref_t<typename DstBuffer::type>, remove_cvref_t<DstData>>::value,
+            "wrong! SrcBuffer or DstBuffer data type is wrong");
+
+        constexpr auto I0 = Number<0>{};
+        constexpr auto I1 = Number<1>{};
+
+        // src scalar per access on each dim
+        // TODO: don't use this
+        constexpr auto dst_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<DstVectorDim, DstScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto dst_access_lengths = SliceLengths{} / dst_scalar_per_access;
+
+        constexpr auto dst_dim_access_order = DstDimAccessOrder{};
+
+        constexpr auto ordered_dst_access_lengths =
+            container_reorder_given_new2old(dst_access_lengths, dst_dim_access_order);
+
+        // make forward steps
+        const auto dst_forward_steps = generate_tuple(
+            [&](auto i) {
+                Index forward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    forward_step_idx(j) = (i.value == j.value) ? dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst_desc, forward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make forward steps: dst0
+        // WARNING!!!!!!: this logic is only correct if dst/dst0/dst1 can use the same
+        // DstScalarPerVector
+        // TODO: fix this
+        const auto dst0_forward_steps = generate_tuple(
+            [&](auto i) {
+                Index forward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    forward_step_idx(j) = (i.value == j.value) ? dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst0_desc, forward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make forward steps: dst1
+        // WARNING!!!!!!: this logic is only correct if dst/dst0/dst1 can use the same
+        // DstScalarPerVector
+        // TODO: fix this
+        const auto dst1_forward_steps = generate_tuple(
+            [&](auto i) {
+                Index forward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    forward_step_idx(j) = (i.value == j.value) ? dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst1_desc, forward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make backward steps
+        const auto dst_backward_steps = generate_tuple(
+            [&](auto i) {
+                Index backward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    backward_step_idx(j) = (i.value == j.value) ? -dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst_desc, backward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make backward steps: dst0
+        // WARNING!!!!!!: this logic is only correct if dst/dst0/dst1 can use the same
+        // DstScalarPerVector
+        // TODO: fix this
+        const auto dst0_backward_steps = generate_tuple(
+            [&](auto i) {
+                Index backward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    backward_step_idx(j) = (i.value == j.value) ? -dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst0_desc, backward_step_idx);
+            },
+            Number<nDim>{});
+
+        // make backward steps: dst1
+        // WARNING!!!!!!: this logic is only correct if dst/dst0/dst1 can use the same
+        // DstScalarPerVector
+        // TODO: fix this
+        const auto dst1_backward_steps = generate_tuple(
+            [&](auto i) {
+                Index backward_step_idx;
+
+                static_for<0, nDim, 1>{}([&](auto j) {
+                    backward_step_idx(j) = (i.value == j.value) ? -dst_scalar_per_access[i] : 0;
+                });
+
+                return make_tensor_coordinate_step(dst1_desc, backward_step_idx);
+            },
+            Number<nDim>{});
+
+        // loop over tensor and copy
+        static_ford<decltype(ordered_dst_access_lengths)>{}([&](auto ordered_dst_access_idx) {
+            // judge move forward or move backward
+            constexpr auto forward_sweep = [&]() {
+                StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+                forward_sweep_(I0) = true;
+
+                static_for<1, nDim, 1>{}([&](auto i) {
+                    index_t tmp = ordered_dst_access_idx[I0];
+
+                    static_for<1, i, 1>{}([&](auto j) {
+                        tmp = tmp * ordered_dst_access_lengths[j] + ordered_dst_access_idx[j];
+                    });
+
+                    forward_sweep_(i) = tmp % 2 == 0;
+                });
+
+                return forward_sweep_;
+            }();
+
+            // calculate dst data index
+            constexpr auto dst_data_idx = [&]() {
+                Index ordered_idx;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    ordered_idx(i) = forward_sweep[i] ? ordered_dst_access_idx[i]
+                                                      : ordered_dst_access_lengths[i] - 1 -
+                                                            ordered_dst_access_idx[i];
+                });
+
+                return container_reorder_given_old2new(ordered_idx, dst_dim_access_order) *
+                       dst_scalar_per_access;
+            }();
+
+            constexpr auto dst_data_idx_seq = generate_sequence_v2(
+                [&](auto i) { return Number<dst_data_idx[i]>{}; }, Number<dst_data_idx.Size()>{});
+
+            const bool is_dst_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(dst_desc, dst_coord_);
+
+            using dst_vector_type = vector_type_maker_t<DstData, DstScalarPerVector>;
+            using dst_vector_t    = typename dst_vector_type::type;
+
+            // copy data from dst_thread_scratch_ into dst_vector_container
+            auto dst_vector_container = dst_vector_type{
+                dst_thread_scratch_.template GetAsType<dst_vector_t>(dst_data_idx_seq)};
+
+            // apply DstElementwiseOperation on dst_vector_container
+            static_for<0, DstScalarPerVector, 1>{}([&](auto i) {
+                dst_vector_container.template AsType<DstData>()(i) =
+                    dst_element_op_(dst_vector_container.template AsType<DstData>()[i]);
+            });
+
+            // copy data from dst_vector_container to dst_buf
+            dst_buf.template Set<dst_vector_t>(
+                dst_coord_.GetOffset(),
+                is_dst_valid,
+                dst_vector_container.template AsType<dst_vector_t>()[I0]);
+
+            constexpr auto move_on_dim = [&]() constexpr
+            {
+                StaticallyIndexedArray<bool, nDim> move_on_dim_;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    move_on_dim_(i) = ordered_dst_access_idx[i] < ordered_dst_access_lengths[i] - 1;
+
+                    static_for<i + 1, nDim, 1>{}([&](auto j) {
+                        move_on_dim_(i) &=
+                            ordered_dst_access_idx[j] == ordered_dst_access_lengths[j] - 1;
+                    });
+                });
+
+                return move_on_dim_;
+            }
+            ();
+
+            // move dst coord
+            static_for<0, nDim, 1>{}([&](auto i) {
+                if constexpr(move_on_dim[i])
+                {
+                    if constexpr(forward_sweep[i])
+                    {
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_forward_steps[dst_dim_access_order[i]]);
+                    }
+                    else
+                    {
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_backward_steps[dst_dim_access_order[i]]);
+                    }
+                }
+            });
+        });
+
+        // move dst coordinate back to slice origin (or not)
+        if constexpr(DstResetCoordinateAfterRun)
+        {
+            const auto dst_reset_step =
+                make_tensor_coordinate_step(dst_desc, GetDstCoordinateResetStep());
+
+            move_tensor_coordinate(dst_desc, dst_coord_, dst_reset_step);
+        }
+    }
+
+    __device__ static constexpr auto GetSrcCoordinateResetStep()
+    {
+        constexpr auto I0 = Number<0>{};
+
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto src_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<SrcVectorDim, SrcScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto src_access_lengths = SliceLengths{} / src_scalar_per_access;
+
+        constexpr auto src_dim_access_order = SrcDimAccessOrder{};
+
+        constexpr auto ordered_src_access_lengths =
+            container_reorder_given_new2old(src_access_lengths, src_dim_access_order);
+
+        // judge move forward or move backward during the last iteration
+        constexpr auto forward_sweep = [&]() {
+            StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+            forward_sweep_(I0) = true;
+
+            // TODO: BUG: should start at 1
+            static_for<1, nDim, 1>{}([&](auto i) {
+                index_t tmp = ordered_src_access_lengths[I0] - 1;
+
+                static_for<1, i, 1>{}([&](auto j) {
+                    tmp = tmp * ordered_src_access_lengths[j] + ordered_src_access_lengths[j] - 1;
+                });
+
+                forward_sweep_(i) = tmp % 2 == 0;
+            });
+
+            return forward_sweep_;
+        }();
+
+        // calculate src data index after last iteration in RunRead(), if it has not being reset by
+        // RunRead()
+        constexpr auto src_data_idx = [&]() {
+            Index ordered_idx;
+
+            static_for<0, nDim, 1>{}([&](auto i) {
+                ordered_idx(i) = forward_sweep[i] ? ordered_src_access_lengths[i] - 1 : 0;
+            });
+
+            return container_reorder_given_old2new(ordered_idx, src_dim_access_order) *
+                   src_scalar_per_access;
+        }();
+
+        //
+        constexpr auto reset_src_data_step = [&]() {
+            Index reset_src_data_step_;
+
+            static_for<0, nDim, 1>{}([&](auto i) { reset_src_data_step_(i) = -src_data_idx[i]; });
+
+            return reset_src_data_step_;
+        }();
+
+        return reset_src_data_step;
+    }
+
+    __device__ static constexpr auto GetDstCoordinateResetStep()
+    {
+        constexpr auto I0 = Number<0>{};
+
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto dst_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<DstVectorDim, DstScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto dst_access_lengths = SliceLengths{} / dst_scalar_per_access;
+
+        constexpr auto dst_dim_access_order = DstDimAccessOrder{};
+
+        constexpr auto ordered_dst_access_lengths =
+            container_reorder_given_new2old(dst_access_lengths, dst_dim_access_order);
+
+        // judge move forward or move backward during the last iteration
+        constexpr auto forward_sweep = [&]() {
+            StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+            forward_sweep_(I0) = true;
+
+            static_for<1, nDim, 1>{}([&](auto i) {
+                index_t tmp = ordered_dst_access_lengths[I0] - 1;
+
+                static_for<1, i, 1>{}([&](auto j) {
+                    tmp = tmp * ordered_dst_access_lengths[j] + ordered_dst_access_lengths[j] - 1;
+                });
+
+                forward_sweep_(i) = tmp % 2 == 0;
+            });
+
+            return forward_sweep_;
+        }();
+
+        // calculate dst data index after last iteration in RunWrite(), if it has not being reset by
+        // RunWrite()
+        constexpr auto dst_data_idx = [&]() {
+            Index ordered_idx;
+
+            static_for<0, nDim, 1>{}([&](auto i) {
+                ordered_idx(i) = forward_sweep[i] ? ordered_dst_access_lengths[i] - 1 : 0;
+            });
+
+            return container_reorder_given_old2new(ordered_idx, dst_dim_access_order) *
+                   dst_scalar_per_access;
+        }();
+
+        //
+        constexpr auto reset_dst_data_step = [&]() {
+            Index reset_dst_data_step_;
+
+            static_for<0, nDim, 1>{}([&](auto i) { reset_dst_data_step_(i) = -dst_data_idx[i]; });
+
+            return reset_dst_data_step_;
+        }();
+
+        return reset_dst_data_step;
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrcSliceWindow(const SrcDesc& src_desc,
+                                       const Index& src_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx =
+            SrcResetCoordinateAfterRun ? src_slice_origin_step_idx
+                                       : src_slice_origin_step_idx + GetSrcCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src_desc, src_coord_, adjusted_step);
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrcSliceWindow(const SrcDesc& src_desc,
+                                       const Index& src_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx =
+            SrcResetCoordinateAfterRun ? src_slice_origin_step_idx
+                                       : src_slice_origin_step_idx + GetSrcCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src_desc, src_coord_, adjusted_step);
+    }
+
+    // dst_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveDstSliceWindow(const DstDesc& dst_desc,
+                                       const Dst0Desc dst0_desc,
+                                       const Dst1Desc dst1_desc,
+                                       const Index& dst_slice_origin_step_idx)
+    {
+        // if dst coord was not reset by RunWrite(), then need to adjust the step here
+        const auto adjusted_step_idx =
+            DstResetCoordinateAfterRun ? dst_slice_origin_step_idx
+                                       : dst_slice_origin_step_idx + GetDstCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(dst_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(dst_desc, dst_coord_, adjusted_step);
+        move_tensor_coordinate(dst0_desc, dst0_coord_, adjusted_step);
+        move_tensor_coordinate(dst1_desc, dst1_coord_, adjusted_step);
+    }
+
+    __device__ static constexpr auto GetSrcThreadScratchDescriptor()
+    {
+        constexpr auto src_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<SrcVectorDim, SrcScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto src_access_lengths = SliceLengths{} / src_scalar_per_access;
+
+        constexpr auto src_access_lengths_and_vector_length = container_push_back(
+            sequence_to_tuple_of_number(src_access_lengths), Number<SrcScalarPerVector>{});
+
+        // 1st stage of transforms
+        constexpr auto desc0 =
+            make_naive_tensor_descriptor_packed(src_access_lengths_and_vector_length);
+
+        // 2nd stage of transforms
+        constexpr auto transforms = generate_tuple(
+            [&](auto i) {
+                if constexpr(i == SrcVectorDim)
+                {
+                    return make_merge_transform_v3_division_mod(
+                        make_tuple(src_access_lengths_and_vector_length[i],
+                                   src_access_lengths_and_vector_length[Number<nDim>{}]));
+                }
+                else
+                {
+                    return make_pass_through_transform(src_access_lengths_and_vector_length[i]);
+                }
+            },
+            Number<nDim>{});
+
+        constexpr auto low_dim_idss = generate_tuple(
+            [&](auto i) {
+                if constexpr(i == SrcVectorDim)
+                {
+                    return Sequence<i.value, nDim>{};
+                }
+                else
+                {
+                    return Sequence<i.value>{};
+                }
+            },
+            Number<nDim>{});
+
+        constexpr auto up_dim_idss =
+            generate_tuple([&](auto i) { return Sequence<i.value>{}; }, Number<nDim>{});
+
+        return transform_tensor_descriptor(desc0, transforms, low_dim_idss, up_dim_idss);
+    }
+
+    __device__ static constexpr auto GetDstThreadScratchDescriptor()
+    {
+        // 1st stage of transforms
+        constexpr auto dst_scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<DstVectorDim, DstScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto dst_access_lengths = SliceLengths{} / dst_scalar_per_access;
+
+        constexpr auto dst_access_lengths_and_vector_length = container_push_back(
+            sequence_to_tuple_of_number(dst_access_lengths), Number<DstScalarPerVector>{});
+
+        constexpr auto desc0 =
+            make_naive_tensor_descriptor_packed(dst_access_lengths_and_vector_length);
+
+        // 2nd stage of transforms
+        constexpr auto transforms = generate_tuple(
+            [&](auto i) {
+                if constexpr(i == DstVectorDim)
+                {
+                    return make_merge_transform_v3_division_mod(
+                        make_tuple(dst_access_lengths_and_vector_length[i],
+                                   dst_access_lengths_and_vector_length[Number<nDim>{}]));
+                }
+                else
+                {
+                    return make_pass_through_transform(dst_access_lengths_and_vector_length[i]);
+                }
+            },
+            Number<nDim>{});
+
+        constexpr auto low_dim_idss = generate_tuple(
+            [&](auto i) {
+                if constexpr(i == DstVectorDim)
+                {
+                    return Sequence<i.value, nDim>{};
+                }
+                else
+                {
+                    return Sequence<i.value>{};
+                }
+            },
+            Number<nDim>{});
+
+        constexpr auto up_dim_idss =
+            generate_tuple([&](auto i) { return Sequence<i.value>{}; }, Number<nDim>{});
+
+        return transform_tensor_descriptor(desc0, transforms, low_dim_idss, up_dim_idss);
+    }
+
+    private:
+    static constexpr auto src_thread_scratch_desc_ = decltype(GetSrcThreadScratchDescriptor()){};
+    static constexpr auto dst_thread_scratch_desc_ = decltype(GetDstThreadScratchDescriptor()){};
+
+    StaticTensorTupleOfVectorBuffer<AddressSpaceEnum_t::Vgpr,
+                                    SrcData,
+                                    SrcScalarPerVector,
+                                    decltype(src_thread_scratch_desc_),
+                                    true>
+        src_thread_scratch_;
+
+    StaticTensorTupleOfVectorBuffer<AddressSpaceEnum_t::Vgpr,
+                                    DstData,
+                                    DstScalarPerVector,
+                                    decltype(dst_thread_scratch_desc_),
+                                    true>
+        dst_thread_scratch_;
+
+    SrcCoord src_coord_;
+    DstCoord dst_coord_;
+    const SrcElementwiseOperation src_element_op_;
+    const DstElementwiseOperation dst_element_op_;
+};
+
+} // namespace ck
+#endif

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v4r1.hpp

+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V4R1_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V4R1_HPP
+
+#include "common_header.hpp"
+#include "tensor_descriptor.hpp"
+#include "tensor_descriptor_helper.hpp"
+
+namespace ck {
+// Assume:
+//   1. src:
+//     1. SrcDesc is known at compile-time
+//     2. SrcBuffer is DynamicBuffer
+//     3. src_ref_idx is known at run-time
+//     4. SrcRefToOriginDisplacement is known at compile-time
+//     5. use #-step
+//   2. dst:
+//     1. DstDesc is known at compile-time
+//     2. DstBuffer is StaticBuffer
+//     3. DstOriginIdx is known at compile-time
+//     4. use direct address calculation
+//   3. vector access on src
+template <typename SrcData,
+          typename DstData,
+          typename SrcDesc,
+          typename DstDesc,
+          typename SliceLengths,
+          typename DimAccessOrder,
+          typename SrcVectorTensorLengths,
+          typename SrcVectorTensorContiguousDimOrder,
+          typename enable_if<SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
+                             bool>::type = false>
+struct ThreadwiseTensorSliceTransfer_v4r1
+{
+    static constexpr auto I0 = Number<0>{};
+    static constexpr auto I1 = Number<1>{};
+
+    static constexpr index_t nDim = SliceLengths::Size();
+
+    using Index = MultiIndex<nDim>;
+
+    using SrcCoord = decltype(make_tensor_coordinate(SrcDesc{}, Index{}));
+
+    using SrcCoordStep = decltype(make_tensor_coordinate_step(SrcDesc{}, Index{}));
+
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v4r1(const Index& src_ref_idx)
+        : src_ref_coord_(make_tensor_coordinate(SrcDesc{}, src_ref_idx))
+    {
+        static_assert(SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
+                      "wrong! SrcDesc and DstDesc need to known at compile-time");
+
+        static_for<0, nDim, 1>{}([](auto i) {
+            static_assert(SliceLengths::At(i) % SrcVectorTensorLengths::At(i) == 0, "wrong!");
+        });
+    }
+
+    template <typename SrcRefToOriginDisplacement,
+              typename DstOriginIdx,
+              typename SrcBuffer,
+              typename DstBuffer>
+    __device__ void Run(const SrcDesc&,
+                        const SrcRefToOriginDisplacement&,
+                        const SrcBuffer& src_buf,
+                        const DstDesc&,
+                        const DstOriginIdx&,
+                        DstBuffer& dst_buf) const
+    {
+        static_assert(SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
+                      "wrong! SrcDesc and DstDesc need to known at compile-time");
+
+        static_assert(
+            is_same<remove_cvref_t<typename SrcBuffer::type>, remove_cvref_t<SrcData>>::value &&
+                is_same<remove_cvref_t<typename DstBuffer::type>, remove_cvref_t<DstData>>::value,
+            "wrong! SrcBuffer or DstBuffer data type is wrong");
+
+        static_assert(DstBuffer::IsStaticBuffer(), "wrong! DstBuffer need to be StaticBuffer");
+
+        static_assert(is_known_at_compile_time<remove_cvref_t<SrcRefToOriginDisplacement>>::value &&
+                          is_known_at_compile_time<remove_cvref_t<DstOriginIdx>>::value,
+                      "wrong! SrcOriginToRefDistance and DstOriginToRefDistance need to be known "
+                      "at compile-time");
+
+        // SrcDesc and DstDesc are known at compile-time
+        constexpr auto src_desc = remove_cvref_t<SrcDesc>{};
+        constexpr auto dst_desc = remove_cvref_t<DstDesc>{};
+
+        // SrcOriginToRefDisttance and DstOriginToRefDistance are known at compile-time
+        constexpr auto src_ref_to_origin_disp_idx = to_multi_index(SrcRefToOriginDisplacement{});
+        constexpr auto dst_origin_idx             = to_multi_index(DstOriginIdx{});
+
+        // tensor descriptor for src_vector
+        constexpr auto src_vector_tensor_lengths = SrcVectorTensorLengths{};
+
+        constexpr auto src_vector_tensor_strides = container_reorder_given_old2new(
+            container_reverse_exclusive_scan(
+                container_reorder_given_new2old(src_vector_tensor_lengths,
+                                                SrcVectorTensorContiguousDimOrder{}),
+                math::multiplies{},
+                I1),
+            SrcVectorTensorContiguousDimOrder{});
+
+        constexpr auto src_vector_desc =
+            make_naive_tensor_descriptor(sequence_to_tuple_of_number(src_vector_tensor_lengths),
+                                         sequence_to_tuple_of_number(src_vector_tensor_strides));
+
+        // access order and lengths
+        constexpr auto access_lengths = SliceLengths{} / src_vector_tensor_lengths;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        static_ford<decltype(ordered_access_lengths)>{}([&](auto ordered_access_idx) {
+            // position in slice window
+            constexpr auto data_to_origin_disp_idx =
+                ordered_access_idx.ReorderGivenOld2New(dim_access_order) *
+                src_vector_tensor_lengths;
+
+            // src coordinate at starting point of src_vector
+            constexpr auto src_ref_to_data_disp_idx =
+                src_ref_to_origin_disp_idx + data_to_origin_disp_idx;
+
+            constexpr auto src_ref_to_data_disp_coord_step =
+                make_tensor_coordinate_step(src_desc, src_ref_to_data_disp_idx);
+
+            auto src_data_coord = src_ref_coord_;
+
+            move_tensor_coordinate(src_desc, src_data_coord, src_ref_to_data_disp_coord_step);
+
+            vector_type_maker_t<SrcData, src_vector_desc.GetElementSpaceSize()> src_vector;
+
+            using src_vector_t = typename decltype(src_vector)::type;
+
+            const bool is_src_valid = coordinate_has_valid_offset_assuming_visible_index_is_valid(
+                src_desc, src_data_coord);
+
+            // copy data from src_buf into src_vector
+            src_vector.template AsType<src_vector_t>()(I0) =
+                src_buf.template Get<src_vector_t>(src_data_coord.GetOffset(), is_src_valid);
+
+            // copy data from src_vector into dst_buf (also cast from SrcData to DstData)
+            static_ford<SrcVectorTensorLengths>{}([&](auto src_vector_idx_) {
+                constexpr auto src_vector_idx = to_multi_index(src_vector_idx_);
+
+                constexpr index_t src_vector_offset =
+                    src_vector_desc.CalculateOffset(src_vector_idx);
+
+                constexpr index_t dst_offset = dst_desc.CalculateOffset(
+                    dst_origin_idx + data_to_origin_disp_idx + src_vector_idx);
+
+                dst_buf(Number<dst_offset>{}) = type_convert<DstData>(
+                    src_vector.template AsType<DstData>()[Number<src_vector_offset>{}]);
+            });
+        });
+    }
+
+    template <typename SrcSliceMoveStepIdx>
+    __device__ void MoveSrcSliceWindow(const SrcDesc&,
+                                       const SrcSliceMoveStepIdx& src_slice_move_step_idx)
+    {
+        constexpr auto src_desc = SrcDesc{};
+
+        const auto src_slice_move_step_iter =
+            make_tensor_coordinate_step(src_desc, to_multi_index(src_slice_move_step_idx));
+
+        move_tensor_coordinate(SrcDesc{}, src_ref_coord_, src_slice_move_step_iter);
+    }
+
+    private:
+    SrcCoord src_ref_coord_;
+};
+
+} // namespace ck
+#endif

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v2.hpp → composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v5r1.hpp

-#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V2_HPP
-#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V2_HPP
+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V5R1_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V5R1_HPP
 
 #include "common_header.hpp"
 #include "tensor_descriptor.hpp"
@@ -30,7 +30,7 @@ template <typename SliceLengths,
           bool DstResetCoordinateAfterRun> // control whether to move back dst coordinate after each
                                            // RunWrite(),  will be fused with MoveDstSliceWindow to
                                            // save addr computation
-struct ThreadwiseTensorSliceTransfer_v3r1
+struct ThreadwiseTensorSliceTransfer_v5r1
 {
     static constexpr auto I0 = Number<0>{};
     static constexpr auto I1 = Number<1>{};
@@ -44,7 +44,7 @@ struct ThreadwiseTensorSliceTransfer_v3r1
     using SrcCoordStep = decltype(make_tensor_coordinate_step(SrcDesc{}, Index{}));
     using DstCoordStep = decltype(make_tensor_coordinate_step(DstDesc{}, Index{}));
 
-    __device__ constexpr ThreadwiseTensorSliceTransfer_v3r1(const SrcDesc& src_desc,
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v5r1(const SrcDesc& src_desc,
                                                             const Index& src_slice_origin,
                                                             const DstDesc& dst_desc,
                                                             const Index& dst_slice_origin)
@@ -608,169 +608,5 @@ struct ThreadwiseTensorSliceTransfer_v3r1
     DstCoord dst_coord_;
 };
 
-// Assume:
-//   1. src:
-//     1. SrcDesc is known at compile-time
-//     2. SrcBuffer is DynamicBuffer
-//     3. src_ref_idx is known at run-time
-//     4. SrcRefToOriginDisplacement is known at compile-time
-//     5. use #-step
-//   2. dst:
-//     1. DstDesc is known at compile-time
-//     2. DstBuffer is StaticBuffer
-//     3. DstOriginIdx is known at compile-time
-//     4. use direct address calculation
-//   3. vector access on src
-template <typename SrcData,
-          typename DstData,
-          typename SrcDesc,
-          typename DstDesc,
-          typename SliceLengths,
-          typename DimAccessOrder,
-          typename SrcVectorTensorLengths,
-          typename SrcVectorTensorContiguousDimOrder,
-          typename enable_if<SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
-                             bool>::type = false>
-struct ThreadwiseTensorSliceTransfer_v4r1
-{
-    static constexpr auto I0 = Number<0>{};
-    static constexpr auto I1 = Number<1>{};
-
-    static constexpr index_t nDim = SliceLengths::Size();
-
-    using Index = MultiIndex<nDim>;
-
-    using SrcCoord = decltype(make_tensor_coordinate(SrcDesc{}, Index{}));
-
-    using SrcCoordStep = decltype(make_tensor_coordinate_step(SrcDesc{}, Index{}));
-
-    __device__ constexpr ThreadwiseTensorSliceTransfer_v4r1(const Index& src_ref_idx)
-        : src_ref_coord_(make_tensor_coordinate(SrcDesc{}, src_ref_idx))
-    {
-        static_assert(SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
-                      "wrong! SrcDesc and DstDesc need to known at compile-time");
-
-        static_for<0, nDim, 1>{}([](auto i) {
-            static_assert(SliceLengths::At(i) % SrcVectorTensorLengths::At(i) == 0, "wrong!");
-        });
-    }
-
-    template <typename SrcRefToOriginDisplacement,
-              typename DstOriginIdx,
-              typename SrcBuffer,
-              typename DstBuffer>
-    __device__ void Run(const SrcDesc&,
-                        const SrcRefToOriginDisplacement&,
-                        const SrcBuffer& src_buf,
-                        const DstDesc&,
-                        const DstOriginIdx&,
-                        DstBuffer& dst_buf) const
-    {
-        static_assert(SrcDesc::IsKnownAtCompileTime() && DstDesc::IsKnownAtCompileTime(),
-                      "wrong! SrcDesc and DstDesc need to known at compile-time");
-
-        static_assert(
-            is_same<remove_cvref_t<typename SrcBuffer::type>, remove_cvref_t<SrcData>>::value &&
-                is_same<remove_cvref_t<typename DstBuffer::type>, remove_cvref_t<DstData>>::value,
-            "wrong! SrcBuffer or DstBuffer data type is wrong");
-
-        static_assert(DstBuffer::IsStaticBuffer(), "wrong! DstBuffer need to be StaticBuffer");
-
-        static_assert(is_known_at_compile_time<remove_cvref_t<SrcRefToOriginDisplacement>>::value &&
-                          is_known_at_compile_time<remove_cvref_t<DstOriginIdx>>::value,
-                      "wrong! SrcOriginToRefDistance and DstOriginToRefDistance need to be known "
-                      "at compile-time");
-
-        // SrcDesc and DstDesc are known at compile-time
-        constexpr auto src_desc = remove_cvref_t<SrcDesc>{};
-        constexpr auto dst_desc = remove_cvref_t<DstDesc>{};
-
-        // SrcOriginToRefDisttance and DstOriginToRefDistance are known at compile-time
-        constexpr auto src_ref_to_origin_disp_idx = to_multi_index(SrcRefToOriginDisplacement{});
-        constexpr auto dst_origin_idx             = to_multi_index(DstOriginIdx{});
-
-        // tensor descriptor for src_vector
-        constexpr auto src_vector_tensor_lengths = SrcVectorTensorLengths{};
-
-        constexpr auto src_vector_tensor_strides = container_reorder_given_old2new(
-            container_reverse_exclusive_scan(
-                container_reorder_given_new2old(src_vector_tensor_lengths,
-                                                SrcVectorTensorContiguousDimOrder{}),
-                math::multiplies{},
-                I1),
-            SrcVectorTensorContiguousDimOrder{});
-
-        constexpr auto src_vector_desc =
-            make_naive_tensor_descriptor(sequence_to_tuple_of_number(src_vector_tensor_lengths),
-                                         sequence_to_tuple_of_number(src_vector_tensor_strides));
-
-        // access order and lengths
-        constexpr auto access_lengths = SliceLengths{} / src_vector_tensor_lengths;
-
-        constexpr auto dim_access_order = DimAccessOrder{};
-
-        constexpr auto ordered_access_lengths =
-            container_reorder_given_new2old(access_lengths, dim_access_order);
-
-        static_ford<decltype(ordered_access_lengths)>{}([&](auto ordered_access_idx) {
-            // position in slice window
-            constexpr auto data_to_origin_disp_idx =
-                ordered_access_idx.ReorderGivenOld2New(dim_access_order) *
-                src_vector_tensor_lengths;
-
-            // src coordinate at starting point of src_vector
-            constexpr auto src_ref_to_data_disp_idx =
-                src_ref_to_origin_disp_idx + data_to_origin_disp_idx;
-
-            constexpr auto src_ref_to_data_disp_coord_step =
-                make_tensor_coordinate_step(src_desc, src_ref_to_data_disp_idx);
-
-            auto src_data_coord = src_ref_coord_;
-
-            move_tensor_coordinate(src_desc, src_data_coord, src_ref_to_data_disp_coord_step);
-
-            vector_type_maker_t<SrcData, src_vector_desc.GetElementSpaceSize()> src_vector;
-
-            using src_vector_t = typename decltype(src_vector)::type;
-
-            const bool is_src_valid = coordinate_has_valid_offset_assuming_visible_index_is_valid(
-                src_desc, src_data_coord);
-
-            // copy data from src_buf into src_vector
-            src_vector.template AsType<src_vector_t>()(I0) =
-                src_buf.template Get<src_vector_t>(src_data_coord.GetOffset(), is_src_valid);
-
-            // copy data from src_vector into dst_buf (also cast from SrcData to DstData)
-            static_ford<SrcVectorTensorLengths>{}([&](auto src_vector_idx_) {
-                constexpr auto src_vector_idx = to_multi_index(src_vector_idx_);
-
-                constexpr index_t src_vector_offset =
-                    src_vector_desc.CalculateOffset(src_vector_idx);
-
-                constexpr index_t dst_offset = dst_desc.CalculateOffset(
-                    dst_origin_idx + data_to_origin_disp_idx + src_vector_idx);
-
-                dst_buf(Number<dst_offset>{}) = type_convert<DstData>(
-                    src_vector.template AsType<DstData>()[Number<src_vector_offset>{}]);
-            });
-        });
-    }
-
-    template <typename SrcSliceMoveStepIdx>
-    __device__ void MoveSrcSliceWindow(const SrcDesc&,
-                                       const SrcSliceMoveStepIdx& src_slice_move_step_idx)
-    {
-        constexpr auto src_desc = SrcDesc{};
-
-        const auto src_slice_move_step_iter =
-            make_tensor_coordinate_step(src_desc, to_multi_index(src_slice_move_step_idx));
-
-        move_tensor_coordinate(SrcDesc{}, src_ref_coord_, src_slice_move_step_iter);
-    }
-
-    private:
-    SrcCoord src_ref_coord_;
-};
-
 } // namespace ck
 #endif

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r1.hpp

+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R1_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R1_HPP
+
+#include "common_header.hpp"
+#include "tensor_descriptor.hpp"
+#include "tensor_descriptor_helper.hpp"
+
+namespace ck {
+
+// Do following things to avoid "alloca" in LLVM-IR, which would cause scratch memory
+// and sometimes useless instructions:
+//   1. Don't save a reference to tensor descriptor in class, pass in tensor descriptor as argument
+//   instead
+//   2. Don't construct a new tensor coordinate everytime when using it, update and reuse the same
+//   tensor coordinate instead
+//   3. Don't use a pointer to VGPR buffer, use vector instead
+
+// Assume:
+//   1. src_desc and dst_desc are not known at compile-time
+//   2. SrcBuffer and DstBuffer are DynamicBuffer
+//   3. src_slice_origin and dst_slice_origin are not known at compile-time,
+template <typename SrcData,
+          typename DstData,
+          typename SrcDesc,
+          typename DstDesc,
+          typename ElementwiseOperation,
+          typename SliceLengths,
+          typename DimAccessOrder,
+          index_t VectorDim,
+          index_t ScalarPerVector,
+          InMemoryDataOperationEnum_t DstInMemOp,
+          bool SrcResetCoordinateAfterRun,
+          bool DstResetCoordinateAfterRun>
+struct ThreadwiseTensorSliceTransfer_v6r1
+{
+    static constexpr index_t nDim = SliceLengths::Size();
+
+    using Index = MultiIndex<nDim>;
+
+    using SrcCoord = decltype(make_tensor_coordinate(SrcDesc{}, Index{}));
+    using DstCoord = decltype(make_tensor_coordinate(DstDesc{}, Index{}));
+
+    using SrcCoordStep = decltype(make_tensor_coordinate_step(SrcDesc{}, Index{}));
+    using DstCoordStep = decltype(make_tensor_coordinate_step(DstDesc{}, Index{}));
+
+    static constexpr auto I0 = Number<0>{};
+
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v6r1(const SrcDesc& src_desc,
+                                                            const Index& src_slice_origin,
+                                                            const DstDesc& dst_desc,
+                                                            const Index& dst_slice_origin,
+                                                            const ElementwiseOperation& element_op)
+        : src_coord_(make_tensor_coordinate(src_desc, src_slice_origin)),
+          dst_coord_(make_tensor_coordinate(dst_desc, dst_slice_origin)),
+          element_op_(element_op)
+    {
+        static_assert(SliceLengths::At(Number<VectorDim>{}) % ScalarPerVector == 0,
+                      "wrong! cannot evenly divide");
+    }
+
+    __device__ void SetSrcSliceOrigin(const SrcDesc& src_desc, const Index& src_slice_origin_idx)
+    {
+        src_coord_ = make_tensor_coordinate(src_desc, src_slice_origin_idx);
+    }
+
+    __device__ void SetDstSliceOrigin(const DstDesc& dst_desc, const Index& dst_slice_origin_idx)
+    {
+        dst_coord_ = make_tensor_coordinate(dst_desc, dst_slice_origin_idx);
+    }
+
+    template <typename SrcBuffer, typename DstBuffer>
+    __device__ void Run(const SrcDesc& src_desc,
+                        const SrcBuffer& src_buf,
+                        const DstDesc& dst_desc,
+                        DstBuffer& dst_buf)
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        auto make_forward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index forward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        forward_step_idx(j) = (i.value == j.value) ? scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, forward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        auto make_backward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index backward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        backward_step_idx(j) = (i.value == j.value) ? -scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, backward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        // make forward steps
+        const auto src_forward_steps = make_forward_steps(src_desc);
+        const auto dst_forward_steps = make_forward_steps(dst_desc);
+
+        // make backward steps
+        const auto src_backward_steps = make_backward_steps(src_desc);
+        const auto dst_backward_steps = make_backward_steps(dst_desc);
+
+        // loop over slice window
+        static_ford<decltype(ordered_access_lengths)>{}([&](auto ordered_access_idx) {
+            // judge move forward or move backward
+            constexpr auto forward_sweep = [&]() {
+                StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+                forward_sweep_(I0) = true;
+
+                static_for<1, nDim, 1>{}([&](auto i) {
+                    index_t tmp = ordered_access_idx[I0];
+
+                    static_for<1, i, 1>{}([&](auto j) {
+                        tmp = tmp * ordered_access_lengths[j] + ordered_access_idx[j];
+                    });
+
+                    forward_sweep_(i) = tmp % 2 == 0;
+                });
+
+                return forward_sweep_;
+            }();
+
+            using src_vector_type = vector_type_maker_t<SrcData, ScalarPerVector>;
+            using src_vector_t    = typename src_vector_type::type;
+
+            using dst_vector_type = vector_type_maker_t<DstData, ScalarPerVector>;
+            using dst_vector_t    = typename dst_vector_type::type;
+
+            const bool is_src_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src_desc, src_coord_);
+
+            // copy data from src_buf into src_vector_container
+            auto src_vector_container = src_vector_type{
+                src_buf.template Get<src_vector_t>(src_coord_.GetOffset(), is_src_valid)};
+
+            auto dst_vector_container = dst_vector_type{};
+
+            // apply pointwise operation
+            static_for<0, ScalarPerVector, 1>{}([&](auto i) {
+                element_op_(dst_vector_container.template AsType<DstData>()(i),
+                            src_vector_container.template AsType<SrcData>()[i]);
+            });
+
+            const bool is_dst_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(dst_desc, dst_coord_);
+
+            // copy data from dst_vector into dst_buf
+            if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::Set)
+            {
+                dst_buf.template Set<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+            else if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::AtomicAdd)
+            {
+                dst_buf.template AtomicAdd<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+
+            constexpr auto move_on_dim = [&]() constexpr
+            {
+                StaticallyIndexedArray<bool, nDim> move_on_dim_;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    move_on_dim_(i) = ordered_access_idx[i] < ordered_access_lengths[i] - 1;
+
+                    static_for<i + 1, nDim, 1>{}([&](auto j) {
+                        move_on_dim_(i) &= ordered_access_idx[j] == ordered_access_lengths[j] - 1;
+                    });
+                });
+
+                return move_on_dim_;
+            }
+            ();
+
+            // move coordinate
+            static_for<0, nDim, 1>{}([&](auto i) {
+                if constexpr(move_on_dim[i])
+                {
+                    if constexpr(forward_sweep[i])
+                    {
+                        move_tensor_coordinate(
+                            src_desc, src_coord_, src_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_forward_steps[dim_access_order[i]]);
+                    }
+                    else
+                    {
+                        move_tensor_coordinate(
+                            src_desc, src_coord_, src_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_backward_steps[dim_access_order[i]]);
+                    }
+                }
+            });
+        });
+
+        // move coordinate back to slice origin (or not)
+        if constexpr(SrcResetCoordinateAfterRun)
+        {
+            const auto src_reset_step =
+                make_tensor_coordinate_step(src_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src_desc, src_coord_, src_reset_step);
+        }
+
+        if constexpr(DstResetCoordinateAfterRun)
+        {
+            const auto dst_reset_step =
+                make_tensor_coordinate_step(dst_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(dst_desc, dst_coord_, dst_reset_step);
+        }
+    }
+
+    __device__ static constexpr auto GetCoordinateResetStep()
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        // judge move forward or move backward during the last iteration
+        constexpr auto forward_sweep = [&]() {
+            StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+            forward_sweep_(I0) = true;
+
+            static_for<1, nDim, 1>{}([&](auto i) {
+                index_t tmp = ordered_access_lengths[I0] - 1;
+
+                static_for<1, i, 1>{}([&](auto j) {
+                    tmp = tmp * ordered_access_lengths[j] + ordered_access_lengths[j] - 1;
+                });
+
+                forward_sweep_(i) = tmp % 2 == 0;
+            });
+
+            return forward_sweep_;
+        }();
+
+        // calculate data index after last iteration in Run(), if it has not being reset
+        constexpr auto data_idx = [&]() {
+            Index ordered_idx;
+
+            static_for<0, nDim, 1>{}([&](auto i) {
+                ordered_idx(i) = forward_sweep[i] ? ordered_access_lengths[i] - 1 : 0;
+            });
+
+            return container_reorder_given_old2new(ordered_idx, dim_access_order) *
+                   scalar_per_access;
+        }();
+
+        //
+        constexpr auto reset_data_step = [&]() {
+            Index reset_data_step_;
+
+            static_for<0, nDim, 1>{}([&](auto i) { reset_data_step_(i) = -data_idx[i]; });
+
+            return reset_data_step_;
+        }();
+
+        return reset_data_step;
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrcSliceWindow(const SrcDesc& src_desc,
+                                       const Index& src_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = SrcResetCoordinateAfterRun
+                                           ? src_slice_origin_step_idx
+                                           : src_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src_desc, src_coord_, adjusted_step);
+    }
+
+    // dst_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveDstSliceWindow(const DstDesc& dst_desc,
+                                       const Index& dst_slice_origin_step_idx)
+    {
+        // if dst coord was not reset by Run(), then need to adjust the step here
+        const auto adjusted_step_idx = DstResetCoordinateAfterRun
+                                           ? dst_slice_origin_step_idx
+                                           : dst_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(dst_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(dst_desc, dst_coord_, adjusted_step);
+    }
+
+    private:
+    SrcCoord src_coord_;
+    DstCoord dst_coord_;
+    const ElementwiseOperation element_op_;
+};
+
+} // namespace ck
+#endif

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r2.hpp

+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R2_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R2_HPP
+
+#include "common_header.hpp"
+#include "tensor_descriptor.hpp"
+#include "tensor_descriptor_helper.hpp"
+
+namespace ck {
+
+// Do following things to avoid "alloca" in LLVM-IR, which would cause scratch memory
+// and sometimes useless instructions:
+//   1. Don't save a reference to tensor descriptor in class, pass in tensor descriptor as argument
+//   instead
+//   2. Don't construct a new tensor coordinate everytime when using it, update and reuse the same
+//   tensor coordinate instead
+//   3. Don't use a pointer to VGPR buffer, use vector instead
+
+// Assume:
+//   1. src0_desc and dst_desc are not known at compile-time
+//   2. SrcBuffer and DstBuffer are DynamicBuffer
+//   3. src_slice_origin and dst_slice_origin are not known at compile-time,
+template <typename Src0Data,
+          typename Src1Data,
+          typename DstData,
+          typename Src0Desc,
+          typename Src1Desc,
+          typename DstDesc,
+          typename ElementwiseOperation,
+          typename SliceLengths,
+          typename DimAccessOrder,
+          index_t VectorDim,
+          index_t ScalarPerVector,
+          InMemoryDataOperationEnum_t DstInMemOp,
+          bool Src0ResetCoordinateAfterRun,
+          bool Src1ResetCoordinateAfterRun,
+          bool DstResetCoordinateAfterRun>
+struct ThreadwiseTensorSliceTransfer_v6r2
+{
+    static constexpr index_t nDim = SliceLengths::Size();
+
+    using Index = MultiIndex<nDim>;
+
+    using Src0Coord = decltype(make_tensor_coordinate(Src0Desc{}, Index{}));
+    using Src1Coord = decltype(make_tensor_coordinate(Src1Desc{}, Index{}));
+    using DstCoord  = decltype(make_tensor_coordinate(DstDesc{}, Index{}));
+
+    using Src0CoordStep = decltype(make_tensor_coordinate_step(Src0Desc{}, Index{}));
+    using Src1CoordStep = decltype(make_tensor_coordinate_step(Src1Desc{}, Index{}));
+    using DstCoordStep  = decltype(make_tensor_coordinate_step(DstDesc{}, Index{}));
+
+    static constexpr auto I0 = Number<0>{};
+
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v6r2(const Src0Desc& src0_desc,
+                                                            const Index& src0_slice_origin,
+                                                            const Src1Desc& src1_desc,
+                                                            const Index& src1_slice_origin,
+                                                            const DstDesc& dst_desc,
+                                                            const Index& dst_slice_origin,
+                                                            const ElementwiseOperation& element_op)
+        : src0_coord_(make_tensor_coordinate(src0_desc, src0_slice_origin)),
+          src1_coord_(make_tensor_coordinate(src1_desc, src1_slice_origin)),
+          dst_coord_(make_tensor_coordinate(dst_desc, dst_slice_origin)),
+          element_op_(element_op)
+    {
+        static_assert(SliceLengths::At(Number<VectorDim>{}) % ScalarPerVector == 0,
+                      "wrong! cannot evenly divide");
+    }
+
+    __device__ void SetSrc0SliceOrigin(const Src0Desc& src0_desc,
+                                       const Index& src0_slice_origin_idx)
+    {
+        src0_coord_ = make_tensor_coordinate(src0_desc, src0_slice_origin_idx);
+    }
+
+    __device__ void SetSrc1SliceOrigin(const Src1Desc& src1_desc,
+                                       const Index& src1_slice_origin_idx)
+    {
+        src1_coord_ = make_tensor_coordinate(src1_desc, src1_slice_origin_idx);
+    }
+
+    __device__ void SetDstSliceOrigin(const DstDesc& dst_desc, const Index& dst_slice_origin_idx)
+    {
+        dst_coord_ = make_tensor_coordinate(dst_desc, dst_slice_origin_idx);
+    }
+
+    template <typename Src0Buffer, typename Src1Buffer, typename DstBuffer>
+    __device__ void Run(const Src0Desc& src0_desc,
+                        const Src0Buffer& src0_buf,
+                        const Src1Desc& src1_desc,
+                        const Src1Buffer& src1_buf,
+                        const DstDesc& dst_desc,
+                        DstBuffer& dst_buf)
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        auto make_forward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index forward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        forward_step_idx(j) = (i.value == j.value) ? scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, forward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        auto make_backward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index backward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        backward_step_idx(j) = (i.value == j.value) ? -scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, backward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        // make forward steps
+        const auto src0_forward_steps = make_forward_steps(src0_desc);
+        const auto src1_forward_steps = make_forward_steps(src1_desc);
+        const auto dst_forward_steps  = make_forward_steps(dst_desc);
+
+        // make backward steps
+        const auto src0_backward_steps = make_backward_steps(src0_desc);
+        const auto src1_backward_steps = make_backward_steps(src1_desc);
+        const auto dst_backward_steps  = make_backward_steps(dst_desc);
+
+        // loop over slice window
+        static_ford<decltype(ordered_access_lengths)>{}([&](auto ordered_access_idx) {
+            // judge move forward or move backward
+            constexpr auto forward_sweep = [&]() {
+                StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+                forward_sweep_(I0) = true;
+
+                static_for<1, nDim, 1>{}([&](auto i) {
+                    index_t tmp = ordered_access_idx[I0];
+
+                    static_for<1, i, 1>{}([&](auto j) {
+                        tmp = tmp * ordered_access_lengths[j] + ordered_access_idx[j];
+                    });
+
+                    forward_sweep_(i) = tmp % 2 == 0;
+                });
+
+                return forward_sweep_;
+            }();
+
+            using src0_vector_type = vector_type_maker_t<Src0Data, ScalarPerVector>;
+            using src0_vector_t    = typename src0_vector_type::type;
+
+            using src1_vector_type = vector_type_maker_t<Src1Data, ScalarPerVector>;
+            using src1_vector_t    = typename src1_vector_type::type;
+
+            using dst_vector_type = vector_type_maker_t<DstData, ScalarPerVector>;
+            using dst_vector_t    = typename dst_vector_type::type;
+
+            const bool is_src0_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src0_desc, src0_coord_);
+
+            const bool is_src1_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src1_desc, src1_coord_);
+
+            // copy data from src0_buf into src0_vector_container
+            auto src0_vector_container = src0_vector_type{
+                src0_buf.template Get<src0_vector_t>(src0_coord_.GetOffset(), is_src0_valid)};
+
+            auto src1_vector_container = src1_vector_type{
+                src1_buf.template Get<src1_vector_t>(src1_coord_.GetOffset(), is_src1_valid)};
+
+            auto dst_vector_container = dst_vector_type{};
+
+            // apply pointwise operation
+            static_for<0, ScalarPerVector, 1>{}([&](auto i) {
+                element_op_(dst_vector_container.template AsType<DstData>()(i),
+                            src0_vector_container.template AsType<Src0Data>()[i],
+                            src1_vector_container.template AsType<Src1Data>()[i]);
+            });
+
+            const bool is_dst_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(dst_desc, dst_coord_);
+
+            // copy data from dst_vector into dst_buf
+            if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::Set)
+            {
+                dst_buf.template Set<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+            else if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::AtomicAdd)
+            {
+                dst_buf.template AtomicAdd<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+
+            constexpr auto move_on_dim = [&]() constexpr
+            {
+                StaticallyIndexedArray<bool, nDim> move_on_dim_;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    move_on_dim_(i) = ordered_access_idx[i] < ordered_access_lengths[i] - 1;
+
+                    static_for<i + 1, nDim, 1>{}([&](auto j) {
+                        move_on_dim_(i) &= ordered_access_idx[j] == ordered_access_lengths[j] - 1;
+                    });
+                });
+
+                return move_on_dim_;
+            }
+            ();
+
+            // move coordinate
+            static_for<0, nDim, 1>{}([&](auto i) {
+                if constexpr(move_on_dim[i])
+                {
+                    if constexpr(forward_sweep[i])
+                    {
+                        move_tensor_coordinate(
+                            src0_desc, src0_coord_, src0_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src1_desc, src1_coord_, src1_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_forward_steps[dim_access_order[i]]);
+                    }
+                    else
+                    {
+                        move_tensor_coordinate(
+                            src0_desc, src0_coord_, src0_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src1_desc, src1_coord_, src1_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_backward_steps[dim_access_order[i]]);
+                    }
+                }
+            });
+        });
+
+        // move coordinate back to slice origin (or not)
+        if constexpr(Src0ResetCoordinateAfterRun)
+        {
+            const auto src0_reset_step =
+                make_tensor_coordinate_step(src0_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src0_desc, src0_coord_, src0_reset_step);
+        }
+
+        if constexpr(Src1ResetCoordinateAfterRun)
+        {
+            const auto src1_reset_step =
+                make_tensor_coordinate_step(src1_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src1_desc, src1_coord_, src1_reset_step);
+        }
+
+        if constexpr(DstResetCoordinateAfterRun)
+        {
+            const auto dst_reset_step =
+                make_tensor_coordinate_step(dst_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(dst_desc, dst_coord_, dst_reset_step);
+        }
+    }
+
+    __device__ static constexpr auto GetCoordinateResetStep()
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        // judge move forward or move backward during the last iteration
+        constexpr auto forward_sweep = [&]() {
+            StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+            forward_sweep_(I0) = true;
+
+            static_for<1, nDim, 1>{}([&](auto i) {
+                index_t tmp = ordered_access_lengths[I0] - 1;
+
+                static_for<1, i, 1>{}([&](auto j) {
+                    tmp = tmp * ordered_access_lengths[j] + ordered_access_lengths[j] - 1;
+                });
+
+                forward_sweep_(i) = tmp % 2 == 0;
+            });
+
+            return forward_sweep_;
+        }();
+
+        // calculate data index after last iteration in Run(), if it has not being reset
+        constexpr auto data_idx = [&]() {
+            Index ordered_idx;
+
+            static_for<0, nDim, 1>{}([&](auto i) {
+                ordered_idx(i) = forward_sweep[i] ? ordered_access_lengths[i] - 1 : 0;
+            });
+
+            return container_reorder_given_old2new(ordered_idx, dim_access_order) *
+                   scalar_per_access;
+        }();
+
+        //
+        constexpr auto reset_data_step = [&]() {
+            Index reset_data_step_;
+
+            static_for<0, nDim, 1>{}([&](auto i) { reset_data_step_(i) = -data_idx[i]; });
+
+            return reset_data_step_;
+        }();
+
+        return reset_data_step;
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrc0SliceWindow(const Src0Desc& src0_desc,
+                                        const Index& src0_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = Src0ResetCoordinateAfterRun
+                                           ? src0_slice_origin_step_idx
+                                           : src0_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src0_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src0_desc, src0_coord_, adjusted_step);
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrc1SliceWindow(const Src1Desc& src1_desc,
+                                        const Index& src1_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = Src1ResetCoordinateAfterRun
+                                           ? src1_slice_origin_step_idx
+                                           : src1_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src1_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src1_desc, src1_coord_, adjusted_step);
+    }
+
+    // dst_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveDstSliceWindow(const DstDesc& dst_desc,
+                                       const Index& dst_slice_origin_step_idx)
+    {
+        // if dst coord was not reset by Run(), then need to adjust the step here
+        const auto adjusted_step_idx = DstResetCoordinateAfterRun
+                                           ? dst_slice_origin_step_idx
+                                           : dst_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(dst_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(dst_desc, dst_coord_, adjusted_step);
+    }
+
+    private:
+    Src0Coord src0_coord_;
+    Src1Coord src1_coord_;
+    DstCoord dst_coord_;
+    const ElementwiseOperation element_op_;
+};
+
+} // namespace ck
+#endif

composable_kernel/include/tensor_operation/threadwise_tensor_slice_transfer_v6r3.hpp

+#ifndef CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R3_HPP
+#define CK_THREADWISE_TENSOR_SLICE_TRANSFER_V6R3_HPP
+
+#include "common_header.hpp"
+#include "tensor_descriptor.hpp"
+#include "tensor_descriptor_helper.hpp"
+
+namespace ck {
+
+// Do following things to avoid "alloca" in LLVM-IR, which would cause scratch memory
+// and sometimes useless instructions:
+//   1. Don't save a reference to tensor descriptor in class, pass in tensor descriptor as argument
+//   instead
+//   2. Don't construct a new tensor coordinate everytime when using it, update and reuse the same
+//   tensor coordinate instead
+//   3. Don't use a pointer to VGPR buffer, use vector instead
+
+// Assume:
+//   1. src0_desc and dst_desc are not known at compile-time
+//   2. SrcBuffer and DstBuffer are DynamicBuffer
+//   3. src_slice_origin and dst_slice_origin are not known at compile-time,
+template <typename Src0Data,
+          typename Src1Data,
+          typename Src2Data,
+          typename DstData,
+          typename Src0Desc,
+          typename Src1Desc,
+          typename Src2Desc,
+          typename DstDesc,
+          typename ElementwiseOperation,
+          typename SliceLengths,
+          typename DimAccessOrder,
+          index_t VectorDim,
+          index_t ScalarPerVector,
+          InMemoryDataOperationEnum_t DstInMemOp,
+          bool Src0ResetCoordinateAfterRun,
+          bool Src1ResetCoordinateAfterRun,
+          bool Src2ResetCoordinateAfterRun,
+          bool DstResetCoordinateAfterRun>
+struct ThreadwiseTensorSliceTransfer_v6r3
+{
+    static constexpr index_t nDim = SliceLengths::Size();
+
+    using Index = MultiIndex<nDim>;
+
+    using Src0Coord = decltype(make_tensor_coordinate(Src0Desc{}, Index{}));
+    using Src1Coord = decltype(make_tensor_coordinate(Src1Desc{}, Index{}));
+    using Src2Coord = decltype(make_tensor_coordinate(Src2Desc{}, Index{}));
+    using DstCoord  = decltype(make_tensor_coordinate(DstDesc{}, Index{}));
+
+    using Src0CoordStep = decltype(make_tensor_coordinate_step(Src0Desc{}, Index{}));
+    using Src1CoordStep = decltype(make_tensor_coordinate_step(Src1Desc{}, Index{}));
+    using Src2CoordStep = decltype(make_tensor_coordinate_step(Src2Desc{}, Index{}));
+    using DstCoordStep  = decltype(make_tensor_coordinate_step(DstDesc{}, Index{}));
+
+    static constexpr auto I0 = Number<0>{};
+
+    __device__ constexpr ThreadwiseTensorSliceTransfer_v6r3(const Src0Desc& src0_desc,
+                                                            const Index& src0_slice_origin,
+                                                            const Src1Desc& src1_desc,
+                                                            const Index& src1_slice_origin,
+                                                            const Src2Desc& src2_desc,
+                                                            const Index& src2_slice_origin,
+                                                            const DstDesc& dst_desc,
+                                                            const Index& dst_slice_origin,
+                                                            const ElementwiseOperation& element_op)
+        : src0_coord_(make_tensor_coordinate(src0_desc, src0_slice_origin)),
+          src1_coord_(make_tensor_coordinate(src1_desc, src1_slice_origin)),
+          src2_coord_(make_tensor_coordinate(src2_desc, src2_slice_origin)),
+          dst_coord_(make_tensor_coordinate(dst_desc, dst_slice_origin)),
+          element_op_(element_op)
+    {
+        static_assert(SliceLengths::At(Number<VectorDim>{}) % ScalarPerVector == 0,
+                      "wrong! cannot evenly divide");
+    }
+
+    __device__ void SetSrc0SliceOrigin(const Src0Desc& src0_desc,
+                                       const Index& src0_slice_origin_idx)
+    {
+        src0_coord_ = make_tensor_coordinate(src0_desc, src0_slice_origin_idx);
+    }
+
+    __device__ void SetSrc1SliceOrigin(const Src1Desc& src1_desc,
+                                       const Index& src1_slice_origin_idx)
+    {
+        src1_coord_ = make_tensor_coordinate(src1_desc, src1_slice_origin_idx);
+    }
+
+    __device__ void SetSrc2SliceOrigin(const Src2Desc& src2_desc,
+                                       const Index& src2_slice_origin_idx)
+    {
+        src2_coord_ = make_tensor_coordinate(src2_desc, src2_slice_origin_idx);
+    }
+
+    __device__ void SetDstSliceOrigin(const DstDesc& dst_desc, const Index& dst_slice_origin_idx)
+    {
+        dst_coord_ = make_tensor_coordinate(dst_desc, dst_slice_origin_idx);
+    }
+
+    template <typename Src0Buffer, typename Src1Buffer, typename Src2Buffer, typename DstBuffer>
+    __device__ void Run(const Src0Desc& src0_desc,
+                        const Src0Buffer& src0_buf,
+                        const Src1Desc& src1_desc,
+                        const Src1Buffer& src1_buf,
+                        const Src2Desc& src2_desc,
+                        const Src2Buffer& src2_buf,
+                        const DstDesc& dst_desc,
+                        DstBuffer& dst_buf)
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        auto make_forward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index forward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        forward_step_idx(j) = (i.value == j.value) ? scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, forward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        auto make_backward_steps = [&](auto desc) {
+            return generate_tuple(
+                [&](auto i) {
+                    Index backward_step_idx;
+
+                    static_for<0, nDim, 1>{}([&](auto j) {
+                        backward_step_idx(j) = (i.value == j.value) ? -scalar_per_access[i] : 0;
+                    });
+
+                    return make_tensor_coordinate_step(desc, backward_step_idx);
+                },
+                Number<nDim>{});
+        };
+
+        // make forward steps
+        const auto src0_forward_steps = make_forward_steps(src0_desc);
+        const auto src1_forward_steps = make_forward_steps(src1_desc);
+        const auto src2_forward_steps = make_forward_steps(src2_desc);
+        const auto dst_forward_steps  = make_forward_steps(dst_desc);
+
+        // make backward steps
+        const auto src0_backward_steps = make_backward_steps(src0_desc);
+        const auto src1_backward_steps = make_backward_steps(src1_desc);
+        const auto src2_backward_steps = make_backward_steps(src2_desc);
+        const auto dst_backward_steps  = make_backward_steps(dst_desc);
+
+        // loop over slice window
+        static_ford<decltype(ordered_access_lengths)>{}([&](auto ordered_access_idx) {
+            // judge move forward or move backward
+            constexpr auto forward_sweep = [&]() {
+                StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+                forward_sweep_(I0) = true;
+
+                static_for<1, nDim, 1>{}([&](auto i) {
+                    index_t tmp = ordered_access_idx[I0];
+
+                    static_for<1, i, 1>{}([&](auto j) {
+                        tmp = tmp * ordered_access_lengths[j] + ordered_access_idx[j];
+                    });
+
+                    forward_sweep_(i) = tmp % 2 == 0;
+                });
+
+                return forward_sweep_;
+            }();
+
+            using src0_vector_type = vector_type_maker_t<Src0Data, ScalarPerVector>;
+            using src0_vector_t    = typename src0_vector_type::type;
+
+            using src1_vector_type = vector_type_maker_t<Src1Data, ScalarPerVector>;
+            using src1_vector_t    = typename src1_vector_type::type;
+
+            using src2_vector_type = vector_type_maker_t<Src2Data, ScalarPerVector>;
+            using src2_vector_t    = typename src2_vector_type::type;
+
+            using dst_vector_type = vector_type_maker_t<DstData, ScalarPerVector>;
+            using dst_vector_t    = typename dst_vector_type::type;
+
+            const bool is_src0_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src0_desc, src0_coord_);
+
+            const bool is_src1_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src1_desc, src1_coord_);
+
+            const bool is_src2_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(src2_desc, src2_coord_);
+
+            // copy data from src0_buf into src0_vector_container
+            auto src0_vector_container = src0_vector_type{
+                src0_buf.template Get<src0_vector_t>(src0_coord_.GetOffset(), is_src0_valid)};
+
+            auto src1_vector_container = src1_vector_type{
+                src1_buf.template Get<src1_vector_t>(src1_coord_.GetOffset(), is_src1_valid)};
+
+            auto src2_vector_container = src2_vector_type{
+                src2_buf.template Get<src2_vector_t>(src2_coord_.GetOffset(), is_src2_valid)};
+
+            auto dst_vector_container = dst_vector_type{};
+
+            // apply pointwise operation
+            static_for<0, ScalarPerVector, 1>{}([&](auto i) {
+                element_op_(dst_vector_container.template AsType<DstData>()(i),
+                            src0_vector_container.template AsType<Src0Data>()[i],
+                            src1_vector_container.template AsType<Src1Data>()[i],
+                            src2_vector_container.template AsType<Src2Data>()[i]);
+            });
+
+            const bool is_dst_valid =
+                coordinate_has_valid_offset_assuming_visible_index_is_valid(dst_desc, dst_coord_);
+
+            // copy data from dst_vector into dst_buf
+            if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::Set)
+            {
+                dst_buf.template Set<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+            else if constexpr(DstInMemOp == InMemoryDataOperationEnum_t::AtomicAdd)
+            {
+                dst_buf.template AtomicAdd<dst_vector_t>(
+                    dst_coord_.GetOffset(),
+                    is_dst_valid,
+                    dst_vector_container.template AsType<dst_vector_t>()[I0]);
+            }
+
+            constexpr auto move_on_dim = [&]() constexpr
+            {
+                StaticallyIndexedArray<bool, nDim> move_on_dim_;
+
+                static_for<0, nDim, 1>{}([&](auto i) {
+                    move_on_dim_(i) = ordered_access_idx[i] < ordered_access_lengths[i] - 1;
+
+                    static_for<i + 1, nDim, 1>{}([&](auto j) {
+                        move_on_dim_(i) &= ordered_access_idx[j] == ordered_access_lengths[j] - 1;
+                    });
+                });
+
+                return move_on_dim_;
+            }
+            ();
+
+            // move coordinate
+            static_for<0, nDim, 1>{}([&](auto i) {
+                if constexpr(move_on_dim[i])
+                {
+                    if constexpr(forward_sweep[i])
+                    {
+                        move_tensor_coordinate(
+                            src0_desc, src0_coord_, src0_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src1_desc, src1_coord_, src1_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src2_desc, src2_coord_, src2_forward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_forward_steps[dim_access_order[i]]);
+                    }
+                    else
+                    {
+                        move_tensor_coordinate(
+                            src0_desc, src0_coord_, src0_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src1_desc, src1_coord_, src1_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            src2_desc, src2_coord_, src2_backward_steps[dim_access_order[i]]);
+
+                        move_tensor_coordinate(
+                            dst_desc, dst_coord_, dst_backward_steps[dim_access_order[i]]);
+                    }
+                }
+            });
+        });
+
+        // move coordinate back to slice origin (or not)
+        if constexpr(Src0ResetCoordinateAfterRun)
+        {
+            const auto src0_reset_step =
+                make_tensor_coordinate_step(src0_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src0_desc, src0_coord_, src0_reset_step);
+        }
+
+        if constexpr(Src1ResetCoordinateAfterRun)
+        {
+            const auto src1_reset_step =
+                make_tensor_coordinate_step(src1_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src1_desc, src1_coord_, src1_reset_step);
+        }
+
+        if constexpr(Src2ResetCoordinateAfterRun)
+        {
+            const auto src2_reset_step =
+                make_tensor_coordinate_step(src2_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(src2_desc, src2_coord_, src2_reset_step);
+        }
+
+        if constexpr(DstResetCoordinateAfterRun)
+        {
+            const auto dst_reset_step =
+                make_tensor_coordinate_step(dst_desc, GetCoordinateResetStep());
+
+            move_tensor_coordinate(dst_desc, dst_coord_, dst_reset_step);
+        }
+    }
+
+    __device__ static constexpr auto GetCoordinateResetStep()
+    {
+        // scalar per access on each dim
+        // TODO: don't use lambda_scalar_per_access
+        constexpr auto scalar_per_access = generate_sequence(
+            detail::lambda_scalar_per_access<VectorDim, ScalarPerVector>{}, Number<nDim>{});
+
+        constexpr auto access_lengths = SliceLengths{} / scalar_per_access;
+
+        constexpr auto dim_access_order = DimAccessOrder{};
+
+        constexpr auto ordered_access_lengths =
+            container_reorder_given_new2old(access_lengths, dim_access_order);
+
+        // judge move forward or move backward during the last iteration
+        constexpr auto forward_sweep = [&]() {
+            StaticallyIndexedArray<bool, nDim> forward_sweep_;
+
+            forward_sweep_(I0) = true;
+
+            static_for<1, nDim, 1>{}([&](auto i) {
+                index_t tmp = ordered_access_lengths[I0] - 1;
+
+                static_for<1, i, 1>{}([&](auto j) {
+                    tmp = tmp * ordered_access_lengths[j] + ordered_access_lengths[j] - 1;
+                });
+
+                forward_sweep_(i) = tmp % 2 == 0;
+            });
+
+            return forward_sweep_;
+        }();
+
+        // calculate data index after last iteration in Run(), if it has not being reset
+        constexpr auto data_idx = [&]() {
+            Index ordered_idx;
+
+            static_for<0, nDim, 1>{}([&](auto i) {
+                ordered_idx(i) = forward_sweep[i] ? ordered_access_lengths[i] - 1 : 0;
+            });
+
+            return container_reorder_given_old2new(ordered_idx, dim_access_order) *
+                   scalar_per_access;
+        }();
+
+        //
+        constexpr auto reset_data_step = [&]() {
+            Index reset_data_step_;
+
+            static_for<0, nDim, 1>{}([&](auto i) { reset_data_step_(i) = -data_idx[i]; });
+
+            return reset_data_step_;
+        }();
+
+        return reset_data_step;
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrc0SliceWindow(const Src0Desc& src0_desc,
+                                        const Index& src0_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = Src0ResetCoordinateAfterRun
+                                           ? src0_slice_origin_step_idx
+                                           : src0_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src0_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src0_desc, src0_coord_, adjusted_step);
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrc1SliceWindow(const Src1Desc& src1_desc,
+                                        const Index& src1_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = Src1ResetCoordinateAfterRun
+                                           ? src1_slice_origin_step_idx
+                                           : src1_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src1_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src1_desc, src1_coord_, adjusted_step);
+    }
+
+    // src_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveSrc2SliceWindow(const Src2Desc& src2_desc,
+                                        const Index& src2_slice_origin_step_idx)
+    {
+        // if src coord was not reset by RunRead(), then need to adjust the step here
+        const auto adjusted_step_idx = Src2ResetCoordinateAfterRun
+                                           ? src2_slice_origin_step_idx
+                                           : src2_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(src2_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(src2_desc, src2_coord_, adjusted_step);
+    }
+
+    // dst_slice_origin_step_idx need to be known at compile-time, for performance reason
+    __device__ void MoveDstSliceWindow(const DstDesc& dst_desc,
+                                       const Index& dst_slice_origin_step_idx)
+    {
+        // if dst coord was not reset by Run(), then need to adjust the step here
+        const auto adjusted_step_idx = DstResetCoordinateAfterRun
+                                           ? dst_slice_origin_step_idx
+                                           : dst_slice_origin_step_idx + GetCoordinateResetStep();
+
+        // is it OK to construct a new step every time?
+        const auto adjusted_step = make_tensor_coordinate_step(dst_desc, adjusted_step_idx);
+
+        move_tensor_coordinate(dst_desc, dst_coord_, adjusted_step);
+    }
+
+    private:
+    Src0Coord src0_coord_;
+    Src1Coord src1_coord_;
+    Src2Coord src2_coord_;
+    DstCoord dst_coord_;
+    const ElementwiseOperation element_op_;
+};
+
+} // namespace ck
+#endif

composable_kernel/include/utility/amd_buffer_addressing.hpp

@@ -31,7 +31,7 @@ __device__ int32x4_t make_wave_buffer_resource(T* p_wave, index_t element_space_
     return wave_buffer_resource.content;
 }
 
-// load
+// buffer load i8
 __device__ int8_t
 llvm_amdgcn_raw_buffer_load_i8(int32x4_t srsrc,
                                index_t voffset,
@@ -50,6 +50,7 @@ llvm_amdgcn_raw_buffer_load_i8x4(int32x4_t srsrc,
                                  index_t soffset,
                                  index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.load.v4i8");
 
+// buffer load i16
 __device__ ushort
 llvm_amdgcn_raw_buffer_load_i16(int32x4_t srsrc,
                                 index_t voffset,
@@ -68,6 +69,7 @@ llvm_amdgcn_raw_buffer_load_i16x4(int32x4_t srsrc,
                                   index_t soffset,
                                   index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.load.v4i16");
 
+// buffer load i32
 __device__ int32_t
 llvm_amdgcn_raw_buffer_load_i32(int32x4_t srsrc,
                                 index_t voffset,
@@ -85,7 +87,7 @@ llvm_amdgcn_raw_buffer_load_i32x4(int32x4_t srsrc,
                                   index_t voffset,
                                   index_t soffset,
                                   index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.load.v4i32");
-// half
+// buffer load fp16
 __device__ half_t
 llvm_amdgcn_raw_buffer_load_fp16(int32x4_t srsrc,
                                  index_t voffset,
@@ -104,7 +106,7 @@ llvm_amdgcn_raw_buffer_load_fp16x4(int32x4_t srsrc,
                                    index_t soffset,
                                    index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.load.v4f16");
 
-// float
+// buffer load fp32
 __device__ float
 llvm_amdgcn_raw_buffer_load_fp32(int32x4_t srsrc,
                                  index_t voffset,
@@ -123,7 +125,7 @@ llvm_amdgcn_raw_buffer_load_fp32x4(int32x4_t srsrc,
                                    index_t soffset,
                                    index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.load.v4f32");
 
-// store
+// buffer store i8
 __device__ void
 llvm_amdgcn_raw_buffer_store_i8(int8_t vdata,
                                 int32x4_t rsrc,
@@ -145,6 +147,7 @@ llvm_amdgcn_raw_buffer_store_i8x4(int8x4_t vdata,
                                   index_t soffset,
                                   index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.store.v4i8");
 
+// buffer store i16
 __device__ void
 llvm_amdgcn_raw_buffer_store_i16(ushort vdata,
                                  int32x4_t rsrc,
@@ -166,6 +169,7 @@ llvm_amdgcn_raw_buffer_store_i16x4(ushort4_t vdata,
                                    index_t soffset,
                                    index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.store.v4i16");
 
+// buffer store i32
 __device__ void
 llvm_amdgcn_raw_buffer_store_i32(int32_t vdata,
                                  int32x4_t rsrc,
@@ -187,7 +191,7 @@ llvm_amdgcn_raw_buffer_store_i32x4(int32x4_t vdata,
                                    index_t soffset,
                                    index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.store.v4i32");
 
-// half
+// buffer store fp16
 __device__ void
 llvm_amdgcn_raw_buffer_store_fp16(half_t vdata,
                                   int32x4_t rsrc,
@@ -208,7 +212,7 @@ llvm_amdgcn_raw_buffer_store_fp16x4(half4_t vdata,
                                     index_t voffset,
                                     index_t soffset,
                                     index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.store.v4f16");
-// float
+// buffer store fp32
 __device__ void
 llvm_amdgcn_raw_buffer_store_fp32(float vdata,
                                   int32x4_t rsrc,
@@ -229,8 +233,15 @@ llvm_amdgcn_raw_buffer_store_fp32x4(float4_t vdata,
                                     index_t voffset,
                                     index_t soffset,
                                     index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.store.v4f32");
-// atomic add
-// int
+// buffer atomic-add fp16
+__device__ half2_t llvm_amdgcn_raw_buffer_atomic_add_fp16x2(
+    half2_t vdata,
+    int32x4_t rsrc,
+    index_t voffset,
+    index_t soffset,
+    index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.atomic.fadd.v2f16");
+
+// buffer atomic-add i32
 __device__ int32_t llvm_amdgcn_raw_buffer_atomic_add_i32(
     int32_t vdata,
     int32x4_t rsrc,
@@ -238,7 +249,7 @@ __device__ int32_t llvm_amdgcn_raw_buffer_atomic_add_i32(
     index_t soffset,
     index_t glc_slc) __asm("llvm.amdgcn.raw.buffer.atomic.add.i32");
 
-// float
+// buffer atomic-add fp32
 __device__ float llvm_amdgcn_raw_buffer_atomic_add_fp32(
     float vdata,
     int32x4_t rsrc,
@@ -752,6 +763,7 @@ __device__ void amd_buffer_atomic_add_impl(const typename vector_type<T, N>::typ
                                            index_t dst_wave_addr_offset)
 {
     static_assert((is_same<T, float>::value && (N == 1 || N == 2 || N == 4)) ||
+                      (is_same<T, half_t>::value && (N == 2 || N == 4 || N == 8)) ||
                       (is_same<T, int32_t>::value && (N == 1 || N == 2 || N == 4)),
                   "wrong! not implemented");
 
@@ -810,6 +822,41 @@ __device__ void amd_buffer_atomic_add_impl(const typename vector_type<T, N>::typ
                                                    0);
         }
     }
+    else if constexpr(is_same<T, half_t>::value)
+    {
+        if constexpr(N == 2)
+        {
+            llvm_amdgcn_raw_buffer_atomic_add_fp16x2(src_thread_data,
+                                                     dst_wave_buffer_resource,
+                                                     dst_thread_addr_offset,
+                                                     dst_wave_addr_offset,
+                                                     0);
+        }
+        else if constexpr(N == 4)
+        {
+            vector_type<half_t, 4> tmp{src_thread_data};
+
+            static_for<0, 2, 1>{}([&](auto i) {
+                llvm_amdgcn_raw_buffer_atomic_add_fp16x2(tmp.AsType<half2_t>()[i],
+                                                         dst_wave_buffer_resource,
+                                                         dst_thread_addr_offset,
+                                                         dst_wave_addr_offset + i * sizeof(half2_t),
+                                                         0);
+            });
+        }
+        else if constexpr(N == 8)
+        {
+            vector_type<half_t, 8> tmp{src_thread_data};
+
+            static_for<0, 4, 1>{}([&](auto i) {
+                llvm_amdgcn_raw_buffer_atomic_add_fp16x2(tmp.AsType<half2_t>()[i],
+                                                         dst_wave_buffer_resource,
+                                                         dst_thread_addr_offset,
+                                                         dst_wave_addr_offset + i * sizeof(half2_t),
+                                                         0);
+            });
+        }
+    }
     else if constexpr(is_same<T, int32_t>::value)
     {
         if constexpr(N == 1)

composable_kernel/include/utility/common_header.hpp

@@ -35,8 +35,8 @@
 #include "dynamic_buffer.hpp"
 #include "is_known_at_compile_time.hpp"
 #include "transpose_vectors.hpp"
-
 #include "inner_product.hpp"
+#include "element_wise_operation.hpp"
 
 // TODO: remove this
 #if CK_USE_AMD_INLINE_ASM

composable_kernel/include/utility/config.hpp

@@ -24,12 +24,16 @@
 #define CK_MIN_BLOCK_PER_CU 2
 #endif
 
-// buffer resourse
+// GPU-specific parameters
 #if defined(CK_AMD_GPU_GFX803) || defined(CK_AMD_GPU_GFX900) || defined(CK_AMD_GPU_GFX906) || \
     defined(CK_AMD_GPU_GFX908) || defined(CK_AMD_GPU_GFX90A)
+// buffer resourse
 #define CK_BUFFER_RESOURCE_3RD_DWORD 0x00020000
+// wave size
+#define CK_GPU_WAVE_SIZE 64
 #elif defined(CK_AMD_GPU_GFX1030)
 #define CK_BUFFER_RESOURCE_3RD_DWORD 0x31014000
+#define CK_GPU_WAVE_SIZE 32
 #endif
 
 // FMA instruction
@@ -141,6 +145,10 @@
 #define CK_WORKAROUND_SWDEV_XXXXXX_THREAD_WISE_COPY_V1R4_TYPE_CONVERT_ISSUE 1
 #endif
 
+#ifndef CK_WORKAROUND_SWDEV_XXXXXX_THREAD_WISE_COPY_V1R5_TYPE_CONVERT_ISSUE
+#define CK_WORKAROUND_SWDEV_XXXXXX_THREAD_WISE_COPY_V1R5_TYPE_CONVERT_ISSUE 1
+#endif
+
 namespace ck {
 
 enum InMemoryDataOperationEnum_t
@@ -152,7 +160,7 @@ enum InMemoryDataOperationEnum_t
 
 enum ActivTypeEnum_t
 {
-    None = 0,
+    None,
     LeakyRelu,
     Sigmoid
 };

composable_kernel/include/utility/utility.hpp

@@ -5,8 +5,12 @@
 
 namespace ck {
 
+__device__ constexpr index_t get_wave_size() { return CK_GPU_WAVE_SIZE; }
+
 __device__ index_t get_thread_local_1d_id() { return threadIdx.x; }
 
+__device__ index_t get_wave_local_1d_id() { return threadIdx.x / get_wave_size(); }
+
 __device__ index_t get_block_1d_id() { return blockIdx.x; }
 
 } // namespace ck

device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_c_shuffle_bias_activation_add_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_bias_activation_add_instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using AddReluAdd  = ck::tensor_operation::element_wise::AddReluAdd;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+static constexpr auto ConvFwd1x1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+
+static constexpr auto ConvFwd1x1S1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+
+static constexpr auto ConvFwdOddC =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::OddC;
+
+// arbitrary conv
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances = std::tuple<
+    // clang-format off
+        //##############################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##############################################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_1x1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##############################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##############################################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, stride 1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##############################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##############################################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// Odd C
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_odd_c_f16_instances = std::tuple<
+    // clang-format off
+        //##############################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##############################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##############################################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##############################################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,    64,    64,    64,     4,  8,   32,   32,    2,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Add_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,  AddReluAdd,    ConvFwdOddC,    64,    32,    64,     4,  8,   32,   32,    1,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances(
+    std::vector<DeviceConvFwdBiasActivationAddPtr<PassThrough, PassThrough, AddReluAdd>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances{});
+    add_device_operation_instances(
+        instances,
+        device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_1x1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances,
+        device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances,
+        device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_odd_c_f16_instances{});
+}
+
+} // namespace device_conv2d_fwd_bias_activation_add_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_bias_activation_atomic_add_instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using AddRelu     = ck::tensor_operation::element_wise::AddRelu;
+
+static constexpr auto InMemoryAtomicAdd = ck::InMemoryDataOperationEnum_t::AtomicAdd;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|               Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise|      GlobalMemory| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################################|       |        |        |        |   Operation|   Operation|   Operation|     DataOperation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################################|       |        |        |        |            |            |            |                  |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 4, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 16>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 4, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 4, 1, 1, 16>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 8, 1, 1, 16>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,   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,           1,           1,             S<1, 1, 4, 1, 1, 32>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 4, 1, 1, 16>,               2>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu, InMemoryAtomicAdd, ConvFwdDefault,    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,           1,           1,             S<1, 1, 4, 1, 1, 16>,               2>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances(
+    std::vector<DeviceConvFwdBiasActivationPtr<PassThrough, PassThrough, AddRelu>>&
+        instance_container)
+{
+    using Instances =
+        device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances;
+
+    const auto instances = Instances{};
+
+    ck::static_for<0, std::tuple_size_v<Instances>, 1>{}([&](auto i) {
+        using Instance = remove_cvref_t<decltype(std::get<i>(instances))>;
+
+        auto instance = Instance{};
+
+        instance_container.push_back(std::make_unique<Instance>(instance));
+    });
+}
+
+} // namespace device_conv2d_fwd_bias_activation_atomic_add_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_c_shuffle_bias_activation_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_bias_activation_instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+using AddRelu     = ck::tensor_operation::element_wise::AddRelu;
+
+static constexpr auto MemorySet = ck::InMemoryDataOperationEnum_t::Set;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+static constexpr auto ConvFwd1x1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+
+static constexpr auto ConvFwd1x1S1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+
+static constexpr auto ConvFwdOddC =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::OddC;
+
+// arbitrary conv
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|           Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise|  GlobalMemory| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################################|       |        |        |        |   Operation|   Operation|   Operation| DataOperation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################################|       |        |        |        |            |            |            |              |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_1x1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|           Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise|  GlobalMemory| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################################|       |        |        |        |   Operation|   Operation|   Operation| DataOperation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################################|       |        |        |        |            |            |            |              |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, stride 1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|           Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise|  GlobalMemory| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################################|       |        |        |        |   Operation|   Operation|   Operation| DataOperation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################################|       |        |        |        |            |            |            |              |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// Odd C
+using device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_odd_c_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|           Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise|  GlobalMemory| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################################|       |        |        |        |   Operation|   Operation|   Operation| DataOperation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################################|       |        |        |        |            |            |            |              |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,    64,    64,    64,     4,  8,   32,   32,    2,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Bias_Activation_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough,     AddRelu,     MemorySet,    ConvFwdOddC,    64,    32,    64,     4,  8,   32,   32,    1,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances(
+    std::vector<DeviceConvFwdBiasActivationPtr<PassThrough, PassThrough, AddRelu>>& instances)
+{
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances{});
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_1x1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances,
+        device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_odd_c_f16_instances{});
+}
+
+} // namespace device_conv2d_fwd_bias_activation_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+static constexpr auto ConvFwd1x1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+
+static constexpr auto ConvFwd1x1S1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+
+static constexpr auto ConvFwdOddC =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::OddC;
+
+// arbitrary conv
+using device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_1x1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+// 1x1, stride 1, pad 0
+using device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+using device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_odd_c_f16_instances = std::tuple<
+    // clang-format off
+        //##########################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle|     CBlockTransferClusterLengths|  CBlockTransfer|
+        //##########################################################################|   Type|    Type|    Type|    Type| Elementwise| Elementwise| Elementwise| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MXdlPerWave_MWaveMPerXdl| ScalarPerVector|
+        //##########################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle| _NBlock_NXdlPerWave_NWaveNPerXdl|   _NWaveNPerXdl|
+        //##########################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                                 |                |
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   256,   256,   128,     4,  8,   32,   32,    4,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   256,   128,   256,     4,  8,   32,   32,    2,    4,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   128,   128,   128,     4,  8,   32,   32,    4,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   256,   128,   128,     4,  8,   32,   32,    2,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   128,   128,    64,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   128,    64,   128,     4,  8,   32,   32,    2,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,    64,    64,    64,     4,  8,   32,   32,    2,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   256,   128,    64,     4,  8,   32,   32,    2,    1,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   256,    64,   128,     4,  8,   32,   32,    1,    2,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  8, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   128,   128,    32,     4,  8,   32,   32,    2,    1,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 32, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,   128,    32,   128,     4,  8,   32,   32,    1,    2,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  4, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 8>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,    64,    64,    32,     4,  8,   32,   32,    2,    1,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>,
+        DeviceConv2dFwdXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,    ConvFwdOddC,    64,    32,    64,     4,  8,   32,   32,    1,    2,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              1,              1,      true,     S<4,  2, 8>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              1,              1,      true,           1,           1,             S<1, 1, 16, 1, 1, 4>,               8>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances(
+    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances)
+{
+    add_device_operation_instances(instances,
+                                   device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instances{});
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_1x1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances{});
+    add_device_operation_instances(
+        instances, device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_odd_c_f16_instances{});
+}
+
+} // namespace device_conv2d_fwd_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_instance {
+
+using F16 = ck::half_t;
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+static constexpr auto ConvFwd1x1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+
+static constexpr auto ConvFwd1x1S1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+
+// Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    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>
+    // clang-format on
+    >;
+
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    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>
+    // clang-format on
+    >;
+
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F16,     F16,     F16,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    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>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances(
+    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances)
+{
+    add_device_operation_instances(instances, device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances{});
+    add_device_operation_instances(instances,
+                                   device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_p0_f16_instances{});
+    add_device_operation_instances(instances,
+                                   device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_s1_p0_f16_instances{});
+}
+
+} // namespace device_conv2d_fwd_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp

+#include <stdlib.h>
+#include "config.hpp"
+#include "device_conv2d_fwd_xdl_nhwc_kyxc_nhwk.hpp"
+#include "element_wise_operation.hpp"
+#include "device_operation_instance.hpp"
+
+namespace ck {
+namespace tensor_operation {
+namespace device {
+namespace device_conv2d_fwd_instance {
+
+using F32 = float;
+
+template <ck::index_t... Is>
+using S = ck::Sequence<Is...>;
+
+using PassThrough = ck::tensor_operation::element_wise::PassThrough;
+
+static constexpr auto ConvFwdDefault =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+
+static constexpr auto ConvFwd1x1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+
+static constexpr auto ConvFwd1x1S1P0 =
+    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+
+// Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   256,   128,     4,  4,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   256,     4,  4,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,   128,     4,  4,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,   128,     4,  4,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,    64,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,    64,   128,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    64,    64,    64,     4,  4,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,   128,    64,     4,  4,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   256,    64,   128,     4,  4,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,   128,    32,     4,  4,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,   128,    32,   128,     4,  4,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    64,    64,    32,     4,  4,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwdDefault,    64,    32,    64,     4,  4,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>
+    // clang-format on
+    >;
+
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_p0_f32_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   256,   256,   128,     4,  4,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   256,   128,   256,     4,  4,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   128,   128,   128,     4,  4,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   256,   128,   128,     4,  4,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   128,   128,    64,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   128,    64,   128,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    64,    64,    64,     4,  4,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   256,   128,    64,     4,  4,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   256,    64,   128,     4,  4,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   128,   128,    32,     4,  4,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,   128,    32,   128,     4,  4,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    64,    64,    32,     4,  4,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough,   ConvFwd1x1P0,    64,    32,    64,     4,  4,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>
+    // clang-format on
+    >;
+
+using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_s1_p0_f32_instances = std::tuple<
+    // clang-format off
+        //################################################################| InData| WeiData| OutData| AccData|          In|         Wei|         Out|    ConvForward| 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| Specialization|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| SrcDstVectorDim|       DstScalar|
+        //################################################################|       |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |                |       PerVector|
+        //################################################################|       |        |        |        |            |            |            |               |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   256,   256,   128,     4,  4,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   256,   128,   256,     4,  4,   32,   32,    2,    4,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   128,   128,   128,     4,  4,   32,   32,    4,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   256,   128,   128,     4,  4,   32,   32,    2,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   128,   128,    64,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   128,    64,   128,     4,  4,   32,   32,    2,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    64,    64,    64,     4,  4,   32,   32,    2,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   256,   128,    64,     4,  4,   32,   32,    2,    1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   256,    64,   128,     4,  4,   32,   32,    1,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   128,   128,    32,     4,  4,   32,   32,    2,    1,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,   128,    32,   128,     4,  4,   32,   32,    1,    2,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    64,    64,    32,     4,  4,   32,   32,    2,    1,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>,
+        DeviceConv2dFwdXdl_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_N_Ho_Wo_K<    F32,     F32,     F32,     F32, PassThrough, PassThrough, PassThrough, ConvFwd1x1S1P0,    64,    32,    64,     4,  4,   32,   32,    1,    2,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      true,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,      true,               7,               1>
+    // clang-format on
+    >;
+
+void add_device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances(
+    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& instances)
+{
+    add_device_operation_instances(instances, device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances{});
+    add_device_operation_instances(instances,
+                                   device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_p0_f32_instances{});
+    add_device_operation_instances(instances,
+                                   device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_1x1_s1_p0_f32_instances{});
+}
+
+} // namespace device_conv2d_fwd_instance
+} // namespace device
+} // namespace tensor_operation
+} // namespace ck

device_operation/device_conv_xdl_instance_f16_f16_f16_nhwc_kyxc_nhwk.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_conv_fwd_xdl_nhwc_kyxc_nhwk.hpp"
-#include "device_conv_instance.hpp"
-#include "element_wise_operation.hpp"
-
-namespace ck {
-namespace tensor_operation {
-namespace device {
-namespace device_conv_instance {
-
-using F16 = ck::half_t;
-using F32 = float;
-
-using NHWC = ck::tensor_layout::convolution::NHWC;
-using KYXC = ck::tensor_layout::convolution::KYXC;
-using NHWK = ck::tensor_layout::convolution::NHWK;
-
-template <ck::index_t... Is>
-using S = ck::Sequence<Is...>;
-
-using PassThrough = ck::tensor_operation::element_wise::PassThrough;
-
-// Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
-using device_conv_fwd_xdl_instances_f16_f16_f16_nhwc_kyxc_nhwk = std::tuple<
-    // clang-format off
-        //##############|    NDim| InData| WeiData| OutData| AccData|     In|    Wei|    Out|           A|           B|           C| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer|  BBlockTransfer|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| CThreadTransfer| CThreadTransfer| ABlockLds| BBlockLds|
-        //##############| Spatial|   Type|    Type|    Type|    Type| Layout| Layout| Layout| Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| SrcDstVectorDim|       DstScalar| AddExtraM| AddExtraN|
-        //##############|        |       |        |        |        |       |       |       |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_N_K1| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1| Lengths_K0_N_K1| Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|                |       PerVector|          |          |
-        //##############|        |       |        |        |        |       |       |       |            |            |            |      |      |      |      |   |     |     |     |     |                |                |               |               |               |               |               |                |                |               |               |              |               |               |                |                |          |          |
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 2, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  8,   32,   32,    2,    2,      S<1, 4, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 1, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,      S<1, 1, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  8,   32,   32,    2,    1,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 1, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  8,   32,   32,    1,    2,      S<1, 1, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    64,    32,     4,  8,   32,   32,    2,    1,      S<1, 4, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 2, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F16,     F16,     F16,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  8,   32,   32,    1,    2,      S<1, 2, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,      S<1, 4, 8>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>
-    // clang-format on
-    >;
-
-template <>
-void add_device_conv_fwd_instance<2, F16, F16, F16, NHWC, KYXC, NHWK>(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& device_conv_instances)
-{
-    using DeviceConvs = device_conv_fwd_xdl_instances_f16_f16_f16_nhwc_kyxc_nhwk;
-
-    const auto device_convs = DeviceConvs{};
-
-    ck::static_for<0, std::tuple_size_v<DeviceConvs>, 1>{}([&](auto i) {
-        using Conv = remove_cvref_t<decltype(std::get<i>(device_convs))>;
-
-        auto conv = Conv{};
-
-        device_conv_instances.push_back(std::make_unique<Conv>(conv));
-    });
-}
-
-} // namespace device_conv_instance
-} // namespace device
-} // namespace tensor_operation
-} // namespace ck

device_operation/device_conv_xdl_instance_f32_f32_f32_nhwc_kyxc_nhwk.cpp

-#include <stdlib.h>
-#include "config.hpp"
-#include "device_conv_fwd_xdl_nhwc_kyxc_nhwk.hpp"
-#include "device_conv_instance.hpp"
-#include "element_wise_operation.hpp"
-
-namespace ck {
-namespace tensor_operation {
-namespace device {
-namespace device_conv_instance {
-
-using F16 = ck::half_t;
-using F32 = float;
-
-using NHWC = ck::tensor_layout::convolution::NHWC;
-using KYXC = ck::tensor_layout::convolution::KYXC;
-using NHWK = ck::tensor_layout::convolution::NHWK;
-
-template <ck::index_t... Is>
-using S = ck::Sequence<Is...>;
-
-using PassThrough = ck::tensor_operation::element_wise::PassThrough;
-
-// Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
-using device_conv_fwd_xdl_instances_f32_f32_f32_nhwc_kyxc_nhwk = std::tuple<
-    // clang-format off
-        //##############|    NDim| InData| WeiData| OutData| AccData|     In|    Wei|    Out|           A|           B|           C| Block|  MPer|  NPer| K0Per| K1| MPer| NPer| MXdl| NXdl|  ABlockTransfer|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer|  BBlockTransfer|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| CThreadTransfer| CThreadTransfer| ABlockLds| BBlockLds|
-        //##############| Spatial|   Type|    Type|    Type|    Type| Layout| Layout| Layout| Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| SrcDstVectorDim|       DstScalar| AddExtraM| AddExtraN|
-        //##############|        |       |        |        |        |       |       |       |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_N_K1| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1| Lengths_K0_N_K1| Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|                |       PerVector|          |          |
-        //##############|        |       |        |        |        |       |       |       |            |            |            |      |      |      |      |   |     |     |     |     |                |                |               |               |               |               |               |                |                |               |               |              |               |               |                |                |          |          |
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  4,   32,   32,    4,    2,      S<1, 4, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  4,   32,   32,    2,    4,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  4,   32,   32,    4,    2,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  4,   32,   32,    2,    2,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  4,   32,   32,    2,    2,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 2, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  4,   32,   32,    2,    2,      S<1, 2, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    64,    64,     4,  4,   32,   32,    2,    2,      S<1, 4, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  4,   32,   32,    2,    1,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 1, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  4,   32,   32,    1,    2,      S<1, 1, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 2, 4>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,   128,    32,     4,  4,   32,   32,    2,    1,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 1, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,   128,    32,   128,     4,  4,   32,   32,    1,    2,      S<1, 1, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    64,    32,     4,  4,   32,   32,    2,    1,      S<1, 4, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 2, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>,
-        DeviceConvFwdXdl<       2,    F32,     F32,     F32,     F32,   NHWC,   KYXC,   NHWK, PassThrough, PassThrough, PassThrough,    64,    32,    64,     4,  4,   32,   32,    1,    2,      S<1, 2, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              4,              4,      S<1, 4, 4>,     S<4, 16, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              4,              4,               7,               1,      true,      true>
-    // clang-format on
-    >;
-
-template <>
-void add_device_conv_fwd_instance<2, F32, F32, F32, NHWC, KYXC, NHWK>(
-    std::vector<DeviceConvFwdPtr<PassThrough, PassThrough, PassThrough>>& device_conv_instances)
-{
-    using DeviceConvs = device_conv_fwd_xdl_instances_f32_f32_f32_nhwc_kyxc_nhwk;
-
-    const auto device_convs = DeviceConvs{};
-
-    ck::static_for<0, std::tuple_size_v<DeviceConvs>, 1>{}([&](auto i) {
-        using Conv = remove_cvref_t<decltype(std::get<i>(device_convs))>;
-
-        auto conv = Conv{};
-
-        device_conv_instances.push_back(std::make_unique<Conv>(conv));
-    });
-}
-
-} // namespace device_conv_instance
-} // namespace device
-} // namespace tensor_operation
-} // namespace ck

device_operation/device_gemm_xdl_instance_f16_f16_f16_km_kn_mn.cpp

@@ -21,22 +21,23 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 // Compilation parameters for a[k, m] * b[k, n] = c[m, n]
-using device_gemm_xdl_instance_f16_f16_f16_km_kn_mn = 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| ABlockTransfer|  BBlockTransfer|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| CThreadTransfer| CThreadTransfer| ABlockLds| BBlockLds|
-        //##########|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| SrcDstVectorDim|       DstScalar| AddExtraM| AddExtraN|
-        //##########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_N_K1| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1| Lengths_K0_N_K1| Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|                |       PerVector|          |          |
-        //##########|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |                |               |               |               |               |               |                |                |               |               |              |               |               |                |                |          |          |
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 4, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 2, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              4,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 1, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              1,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Row,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,      S<1, 1, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,             1,              2,              8,               7,               1,      true,      true>
-    // clang-format on
-    >;
+using device_gemm_xdl_instance_f16_f16_f16_km_kn_mn =
+    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|
+        //##########|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>
+        // clang-format on
+        >;
 
 template <>
 void add_device_gemm_instance<F16, F16, F16, Col, Row, Row>(

device_operation/device_gemm_xdl_instance_f16_f16_f16_km_nk_mn.cpp

@@ -21,22 +21,23 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 // Compilation parameters for a[k, m] * b[n, k] = c[m, n]
-using device_gemm_xdl_instance_f16_f16_f16_km_nk_mn = 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| ABlockTransfer|  BBlockTransfer|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| CThreadTransfer| CThreadTransfer| ABlockLds| BBlockLds|
-        //##########|  Type|  Type|  Type|    Type|        |        |        | Elementwise| Elementwise| Elementwise|  Size| Block| Block| Block|   |  XDL|  XDL|  Per|  Per|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar|     ThreadSlice|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| SrcDstVectorDim|       DstScalar| AddExtraM| AddExtraN|
-        //##########|      |      |      |        |        |        |        |   Operation|   Operation|   Operation|      |      |      |      |   |     |     | Wave| Wave| Lengths_K0_N_K1| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1| Lengths_K0_N_K1| Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|                |       PerVector|          |          |
-        //##########|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |                |               |               |               |               |               |                |                |               |               |              |               |               |                |                |          |          |
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   256,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   256,     4,  8,   32,   32,    2,    4,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 4, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,   128,     4,  8,   32,   32,    4,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,   128,    64,     4,  8,   32,   32,    2,    2,      S<1, 4, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              4,              8,      S<1, 2, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   128,    64,   128,     4,  8,   32,   32,    2,    2,      S<1, 2, 8>,     S<4, 32, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 4, 8>,     S<4, 32, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,   128,    64,     4,  8,   32,   32,    2,    1,      S<1, 2, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              2,              8,      S<1, 1, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>,
-        DeviceGemmXdl<  F16,   F16,   F16,     F32,     Col,      Col,    Row, PassThrough, PassThrough, PassThrough,   256,    64,   128,     4,  8,   32,   32,    1,    2,      S<1, 1, 8>,     S<4, 64, 1>,     S<0, 2, 1>,     S<0, 2, 1>,              1,              1,              8,      S<1, 2, 8>,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,               7,               1,      true,      true>
-    // clang-format on
-    >;
+using device_gemm_xdl_instance_f16_f16_f16_km_nk_mn =
+    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|
+        //##########|      |      |      |        |        |        |        |            |            |            |      |      |      |      |   |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |                |                |
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>,
+        DeviceGemmXdl<  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>
+        // clang-format on
+        >;
 
 template <>
 void add_device_gemm_instance<F16, F16, F16, Col, Col, Row>(