topk_softmax (#1592)

* topk_softmax

* remove some file

* fix atomix linear_offset

* address various comment, and change sfc get_index api to static(tuple)

include/ck_tile/core/utility/magic_div.hpp

@@ -59,8 +59,16 @@ struct magic_division32_bit_range
     CK_TILE_DEVICE static constexpr uint32_t
     do_magic_division(uint32_t dividend, uint32_t multiplier, uint32_t shift)
     {
-        uint32_t tmp = __umulhi(dividend, multiplier);
-        return (tmp + dividend) >> shift;
+        if(__builtin_is_constant_evaluated())
+        {
+            uint32_t tmp = (static_cast<uint64_t>(dividend) * multiplier) >> 32;
+            return (tmp + dividend) >> shift;
+        }
+        else
+        {
+            uint32_t tmp = __umulhi(dividend, multiplier);
+            return (tmp + dividend) >> shift;
+        }
     }
 
     CK_TILE_HOST static constexpr uint32_t
@@ -77,9 +85,18 @@ struct magic_division32_bit_range
     CK_TILE_DEVICE static constexpr int32_t
     do_magic_division(int32_t dividend_i32, uint32_t multiplier, uint32_t shift)
     {
-        uint32_t dividend_u32 = bit_cast<uint32_t>(dividend_i32);
-        uint32_t tmp          = __umulhi(dividend_u32, multiplier);
-        return (tmp + dividend_u32) >> shift;
+        if(__builtin_is_constant_evaluated())
+        {
+            uint32_t dividend_u32 = bit_cast<uint32_t>(dividend_i32);
+            uint32_t tmp          = (static_cast<uint64_t>(dividend_u32) * multiplier) >> 32;
+            return (tmp + dividend_u32) >> shift;
+        }
+        else
+        {
+            uint32_t dividend_u32 = bit_cast<uint32_t>(dividend_i32);
+            uint32_t tmp          = __umulhi(dividend_u32, multiplier);
+            return (tmp + dividend_u32) >> shift;
+        }
     }
 
     CK_TILE_HOST static constexpr int32_t

include/ck_tile/host.hpp

@@ -24,5 +24,6 @@
 #include "ck_tile/host/reference/reference_layernorm2d_fwd.hpp"
 #include "ck_tile/host/reference/reference_reduce.hpp"
 #include "ck_tile/host/reference/reference_softmax.hpp"
+#include "ck_tile/host/reference/reference_topk.hpp"
 #include "ck_tile/host/stream_config.hpp"
 #include "ck_tile/host/timer.hpp"

include/ck_tile/host/fill.hpp

@@ -10,6 +10,7 @@
 #include <random>
 #include <type_traits>
 #include <utility>
+#include <unordered_set>
 
 #include "ck_tile/core.hpp"
 
@@ -41,6 +42,73 @@ struct FillUniformDistribution
     }
 };
 
+namespace impl {
+
+// clang-format off
+template<index_t bytes> struct RawIntegerType_ {};
+template<> struct RawIntegerType_<1> { using type = uint8_t;};
+template<> struct RawIntegerType_<2> { using type = uint16_t;};
+template<> struct RawIntegerType_<4> { using type = uint32_t;};
+template<> struct RawIntegerType_<8> { using type = uint64_t;};
+// clang-format on
+
+template <typename T>
+using RawIntegerType = typename RawIntegerType_<sizeof(T)>::type;
+} // namespace impl
+
+// Note: this struct will have no const-ness will generate random
+template <typename T>
+struct FillUniformDistribution_Unique
+{
+    float a_{-5.f};
+    float b_{5.f};
+    std::optional<uint32_t> seed_{11939};
+
+    std::mt19937 gen_{};
+    std::unordered_set<impl::RawIntegerType<T>> set_{};
+
+    FillUniformDistribution_Unique(float a                      = -5.f,
+                                   float b                      = 5.f,
+                                   std::optional<uint32_t> seed = {11939})
+        : a_(a),
+          b_(b),
+          seed_(seed),
+          gen_{seed_.has_value() ? *seed_ : std::random_device{}()},
+          set_{}
+    {
+    }
+
+    template <typename ForwardIter>
+    void operator()(ForwardIter first, ForwardIter last)
+    {
+        std::mt19937& gen = gen_;
+        std::uniform_real_distribution<float> dis(a_, b_);
+        auto& set = set_;
+        std::generate(first, last, [&dis, &gen, &set]() {
+            T v = static_cast<T>(0);
+            do
+            {
+                v = ck_tile::type_convert<T>(dis(gen));
+            } while(set.count(bit_cast<impl::RawIntegerType<T>>(v)) == 1);
+            set.insert(bit_cast<impl::RawIntegerType<T>>(v));
+
+            return v;
+        });
+    }
+
+    template <typename ForwardRange>
+    auto operator()(ForwardRange&& range)
+        -> std::void_t<decltype(std::declval<FillUniformDistribution_Unique&>()(
+            std::begin(std::forward<ForwardRange>(range)),
+            std::end(std::forward<ForwardRange>(range))))>
+    {
+        (*this)(std::begin(std::forward<ForwardRange>(range)),
+                std::end(std::forward<ForwardRange>(range)));
+    }
+
+    void clear() { set_.clear(); }
+};
+
 template <typename T>
 struct FillNormalDistribution
 {

include/ck_tile/host/host_tensor.hpp

@@ -11,6 +11,7 @@
 #include <thread>
 #include <utility>
 #include <vector>
+#include <functional>
 
 #include "ck_tile/core.hpp"
 #include "ck_tile/host/ranges.hpp"
@@ -545,6 +546,28 @@ struct HostTensor
 
     typename Data::size_type size() const { return mData.size(); }
 
+    // return a slice of this tensor
+    // for simplicity we just copy the data and return a new tensor
+    auto slice(std::vector<size_t> s_begin, std::vector<size_t> s_end) const
+    {
+        assert(s_begin.size() == s_end.size());
+        assert(s_begin.size() == get_num_of_dimension());
+
+        std::vector<size_t> s_len(s_begin.size());
+        std::transform(
+            s_end.begin(), s_end.end(), s_begin.begin(), s_len.begin(), std::minus<size_t>{});
+        HostTensor<T> sliced_tensor(s_len);
+
+        sliced_tensor.ForEach([&](auto& self, auto idx) {
+            std::vector<size_t> src_idx(idx.size());
+            std::transform(
+                idx.begin(), idx.end(), s_begin.begin(), src_idx.begin(), std::plus<size_t>{});
+            self(idx) = operator()(src_idx);
+        });
+
+        return sliced_tensor;
+    }
+
     template <typename U = T>
     auto AsSpan() const
     {

include/ck_tile/host/reference/reference_softmax.hpp

 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 
@@ -9,43 +9,81 @@
 
 namespace ck_tile {
 
-template <typename ADataType, typename AccDataType, typename BDataType>
-CK_TILE_HOST void reference_softmax(const HostTensor<ADataType>& a_m_n,
-                                    HostTensor<BDataType>& b_m_n)
+template <typename InputType, typename ComputeType, typename OutputType = ComputeType>
+CK_TILE_HOST void
+reference_softmax(const HostTensor<InputType>& x, HostTensor<OutputType>& y, index_t dim = -1)
 {
-    auto f = [&](auto m) {
-        const int N = a_m_n.mDesc.get_lengths()[1];
+    index_t rank = x.get_num_of_dimension();
+    assert(rank == y.get_num_of_dimension());
+    assert(dim == -1 || dim < rank);
 
-        AccDataType v_max = ck_tile::numeric<ADataType>::Lowest();
+    index_t target_dim  = dim == -1 ? (rank - 1) : dim;
+    index_t softmax_len = x.get_length(target_dim);
+    index_t n_parallel  = x.get_element_size() / softmax_len;
+    auto x_len          = x.get_lengths();
 
-        // max
-        for(int n = 0; n < N; ++n)
-        {
-            const ADataType v_a = a_m_n(m, n);
+    auto f = [&](auto i_element) {
+        std::vector<size_t> coord = [&]() {
+            std::vector<size_t> t_(rank, 0);
+            size_t r = i_element;
+            for(index_t i = rank - 1; i >= 0; i--)
+            {
+                if(i == target_dim)
+                    continue;
+                t_[i] = r % x_len[i];
+                r     = r / x_len[i];
+            }
+            return t_;
+        }();
+
+        ComputeType v_max = -ck_tile::numeric<ComputeType>::infinity();
 
-            v_max = v_max < v_a ? v_a : v_max;
+        // compute max
+        for(auto idx = 0; idx < softmax_len; idx++)
+        {
+            auto c_               = coord;
+            c_[target_dim]        = idx;
+            const ComputeType v_x = ck_tile::type_convert<ComputeType>(x(c_));
+            v_max                 = v_max < v_x ? v_x : v_max;
         }
 
-        AccDataType v_exp_sum = 0;
+        ComputeType v_exp_sum = static_cast<ComputeType>(0);
 
         // sum
-        for(int n = 0; n < N; ++n)
+        for(auto idx = 0; idx < softmax_len; idx++)
         {
-            const ADataType v_a = a_m_n(m, n);
+            auto c_        = coord;
+            c_[target_dim] = idx;
 
-            v_exp_sum += ck_tile::exp(v_a - v_max);
+            const ComputeType v_x = ck_tile::type_convert<ComputeType>(x(c_));
+
+            v_exp_sum += ck_tile::exp(v_x - v_max);
         }
 
         // elementwise
-        for(int n = 0; n < N; ++n)
+        for(auto idx = 0; idx < softmax_len; idx++)
         {
-            const ADataType v_a = a_m_n(m, n);
+            auto c_        = coord;
+            c_[target_dim] = idx;
+
+            const ComputeType v_x = ck_tile::type_convert<ComputeType>(x(c_));
+
+            auto out = ck_tile::exp(v_x - v_max) / v_exp_sum;
 
-            b_m_n(m, n) = ck_tile::exp(v_a - v_max) / v_exp_sum;
+            y(c_) = ck_tile::type_convert<OutputType>(out);
         }
     };
 
-    make_ParallelTensorFunctor(f,
-                               b_m_n.mDesc.get_lengths()[0])(std::thread::hardware_concurrency());
+    make_ParallelTensorFunctor(f, n_parallel)(std::thread::hardware_concurrency());
+}
+
+template <typename InputType, typename ComputeType, typename OutputType = ComputeType>
+CK_TILE_HOST auto reference_softmax(const HostTensor<InputType>& x, index_t dim = -1)
+{
+    HostTensor<OutputType> y(x.get_lengths(), x.get_strides());
+
+    reference_softmax<InputType, ComputeType, OutputType>(x, y, dim);
+
+    return y;
 }
 } // namespace ck_tile

include/ck_tile/host/reference/reference_topk.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include "ck_tile/host/host_tensor.hpp"
+#include <thread>
+#include <numeric>
+#include <functional>
+#include <utility>
+#include <algorithm>
+
+namespace ck_tile {
+
+/*
+    similiar to torch.topk()
+    x (Tensor) – the input tensor.
+    k (int) – the k in “top-k”
+    dim (int, optional) – the dimension to sort along
+    largest (bool, optional) – largest or smallest elements
+    sorted (bool, optional) – elements in sorted order or not
+
+    output:
+    y_values
+    y_indices
+
+    https://github.com/pytorch/pytorch/blob/main/aten/src/ATen/native/TopKImpl.h
+*/
+template <typename DataType, typename IndexType = index_t>
+CK_TILE_HOST void reference_topk(const HostTensor<DataType>& x,
+                                 HostTensor<DataType>& y_values,
+                                 HostTensor<IndexType>& y_indices,
+                                 index_t k,
+                                 index_t dim  = -1,
+                                 bool largest = true,
+                                 bool sorted  = true)
+{
+    // rank must be the same
+    index_t rank = x.get_num_of_dimension();
+    assert(rank == y_values.get_num_of_dimension());
+    assert(rank == y_indices.get_num_of_dimension());
+    assert(dim == -1 || dim < rank);
+
+    index_t topk_dim     = dim == -1 ? (rank - 1) : dim;
+    index_t topk_src_len = x.get_length(topk_dim);
+    auto x_len           = x.get_lengths();
+
+    assert(k <= topk_src_len);
+    assert(k == y_values.get_length(topk_dim) && k == y_indices.get_length(topk_dim));
+
+    index_t n_parallel = x.get_element_size() / topk_src_len;
+
+    // clang-format off
+    auto f = [&](auto i_element) {
+        std::vector<size_t> topk_coord = [&](){
+            std::vector<size_t> t_(rank, 0);
+            size_t r = i_element;
+            for(index_t i = rank - 1; i >= 0; i--) {
+                if(i == topk_dim)          continue; // topk dim should be zero
+                t_[i] = r % x_len[i];      r = r / x_len[i];
+            }
+            return t_;
+        }();
+
+        using elem_t = std::pair<DataType, IndexType>;
+        std::vector<elem_t> q = [&](){
+            std::vector<elem_t> t_(topk_src_len);
+            for(index_t i = 0; i < topk_src_len; i++) {
+                auto c_ = topk_coord;  c_[topk_dim] = i;
+                t_[i].first = x(c_);   t_[i].second = i;
+            }
+            return t_;
+        }();
+
+        // run topk
+        if(largest) {
+            std::nth_element(q.begin(), q.begin() + k - 1, q.end(),
+            [](const elem_t& lhs, const elem_t& rhs) -> bool { return lhs.first > rhs.first; });
+            if(sorted) {
+                std::sort(q.begin(), q.begin() + k - 1,
+                [](const elem_t& lhs, const elem_t& rhs) -> bool { return lhs.first > rhs.first; });
+            }
+        } else {
+            std::nth_element(q.begin(), q.begin() + k - 1, q.end(),
+            [](const elem_t& lhs, const elem_t& rhs) -> bool { return lhs.first < rhs.first; });
+            if(sorted) {
+                std::sort(q.begin(), q.begin() + k - 1,
+                [](const elem_t& lhs, const elem_t& rhs) -> bool { return lhs.first < rhs.first; });
+            }
+        }
+
+        // write out
+        for(index_t i = 0; i < k; i++) {
+            auto c_ = topk_coord;  c_[topk_dim] = i;
+            y_values(c_) = q[i].first;  y_indices(c_) = q[i].second;
+        }
+    };
+    // clang-format on
+
+    make_ParallelTensorFunctor(f, n_parallel)(std::thread::hardware_concurrency());
+}
+
+// TODO: if using this method, the return tensor would be dense(no stride)
+template <typename DataType, typename IndexType = index_t>
+CK_TILE_HOST auto reference_topk(const HostTensor<DataType>& x,
+                                 index_t k,
+                                 index_t dim  = -1,
+                                 bool largest = true,
+                                 bool sorted  = true)
+{
+    auto lens          = x.get_lengths();
+    index_t target_dim = (dim == -1) ? (lens.size() - 1) : dim;
+    assert(target_dim < lens.size());
+    assert(k <= lens[target_dim]);
+    lens[target_dim] = k;
+    HostTensor<DataType> y_values(lens);
+    HostTensor<IndexType> y_indices(lens);
+
+    reference_topk<DataType, IndexType>(x, y_values, y_indices, k, dim, largest, sorted);
+
+    return ck_tile::make_tuple(y_values, y_indices);
+}
+} // namespace ck_tile

include/ck_tile/ops/elementwise.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/ops/elementwise/unary_element_wise_operation.hpp"
+#include "ck_tile/ops/common/tensor_layout.hpp"

include/ck_tile/ops/elementwise/unary_element_wise_operation.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include <type_traits>
+
+namespace ck_tile {
+namespace element_wise {
+
+#if 0
+struct PassThroughPack2
+{
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const;
+
+    CK_TILE_HOST_DEVICE constexpr void operator()(ck_tile::half2_t& y, const ck_tile::f8x2_t& x) const
+    {
+        auto t = type_convert<float2_t>(x);
+        y      = type_convert<half2_t>(t);
+    }
+    constexpr const static bool is_pack2_invocable = true;
+};
+#endif
+
+struct PassThrough
+{
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<double, double>(double& y, const double& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, double>(float& y, const double& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<double, float>(double& y, const float& x) const
+    {
+        y = type_convert<double>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, float>(float& y, const float& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp16_t, ck_tile::fp16_t>(ck_tile::fp16_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp16_t, float>(ck_tile::fp16_t& y,
+                                                                const float& x) const
+    {
+        y = type_convert<ck_tile::fp16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::bf16_t, ck_tile::bf16_t>(ck_tile::bf16_t& y, const ck_tile::bf16_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int32_t, int32_t>(int32_t& y, const int32_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::bf16_t, float>(ck_tile::bf16_t& y,
+                                                                const float& x) const
+    {
+        y = type_convert<ck_tile::bf16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, ck_tile::bf16_t>(float& y,
+                                                                const ck_tile::bf16_t& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::bf16_t, ck_tile::fp16_t>(ck_tile::bf16_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = type_convert<ck_tile::bf16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, ck_tile::fp16_t>(float& y,
+                                                                const ck_tile::fp16_t& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int8_t, int8_t>(int8_t& y, const int8_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp16_t, int8_t>(ck_tile::fp16_t& y,
+                                                                 const int8_t& x) const
+    {
+        y = type_convert<ck_tile::fp16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::bf16_t, int8_t>(ck_tile::bf16_t& y,
+                                                                 const int8_t& x) const
+    {
+        y = type_convert<ck_tile::bf16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<uint8_t, uint8_t>(uint8_t& y, const uint8_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int8_t, int32_t>(int8_t& y, const int32_t& x) const
+    {
+        y = type_convert<int8_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int32_t, int8_t>(int32_t& y, const int8_t& x) const
+    {
+        y = type_convert<int32_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int8_t, float>(int8_t& y, const float& x) const
+    {
+        y = type_convert<int8_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, int8_t>(float& y, const int8_t& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+#ifdef CK_TILE_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int4_t, int4_t>(int4_t& y, const int4_t& x) const
+    {
+        y = x;
+    }
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int4_t, int>(int4_t& y, const int& x) const
+    {
+        y = type_convert<int4_t>(x);
+    }
+#endif
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp8_t, ck_tile::fp8_t>(ck_tile::fp8_t& y, const ck_tile::fp8_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, ck_tile::fp8_t>(float& y,
+                                                               const ck_tile::fp8_t& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp8_t, float>(ck_tile::fp8_t& y,
+                                                               const float& x) const
+    {
+        y = type_convert<ck_tile::fp8_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp16_t, ck_tile::fp8_t>(ck_tile::fp16_t& y, const ck_tile::fp8_t& x) const
+    {
+        y = type_convert<ck_tile::fp16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp8_t, ck_tile::fp16_t>(ck_tile::fp8_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = type_convert<ck_tile::fp8_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::bf8_t, ck_tile::bf8_t>(ck_tile::bf8_t& y, const ck_tile::bf8_t& x) const
+    {
+        y = x;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, ck_tile::bf8_t>(float& y,
+                                                               const ck_tile::bf8_t& x) const
+    {
+        y = type_convert<float>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::bf8_t, float>(ck_tile::bf8_t& y,
+                                                               const float& x) const
+    {
+        y = type_convert<ck_tile::bf8_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp16_t, ck_tile::bf8_t>(ck_tile::fp16_t& y, const ck_tile::bf8_t& x) const
+    {
+        y = type_convert<ck_tile::fp16_t>(x);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::bf8_t, ck_tile::fp16_t>(ck_tile::bf8_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = ck_tile::type_convert<ck_tile::bf8_t>(x);
+    }
+};
+
+#if 0
+struct UnaryConvert
+{
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        y = type_convert<Y>(x);
+    }
+};
+
+struct ConvertBF16RTN
+{
+    // convert to bf16 using round to nearest (rtn)
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        // check Y datatype
+        static_assert(std::is_same_v<Y, ck_tile::bf16_t>, "Data type is not supported by this operation!");
+
+        // check X datatype
+        static_assert(std::is_same_v<X, float> || std::is_same_v<X, ck_tile::fp16_t>,
+                      "Data type is not supported by this operation!");
+
+        y = bf16_convert_rtn<Y>(x);
+    }
+};
+
+struct ConvertF8SR
+{
+    // convert to fp8 using stochastic rounding (SR)
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        // check Y datatype
+        static_assert(std::is_same_v<Y, ck_tile::fp8_t> || std::is_same_v<Y, ck_tile::bf8_t>,
+                      "Data type is not supported by this operation!");
+
+        // check X datatype
+        static_assert(std::is_same_v<X, float> || std::is_same_v<X, ck_tile::fp16_t>,
+                      "Data type is not supported by this operation!");
+
+        y = f8_convert_sr<Y>(x);
+    }
+};
+
+struct ConvertF8RNE
+{
+    // convert to fp8 using rounding to nearest even
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        // check Y datatype
+        static_assert(std::is_same_v<Y, ck_tile::fp8_t> || std::is_same_v<Y, ck_tile::bf8_t>,
+                      "Data type is not supported by this operation!");
+
+        // check X datatype
+        static_assert(std::is_same_v<X, float> || std::is_same_v<X, ck_tile::fp16_t>,
+                      "Data type is not supported by this operation!");
+
+        y = f8_convert_rne<Y>(x);
+    }
+};
+#endif
+
+struct Scale
+{
+    CK_TILE_HOST_DEVICE Scale(float scale = 1.f) : scale_(scale) {}
+
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        y = ck_tile::type_convert<Y>(ck_tile::type_convert<float>(x) * scale_);
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp16_t, ck_tile::fp16_t>(ck_tile::fp16_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = ck_tile::type_convert<ck_tile::fp16_t>(scale_) * x;
+    };
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::bf16_t, ck_tile::bf16_t>(ck_tile::bf16_t& y, const ck_tile::bf16_t& x) const
+    {
+        const float x_tmp = ck_tile::type_convert<float>(x);
+        const float y_tmp = scale_ * x_tmp;
+        y                 = ck_tile::type_convert<ck_tile::bf16_t>(y_tmp);
+    };
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, float>(float& y, const float& x) const
+    {
+        y = scale_ * x;
+    };
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<double, double>(double& y, const double& x) const
+    {
+        y = scale_ * x;
+    };
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<int8_t, int8_t>(int8_t& y, const int8_t& x) const
+    {
+        y = ck_tile::type_convert<int8_t>(scale_ * ck_tile::type_convert<float>(x));
+    };
+
+    float scale_;
+};
+
+struct ScaleAndResetNaNToMinusInfinity
+{
+    CK_TILE_HOST_DEVICE ScaleAndResetNaNToMinusInfinity(float scale) : scale_(scale) {}
+
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, float>(float& y, const float& x) const
+    {
+        y = ck_tile::isnan(x) ? -numeric<float>::infinity() : scale_ * x;
+    };
+
+    float scale_;
+};
+
+struct UnaryDivide
+{
+    CK_TILE_HOST_DEVICE UnaryDivide(const int32_t divider = 1) : divider_(divider) {}
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = x / type_convert<T>(divider_);
+    };
+
+    int32_t divider_ = 1;
+};
+
+struct UnarySquare
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, ck_tile::fp16_t> ||
+                          std::is_same_v<T, double> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>
+#ifdef CK_TILE_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
+                          || std::is_same_v<T, int4_t>
+#endif
+                      ,
+                      "Data type is not supported by this operation!");
+        y = x * x;
+    };
+};
+
+struct UnaryAbs
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::abs(x);
+    };
+};
+
+struct UnarySqrt
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::sqrt(x);
+    };
+};
+
+struct Relu
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        y = x > 0 ? x : 0;
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()(ck_tile::bf16_t& y, const ck_tile::bf16_t& x) const
+    {
+        float x_f32 = ck_tile::type_convert<float>(x);
+        float y_f32 = x_f32 > 0 ? x_f32 : 0;
+        y           = ck_tile::type_convert<ck_tile::bf16_t>(y_f32);
+    }
+};
+
+// Fast GeLU
+// https://paperswithcode.com/method/gelu
+// y = 0.5*x*(1+tanh(sqrt(2/pi)*(x+0.044715*x^3)))
+// host code use higher accuracy "exp" and "div"
+// gpu code use lower accuracy "_ocml_exp_f32" and "rcp" function
+struct FastGelu
+{
+    template <typename Y, typename X>
+    CK_TILE_HOST void operator()(Y& y, const X& x) const;
+
+    template <typename Y, typename X>
+    CK_TILE_DEVICE void operator()(Y& y, const X& x) const;
+
+    template <>
+    CK_TILE_HOST void operator()<float, float>(float& y, const float& x) const
+    {
+        // const float u   = -2.f * x * (0.035677f * x * x + 0.797885f);
+        const float c1  = -2.0 * 0.035677f;
+        const float c2  = -2.0 * 0.797885f;
+        const float u   = x * (c1 * x * x + c2);
+        const float emu = exp(u);
+        y               = x / (1.f + emu);
+    }
+
+    // device code, use lower precision "__ocml_exp_f32" and "rcp"
+    template <>
+    CK_TILE_DEVICE void operator()<float, float>(float& y, const float& x) const
+    {
+        // const float u   = 2.f * x * (0.035677f * x * x + 0.797885f);
+        const float c1  = -2.0 * 0.035677f;
+        const float c2  = -2.0 * 0.797885f;
+        const float u   = x * (c1 * x * x + c2);
+        const float emu = __ocml_exp_f32(u);
+
+        y = x * ck_tile::rcp(1.f + emu);
+    }
+
+    template <>
+    CK_TILE_HOST void operator()<ck_tile::fp16_t, ck_tile::fp16_t>(ck_tile::fp16_t& y,
+                                                                   const ck_tile::fp16_t& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, type_convert<float>(x));
+
+        y = type_convert<ck_tile::fp16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_DEVICE void operator()<ck_tile::fp16_t, ck_tile::fp16_t>(ck_tile::fp16_t& y,
+                                                                     const ck_tile::fp16_t& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, type_convert<float>(x));
+
+        y = type_convert<ck_tile::fp16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_HOST void operator()<ck_tile::fp16_t, float>(ck_tile::fp16_t& y, const float& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, x);
+
+        y = type_convert<ck_tile::fp16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_DEVICE void operator()<ck_tile::fp16_t, float>(ck_tile::fp16_t& y, const float& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, x);
+
+        y = type_convert<ck_tile::fp16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_HOST void operator()<ck_tile::bf16_t, float>(ck_tile::bf16_t& y, const float& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, x);
+
+        y = type_convert<ck_tile::bf16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_DEVICE void operator()<ck_tile::bf16_t, float>(ck_tile::bf16_t& y, const float& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, x);
+
+        y = type_convert<ck_tile::bf16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_DEVICE void operator()<ck_tile::bf16_t, ck_tile::bf16_t>(ck_tile::bf16_t& y,
+                                                                     const ck_tile::bf16_t& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, type_convert<float>(x));
+
+        y = type_convert<ck_tile::bf16_t>(y_f);
+    }
+
+    template <>
+    CK_TILE_HOST void operator()<ck_tile::bf16_t, ck_tile::bf16_t>(ck_tile::bf16_t& y,
+                                                                   const ck_tile::bf16_t& x) const
+    {
+        float y_f;
+
+        this->operator()<float, float>(y_f, type_convert<float>(x));
+
+        y = type_convert<ck_tile::bf16_t>(y_f);
+    }
+};
+
+// https://paperswithcode.com/method/gelu
+// y = 0.5*x*(1+erf(x/sqrt(2)))
+struct Gelu
+{
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<float, float>(float& y, const float& x) const
+    {
+        y = 0.5f * x * (1.f + erf(float(0.70710678118f * x)));
+    }
+
+    template <>
+    CK_TILE_HOST_DEVICE void
+    operator()<ck_tile::fp16_t, ck_tile::fp16_t>(ck_tile::fp16_t& y, const ck_tile::fp16_t& x) const
+    {
+        y = ck_tile::fp16_t(0.5) * x *
+            (ck_tile::fp16_t(1) + ck_tile::fp16_t(erf(float(0.70710678118f * x))));
+    }
+};
+
+struct Sigmoid
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+        constexpr T one = type_convert<T>(1);
+        y               = one / (one + ck_tile::exp(-x));
+    };
+};
+
+struct Silu
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+        constexpr T one = type_convert<T>(1);
+        y               = x * (one / (one + ck_tile::exp(-x)));
+    };
+};
+
+struct TanH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::tanh(x);
+    };
+};
+
+struct ACos
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::acos(x);
+    };
+};
+
+struct Neg
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::neg(x);
+    };
+};
+
+struct ATan
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::atan(x);
+    };
+};
+
+struct Sin
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::sin(x);
+    };
+};
+
+struct ASinH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::asinh(x);
+    };
+};
+
+struct Cos
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::cos(x);
+    };
+};
+
+struct ACosH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::acosh(x);
+    };
+};
+
+struct Tan
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::tan(x);
+    };
+};
+
+struct ATanH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::atanh(x);
+    };
+};
+
+struct SinH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::sinh(x);
+    };
+};
+
+struct Ceil
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::ceil(x);
+    };
+};
+
+struct Exp
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::exp(x);
+    };
+};
+
+struct CosH
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::cosh(x);
+    };
+};
+
+struct Floor
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::floor(x);
+    };
+};
+
+struct Log
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::log(x);
+    };
+};
+
+struct ASin
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::asin(x);
+    };
+};
+
+struct Rcp
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int8_t> ||
+                          std::is_same_v<T, int32_t>,
+                      "Data type is not supported by this operation!");
+
+        y = ck_tile::rcp(x);
+    };
+};
+
+struct Swish
+{
+    Swish(float beta = 1.0f) : beta_(beta) {}
+
+    template <typename Y, typename X>
+    CK_TILE_HOST_DEVICE void operator()(Y& y, const X& x) const
+    {
+        static_assert(std::is_same_v<X, float> || std::is_same_v<X, double> ||
+                          std::is_same_v<X, ck_tile::fp16_t>,
+                      "Data type is not supported by this operation!");
+
+        static_assert(std::is_same_v<Y, float> || std::is_same_v<Y, double> ||
+                          std::is_same_v<Y, ck_tile::fp16_t>,
+                      "Data type is not supported by this operation!");
+
+        float bx = -beta_ * type_convert<float>(x);
+        y        = type_convert<Y>(x / (1.f + ck_tile::exp(bx)));
+    };
+
+    const float beta_;
+};
+
+struct SoftRelu
+{
+    SoftRelu(float alpha = 1.f) : alpha_(alpha){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha  = type_convert<T>(alpha_);
+        constexpr T one = type_convert<T>(1);
+        y               = ck_tile::log(one + ck_tile::exp(x * casted_alpha)) / casted_alpha;
+    }
+    const float alpha_;
+};
+
+struct Power
+{
+    Power(float alpha = 0.f, float beta = 1.f, float gamma = 2.f)
+        : alpha_(alpha), beta_(beta), gamma_(gamma){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha     = type_convert<T>(alpha_);
+        T casted_beta      = type_convert<T>(beta_);
+        T casted_gamma     = type_convert<T>(gamma_);
+        T shifted_scaled_x = casted_alpha + casted_beta * x;
+        y                  = ck_tile::pow(shifted_scaled_x, casted_gamma);
+    }
+    const float alpha_;
+    const float beta_;
+    const float gamma_;
+};
+
+struct ClippedRelu
+{
+    ClippedRelu(float alpha = 0.f, float beta = 1.f) : alpha_(alpha), beta_(beta){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha = type_convert<T>(alpha_);
+        T casted_beta  = type_convert<T>(beta_);
+        y              = ck_tile::min(casted_beta, ck_tile::max(casted_alpha, x));
+    }
+    const float alpha_;
+    const float beta_;
+};
+
+struct LeakyRelu
+{
+    LeakyRelu(float alpha = 0.01f) : alpha_(alpha){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha = type_convert<T>(alpha_);
+        y              = x >= 0 ? x : x * casted_alpha;
+    }
+    const float alpha_;
+};
+
+struct Elu
+{
+    Elu(float alpha = 1.f) : alpha_(alpha){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha = type_convert<T>(alpha_);
+        y              = x > 0 ? x : casted_alpha * ck_tile::expm1(x);
+    }
+    const float alpha_;
+};
+
+struct Logistic
+{
+    Logistic(float alpha = 1.f) : alpha_(alpha){};
+
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(T& y, const T& x) const
+    {
+        static_assert(std::is_same_v<T, float> || std::is_same_v<T, double> ||
+                          std::is_same_v<T, ck_tile::fp16_t> || std::is_same_v<T, int32_t> ||
+                          std::is_same_v<T, int8_t>,
+                      "Data type is not supported by this operation!");
+        T casted_alpha  = type_convert<T>(alpha_);
+        constexpr T one = type_convert<T>(1);
+        y               = casted_alpha / (one + ck_tile::exp(-x) * casted_alpha);
+    }
+    const float alpha_;
+};
+
+struct ConvInvscale
+{
+    CK_TILE_HOST_DEVICE
+    ConvInvscale(float scale_in = 1.f, float scale_wei = 1.f, float scale_out = 1.f)
+        : scale_in_(scale_in), scale_wei_(scale_wei), scale_out_(scale_out)
+    {
+    }
+
+    template <typename E, typename C>
+    CK_TILE_HOST_DEVICE void operator()(E& e, const C& c) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp8_t, float>(ck_tile::fp8_t& e,
+                                                               const float& c) const
+    {
+        e = type_convert<ck_tile::fp8_t>(c / scale_in_ / scale_wei_ / scale_out_);
+    };
+
+    float scale_in_;
+    float scale_wei_;
+    float scale_out_;
+};
+
+struct ConvScale
+{
+    CK_TILE_HOST_DEVICE
+    ConvScale(float scale_in = 1.f, float scale_wei = 1.f, float scale_out = 1.f)
+        : scale_in_(scale_in), scale_wei_(scale_wei), scale_out_(scale_out)
+    {
+    }
+
+    template <typename E, typename C>
+    CK_TILE_HOST_DEVICE void operator()(E& e, const C& c) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp8_t, float>(ck_tile::fp8_t& e,
+                                                               const float& c) const
+    {
+        e = type_convert<ck_tile::fp8_t>(c * scale_in_ * scale_wei_ * scale_out_);
+    };
+
+    float scale_in_;
+    float scale_wei_;
+    float scale_out_;
+};
+
+struct ConvScaleRelu
+{
+    CK_TILE_HOST_DEVICE
+    ConvScaleRelu(float scale_in = 1.f, float scale_wei = 1.f, float scale_out = 1.f)
+        : scale_in_(scale_in), scale_wei_(scale_wei), scale_out_(scale_out)
+    {
+    }
+
+    template <typename E, typename C>
+    CK_TILE_HOST_DEVICE void operator()(E& e, const C& c) const;
+
+    template <>
+    CK_TILE_HOST_DEVICE void operator()<ck_tile::fp8_t, float>(ck_tile::fp8_t& e,
+                                                               const float& c) const
+    {
+        float x;
+        Relu{}.template operator()<float>(x, c * scale_in_ * scale_wei_);
+        e = type_convert<ck_tile::fp8_t>(x * scale_out_);
+    };
+
+    float scale_in_;
+    float scale_wei_;
+    float scale_out_;
+};
+
+template <typename DstType, typename SrcType>
+struct Cast
+{
+    template <typename T>
+    CK_TILE_HOST_DEVICE void operator()(DstType& y, const SrcType& x) const
+    {
+        y = ck_tile::type_convert<DstType>(x);
+    };
+};
+
+// support fastconvert of int8 to fp16
+#if 0
+template <typename InputDataType, typename OutputDataType, index_t RegPackNumber>
+struct FastNumericArrayConverter
+{
+};
+
+template <>
+struct FastNumericArrayConverter<uint8_t, ck_tile::fp16_t, 4>
+{
+    using InputArray  = vector_type<uint8_t, 4>;
+    using OutputArray = vector_type<ck_tile::fp16_t, 4>;
+
+    CK_TILE_DEVICE static OutputArray convert(InputArray const& Input)
+    {
+        OutputArray Output;
+
+        uint32_t* half_2       = reinterpret_cast<uint32_t*>(&Output);
+        uint32_t const uint8_4 = reinterpret_cast<uint32_t const&>(Input);
+
+        static constexpr uint32_t byte_selector_01 = 0x05010500;
+        static constexpr uint32_t byte_selector_23 = 0x05030502;
+        static constexpr uint32_t fp16_adder       = 0x64646464;
+        half_2[0] = __builtin_amdgcn_perm(fp16_adder, uint8_4, byte_selector_01);
+        half_2[1] = __builtin_amdgcn_perm(fp16_adder, uint8_4, byte_selector_23);
+
+        static constexpr uint32_t I8s_TO_F16s_MAGIC_NUM = 0x64806480;
+        asm volatile("v_pk_add_f16 %0, %1, %2 neg_lo:[0,1] neg_hi:[0,1]"
+                     : "=v"(half_2[0])
+                     : "v"(half_2[0]), "s"(I8s_TO_F16s_MAGIC_NUM));
+        asm volatile("v_pk_add_f16 %0, %1, %2 neg_lo:[0,1] neg_hi:[0,1]"
+                     : "=v"(half_2[1])
+                     : "v"(half_2[1]), "s"(I8s_TO_F16s_MAGIC_NUM));
+
+        return Output;
+    }
+
+    CK_TILE_DEVICE OutputArray operator()(InputArray const& Input) { return convert(Input); }
+};
+
+template <index_t N>
+struct FastNumericArrayConverter<uint8_t, ck_tile::fp16_t, N>
+{
+    static constexpr int VEC_WIDTH = 4;
+    static_assert(!(N % VEC_WIDTH), "N must be multiple of 4.");
+
+    using InputArray  = vector_type<uint8_t, N>;
+    using OutputArray = vector_type<ck_tile::fp16_t, N>;
+
+    CK_TILE_DEVICE static OutputArray convert(InputArray const& Input)
+    {
+        FastNumericArrayConverter<uint8_t, ck_tile::fp16_t, 4> converter;
+
+        OutputArray Output;
+
+        using Vec_InputArray  = vector_type<uint8_t, 4>;
+        using Vec_OutputArray = vector_type<ck_tile::fp16_t, 4>;
+
+        Vec_OutputArray* half_4_ptr       = reinterpret_cast<Vec_OutputArray*>(&Output);
+        Vec_InputArray const* uint8_4_ptr = reinterpret_cast<Vec_InputArray const*>(&Input);
+
+        static_for<0, N / VEC_WIDTH, 1>{}(
+            [&](auto i) { half_4_ptr[i] = converter(uint8_4_ptr[i]); });
+
+        return Output;
+    }
+
+    CK_TILE_DEVICE OutputArray operator()(InputArray const& Input) { return convert(Input); }
+};
+#endif
+} // namespace element_wise
+} // namespace ck_tile

include/ck_tile/ops/fmha/pipeline/block_fmha_pipeline_qr_ks_vs_async.hpp

@@ -334,7 +334,7 @@ struct BlockFmhaPipelineQRKSVSAsync
         move_tile_window(k_dram_window, {0, kK0});
         __builtin_amdgcn_sched_barrier(0);
 
-        buffer_load_fence(k_dram_window.get_num_access(), q.get_thread_buffer());
+        buffer_load_fence(k_dram_window.get_num_of_access(), q.get_thread_buffer());
         (void)q_element_func; // ??? rocm-6.x if use q element func will have scratch on hdim=64/32
         // auto q_tile = q;      // tile_elementwise_in(q_element_func, q);
 
@@ -359,7 +359,7 @@ struct BlockFmhaPipelineQRKSVSAsync
                     if constexpr(i_k0 < k0_loops - 1)
                         move_tile_window(k_dram_window, {0, kK0});
 
-                    async_load_fence(k_dram_window.get_num_access());
+                    async_load_fence(k_dram_window.get_num_of_access());
                     __builtin_amdgcn_s_barrier();
                     __builtin_amdgcn_sched_barrier(0);
                     gemm_0(s_acc,

include/ck_tile/ops/reduce/block/block_reduce.hpp

@@ -4,9 +4,14 @@
 #pragma once
 
 #include "ck_tile/core.hpp"
+#include <tuple>
 
 namespace ck_tile {
 
+/*
+ * TODO: block_tile_reduce_sync() currently has a limitation
+ * Y dim must have at least one dim not been reduced
+ */
 // synchronize reduce result (cross lane reduction and broadcast on replicated dimension)
 template <typename AccDistributedTensor_, typename ReduceFunc, bool WithBroadcast = true>
 CK_TILE_DEVICE void block_tile_reduce_sync(AccDistributedTensor_& acc_tensor,
@@ -104,6 +109,65 @@ CK_TILE_DEVICE void block_tile_reduce_sync(AccDistributedTensor_& acc_tensor,
     });
 }
 
+/*
+ * this version is faster, using xor to do reduce, no need broadcast anymore
+ * TODO: the limitation is to-be-reduced P dim can only mapping to one R dim?
+ */
+template <typename AccDistributedTensor_, typename ReduceFunc>
+CK_TILE_DEVICE void block_tile_reduce_xor_sync(AccDistributedTensor_& acc_tensor,
+                                               const ReduceFunc& reduce_func)
+{
+    using Dstr             = typename AccDistributedTensor_::StaticTileDistribution;
+    using DstrEncode       = typename Dstr::DstrEncode;
+    using DstrEncodeDetail = typename DstrEncode::detail;
+
+    constexpr index_t NDimP = Dstr::get_num_of_dimension_p();
+    constexpr index_t NDimR = Dstr::get_num_of_dimension_r();
+
+    constexpr index_t idim_p_lane = NDimP - 1;
+
+    constexpr index_t thread_buf_size = AccDistributedTensor_::get_thread_buffer_size();
+
+    // loop over thread data
+    static_for<0, thread_buf_size, 1>{}([&](auto i) {
+        auto v_local = acc_tensor.get_thread_buffer()[i];
+
+        // cross-lane reduce for replication
+        // only reduce on R dimension correspond to lane
+        // (lane id maps to this R dimension)
+        static_for<0, NDimR, 1>{}([&](auto idim_r) {
+            // FIXME: nasty to use does_p_own_r_
+            if constexpr(DstrEncodeDetail::does_p_own_r_[idim_p_lane][idim_r])
+            {
+                constexpr index_t r_length = DstrEncode::rs_lengths_[idim_r];
+
+                constexpr index_t lid_over_rid_derivative =
+                    DstrEncodeDetail::ps_over_rs_derivative_[idim_p_lane][idim_r];
+
+                static_assert(is_power_of_two_integer(r_length),
+                              "wrong! only support power of 2 reduction");
+
+                constexpr index_t nstage = integer_log2_floor(r_length);
+
+                // reduction sweep forward
+                static_for<0, nstage, 1>{}([&](auto istage) {
+                    // xor
+                    index_t src_lane =
+                        __lane_id() ^ (number<lid_over_rid_derivative << istage.value>{}.value);
+
+                    // pull data from remote lane
+                    const auto v_remote = warp_shuffle(v_local, src_lane);
+
+                    // reduce
+                    v_local = reduce_func(v_local, v_remote);
+                });
+            }
+        });
+
+        acc_tensor.get_thread_buffer()(i) = v_local;
+    });
+}
+
 // FIXME: this is for 2D to 1D reduce only, need to support n-D
 template <typename AccDistributedTensor_,
           typename InDistributedTensor_,
@@ -175,6 +239,10 @@ CK_TILE_DEVICE void block_tile_reduce(AccDistributedTensor_& acc_tensor,
 #endif
 }
 
+/*
+ * TODO: block_tile_reduce() currently has a limitation
+ * Y dim must have at least one dim not been reduced
+ */
 template <typename AccDataType_,
           typename InDistributedTensor_,
           index_t... InReduceDims,
@@ -208,4 +276,106 @@ CK_TILE_DEVICE auto block_tile_reduce(const InDistributedTensor_& in_tensor,
     return acc_tensor;
 }
 
+// this version only support 2D->1D reduce (reduce-dim=seq<0, 1>)
+// this version only support in/acc/out datatypes are the same
+// this version will call thread/warp+sync in one function call
+//
+template <typename InDistributedTensor_>
+struct BlockReduce2D
+{
+    using InDistributedTensor = remove_cvref_t<InDistributedTensor_>;
+    using InDataType          = typename InDistributedTensor::DataType;
+
+    CK_TILE_HOST_DEVICE BlockReduce2D(const InDistributedTensor& t_, const InDataType& reduce_init_)
+        : t(t_), reduce_init(reduce_init_)
+    {
+    }
+
+    CK_TILE_HOST_DEVICE constexpr auto MakeDstBlockTile() const
+    {
+        using ReduceDim = sequence<1>; // hard coded
+        constexpr auto acc_dstr =
+            make_static_tile_distribution(ck_tile::detail::make_reduce_tile_distribution_encoding(
+                InDistributedTensor::get_tile_distribution()
+                    .get_static_tile_distribution_encoding(),
+                ReduceDim{}));
+
+        return make_static_distributed_tensor<InDataType>(acc_dstr);
+    }
+
+    // return number of pixels each lane need to reduce
+    CK_TILE_HOST_DEVICE constexpr auto get_reduce_length_y() const
+    {
+        constexpr auto spans = InDistributedTensor::get_distributed_spans();
+    }
+
+    // Here ReducePacksPerXDim is not the same meaning as that in static_uford/sweep_tile_uspan
+    // this is number of packs along the X-dim. We need to compute the Unpacks along the Y dim
+    // internally
+    // For simplicity, we just support along the row dimension, ReducePacksPerXDim is always 2
+    // element , and the first element is always ignored For simplicity, will always try from
+    // right-to-left to find alone which Y dim to split
+    template <typename ReduceFunc,
+              typename ReduceSyncFunc,
+              typename ReducePacksPerXDim = uniform_sequence_gen_t<2, 1>>
+    CK_TILE_HOST_DEVICE auto operator()(const ReduceFunc& reduce_func,
+                                        const ReduceSyncFunc& reduce_sync_func,
+                                        ReducePacksPerXDim = {}) const
+    {
+        constexpr auto spans = InDistributedTensor::get_distributed_spans();
+
+        constexpr auto row_y_unpacks = [&]() {
+            constexpr auto row_y_lengths = typename decltype(spans[number<1>{}])::Impl{};
+            constexpr auto row_y_size =
+                reduce_on_sequence(row_y_lengths, multiplies{}, number<1>{});
+            constexpr auto row_y_packs = ReducePacksPerXDim{}.at(number<1>{});
+
+            static_assert(row_y_size % row_y_packs == 0);
+
+            constexpr auto row_y_slice_size = row_y_size / row_y_packs;
+
+            constexpr auto slice_info = slice_sequence(row_y_lengths, number<row_y_slice_size>{});
+            constexpr auto unpacks    = slice_info[number<1>{}];
+            return unpacks;
+        }();
+
+        auto acc_tensor = MakeDstBlockTile();
+
+        // in-thread reduction
+        // FIXME: hard coded to be 2D to 1D reduction
+        sweep_tile_span(spans[number<0>{}], [&](auto dstr_idx_i0) {
+            constexpr auto acc_dstr_idx = make_tuple(dstr_idx_i0);
+
+            auto acc = acc_tensor[acc_dstr_idx];
+
+            sweep_tile_uspan(
+                spans[number<1>{}],
+                [&](auto... dstr_idx_i1) {
+                    acc = reduce_func(acc, t[make_tuple(dstr_idx_i0, dstr_idx_i1)]...);
+                },
+                row_y_unpacks);
+
+            acc_tensor(acc_dstr_idx) = acc;
+        });
+
+        // TODO: always use xor to do cross-lane reduce
+        block_tile_reduce_xor_sync(acc_tensor, reduce_sync_func);
+
+        return acc_tensor;
+    }
+
+    template <typename ReduceFunc>
+    CK_TILE_HOST_DEVICE auto operator()(const ReduceFunc& reduce_func) const
+    {
+        return operator()(reduce_func, reduce_func);
+    }
+
+    InDistributedTensor t;
+    InDataType reduce_init;
+};
+
+// deduction guide
+template <typename T>
+CK_TILE_HOST_DEVICE_EXTERN BlockReduce2D(const T&, const typename T::DataType&)->BlockReduce2D<T>;
+
 } // namespace ck_tile

include/ck_tile/ops/softmax.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/ops/softmax/block/block_softmax_2d.hpp"
+#include "ck_tile/ops/softmax/block/block_softmax_2d_problem.hpp"
+#include "ck_tile/ops/common/tensor_layout.hpp"

include/ck_tile/ops/softmax/block/block_softmax_2d.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include "ck_tile/ops/reduce.hpp"
+
+#define _BLOCK_SOFTMAX_USE_UNPACK2 0
+
+namespace ck_tile {
+
+/*
+simple 2d softmax implementation, along row (dim=1)
+requirement:
+    1). each row is within a warp
+    2). data type must be a dword
+*/
+template <typename Problem_, typename Policy_ = void>
+struct BlockSoftmax2D
+{
+    using Problem = remove_cvref_t<Problem_>;
+    using Policy  = remove_cvref_t<Policy_>;
+
+    using DataType = typename Problem::DataType;
+
+    template <typename DistributedTensor, index_t dim = 1>
+    CK_TILE_DEVICE void
+    operator()(const DistributedTensor& x, DistributedTensor& y, number<dim> = {})
+    {
+        const auto f_max = [](auto e0, auto e1) { return max(e0, e1); };
+        const auto f_sum = [](auto e0, auto e1) { return e0 + e1; };
+#if _BLOCK_SOFTMAX_USE_UNPACK2
+        const auto f_max3 = [](auto e0, auto e1, auto e2) {
+            float rtn;
+            asm volatile("v_max3_f32 %0, %1, %2, %3" : "=v"(rtn) : "v"(e0), "v"(e1), "v"(e2));
+            return rtn;
+        };
+        const auto f_sum3 = [](auto e0, auto e1, auto e2) { return e0 + e1 + e2; };
+#endif
+
+        // compute row max
+        auto reduce_row_max = BlockReduce2D{x, -numeric<DataType>::infinity()};
+#if _BLOCK_SOFTMAX_USE_UNPACK2
+        auto row_max = reduce_row_max(f_max3, f_max, sequence<1, 2>{});
+#else
+        auto row_max = reduce_row_max(f_max);
+#endif
+        sweep_tile<DistributedTensor>([&](auto idx) {
+            constexpr auto row_id = make_tuple(idx[number<0>{}]);
+            y(idx)                = exp(x[idx] - row_max[row_id]);
+        });
+
+        // compute row sum
+        auto reduce_row_sum = BlockReduce2D<decltype(y)>{y, DataType{0}};
+#if _BLOCK_SOFTMAX_USE_UNPACK2
+        auto row_sum = reduce_row_sum(f_sum3, f_sum, sequence<1, 2>{});
+#else
+        auto row_sum = reduce_row_sum(f_sum);
+#endif
+        // reciprocal
+        auto r = make_static_distributed_tensor<DataType>(row_sum.get_tile_distribution());
+        sweep_tile(row_sum, [&](auto idx) { r(idx) = DataType{1} / row_sum(idx); });
+
+        // scale
+        sweep_tile<DistributedTensor>([&](auto idx) {
+            constexpr auto row_id = make_tuple(idx[number<0>{}]);
+            y(idx)                = y(idx) * r(row_id);
+        });
+    }
+
+    template <typename DistributedTensor, index_t dim = 1>
+    CK_TILE_DEVICE decltype(auto) operator()(const DistributedTensor& x, number<dim> = {})
+    {
+        auto y = DistributedTensor{}; // distributed tensor
+        operator()(x, y, number<dim>{});
+        return y;
+    }
+};
+
+} // namespace ck_tile

include/ck_tile/ops/softmax/block/block_softmax_2d_problem.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+
+namespace ck_tile {
+
+template <typename DataType_>
+struct BlockSoftmax2DProblem
+{
+    using DataType = remove_cvref_t<DataType_>;
+};
+
+} // namespace ck_tile

include/ck_tile/ops/topk.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/ops/topk/block/block_topk_stream_2d.hpp"
+#include "ck_tile/ops/topk/block/block_topk_stream_2d_problem.hpp"
+#include "ck_tile/ops/common/tensor_layout.hpp"

include/ck_tile/ops/topk/block/block_topk_stream_2d.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+
+namespace ck_tile {
+
+/*
+simple 2d topk implementation, along row (dim=1)
+requirement:
+    1). each row is within a warp
+*/
+template <typename Problem_, typename Policy_ = void>
+struct BlockTopkStream2D
+{
+    using Problem = remove_cvref_t<Problem_>;
+    using Policy  = remove_cvref_t<Policy_>;
+
+    using DataType  = typename Problem::DataType;
+    using IndexType = typename Problem::IndexType;
+
+    // TODO: if DataType is subdword, need pack into single dword to use argmax
+    struct ArgmaxPacket
+    {
+        DataType arg;
+        index_t value;
+    };
+
+    template <typename DistributedTensor, typename OutWindow, typename IdxWindow, index_t dim = 1>
+    CK_TILE_DEVICE void operator()(const DistributedTensor& x,
+                                   const OutWindow& out_window,
+                                   const IdxWindow& idx_window,
+                                   index_t k,
+                                   number<dim> = {})
+    {
+        OutWindow out_window_tmp = out_window;
+        IdxWindow idx_window_tmp = idx_window;
+        static_assert(
+            std::is_same_v<typename DistributedTensor::DataType, typename OutWindow::DataType> &&
+            std::is_same_v<typename DistributedTensor::DataType, DataType>);
+        static_assert(std::is_same_v<typename IdxWindow::DataType, IndexType>);
+
+        DistributedTensor x_tmp = x;
+        constexpr auto dst_dist = typename IdxWindow::TileDstr{};
+
+        // argmax for topk
+        const auto f_argmax = [](ArgmaxPacket e0, ArgmaxPacket e1) {
+            return e0.arg > e1.arg ? e0 : e1;
+        };
+
+        for(index_t i_k = 0; i_k < k; i_k++)
+        {
+            constexpr auto span_2d = DistributedTensor::get_distributed_spans();
+            auto packet            = [&]() {
+                auto tmp = make_static_distributed_tensor<ArgmaxPacket>(x.get_tile_distribution());
+
+                sweep_tile_span(span_2d[number<0>{}], [&](auto idx0) {
+                    sweep_tile_span(span_2d[number<1>{}], [&](auto idx1) {
+                        const auto tile_idx = get_x_indices_from_distributed_indices(
+                            tmp.get_tile_distribution(), make_tuple(idx0, idx1));
+                        constexpr auto i_j_idx = make_tuple(idx0, idx1);
+                        ArgmaxPacket t;
+                        t.arg        = x_tmp(i_j_idx); // !!! we reference x here
+                        t.value      = tile_idx.at(number<1>{});
+                        tmp(i_j_idx) = t;
+                    });
+                });
+                return tmp;
+            }();
+
+            auto argmax_init = ArgmaxPacket{-numeric<DataType>::infinity(), 0};
+            auto r = block_tile_reduce<ArgmaxPacket>(packet, sequence<1>{}, f_argmax, argmax_init);
+            block_tile_reduce_xor_sync(r, f_argmax);
+
+            auto o = make_static_distributed_tensor<DataType>(dst_dist);
+            auto i = make_static_distributed_tensor<IndexType>(dst_dist);
+            sweep_tile_span(span_2d[number<0>{}], [&](auto idx0) {
+                sweep_tile_span(span_2d[number<1>{}], [&](auto idx1) {
+                    constexpr auto i_j_idx = make_tuple(idx0, idx1);
+                    ArgmaxPacket tmp       = r(i_j_idx);
+                    o(i_j_idx)             = tmp.arg;
+                    i(i_j_idx)             = tmp.value;
+                });
+            });
+
+            // update value
+            sweep_tile_span(span_2d[number<0>{}], [&](auto idx0) {
+                sweep_tile_span(span_2d[number<1>{}], [&](auto idx1) {
+                    const auto tile_idx = get_x_indices_from_distributed_indices(
+                        x.get_tile_distribution(), make_tuple(idx0, idx1));
+                    auto col_id = tile_idx.at(number<1>{});
+
+                    constexpr auto i_j_idx = make_tuple(idx0, idx1);
+
+                    x_tmp(i_j_idx) = (col_id == r(i_j_idx).value) ? -numeric<DataType>::infinity()
+                                                                  : x_tmp(i_j_idx);
+                });
+            });
+
+            if(threadIdx.x % Problem::ColLanes == 0)
+            {
+                store_tile(out_window_tmp, o);
+                store_tile(idx_window_tmp, i);
+            }
+            move_tile_window(out_window_tmp, {number<0>{}, number<1>{}});
+            move_tile_window(idx_window_tmp, {number<0>{}, number<1>{}});
+        }
+    }
+};
+
+} // namespace ck_tile

include/ck_tile/ops/topk/block/block_topk_stream_2d_problem.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+
+namespace ck_tile {
+
+/*
+simple 2d topk implementation, along row (dim=1)
+requirement:
+    1). each row is within a warp
+*/
+template <typename DataType_, typename IndexType_, index_t ColLanes_>
+struct BlockTopkStream2DProblem
+{
+    using DataType                    = remove_cvref_t<DataType_>;
+    using IndexType                   = remove_cvref_t<IndexType_>;
+    static constexpr index_t ColLanes = ColLanes_;
+};
+} // namespace ck_tile

include/ck_tile/ops/topk_softmax.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/ops/topk_softmax/kernel/topk_softmax_kernel.hpp"
+#include "ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_pipeline.hpp"
+#include "ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_policy.hpp"
+#include "ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_problem.hpp"
+#include "ck_tile/ops/common/tensor_layout.hpp"

include/ck_tile/ops/topk_softmax/kernel/topk_softmax_kernel.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include "ck_tile/ops/common.hpp"
+#include "ck_tile/ops/elementwise.hpp"
+#include "ck_tile/host/hip_check_error.hpp"
+#include <string>
+#include <type_traits>
+
+namespace ck_tile {
+
+struct TopkSoftmaxHostArgs
+{
+    const void* p_input;
+    void* p_output;
+    void* p_indices;
+    index_t num_rows;
+    index_t num_experts;
+    index_t topk;
+    index_t stride_input;  // row stride for input, at least experts
+    index_t stride_output; // row stride for output/indices, at least tpok
+};
+
+template <typename Pipeline_>
+struct TopkSoftmaxKernel
+{
+    using Pipeline = remove_cvref_t<Pipeline_>;
+    using Problem  = remove_cvref_t<typename Pipeline::Problem>;
+
+    using InputType  = typename Problem::InputType;
+    using WeightType = typename Problem::WeightType;
+    using IndexType  = typename Problem::IndexType;
+
+    struct TopkSoftmaxKargs
+    {
+        const void* p_input;
+        void* p_output;
+        void* p_indices;
+        index_t num_rows;
+        index_t num_experts;
+        index_t topk;
+        index_t stride_input;  // row stride for input, at least experts
+        index_t stride_output; // row stride for output/indices, at least tpok
+    };
+
+    using Kargs = TopkSoftmaxKargs;
+    using Hargs = TopkSoftmaxHostArgs;
+
+    CK_TILE_HOST static constexpr auto GridSize(const Hargs& h)
+    {
+        if constexpr(Problem::LaunchType > 0)
+        {
+            int num_cu = [&]() {
+                hipDeviceProp_t dev_prop;
+                hipDevice_t dev;
+                HIP_CHECK_ERROR(hipGetDevice(&dev));
+                HIP_CHECK_ERROR(hipGetDeviceProperties(&dev_prop, dev));
+                return dev_prop.multiProcessorCount;
+            }();
+            return dim3(num_cu * Problem::LaunchType);
+        }
+        else
+        {
+            const int num_warps = (h.num_rows + Problem::RowsPerWarp - 1) / Problem::RowsPerWarp;
+            const int num_blocks =
+                (num_warps + Problem::WarpsPerBlock - 1) / Problem::WarpsPerBlock;
+            return dim3(num_blocks);
+        }
+    }
+
+    CK_TILE_HOST static constexpr auto MakeKargs(const Hargs& h)
+    {
+        Kargs k;
+        k.p_input       = h.p_input;
+        k.p_output      = h.p_output;
+        k.p_indices     = h.p_indices;
+        k.num_rows      = h.num_rows;
+        k.num_experts   = h.num_experts;
+        k.topk          = h.topk;
+        k.stride_input  = h.stride_input;
+        k.stride_output = h.stride_output;
+        return k;
+    }
+
+    CK_TILE_HOST_DEVICE static constexpr auto BlockSize() { return Problem::BlockSize; }
+
+    CK_TILE_DEVICE void operator()(Kargs kargs) const
+    {
+        index_t block_row_id = static_cast<index_t>(blockIdx.x * Problem::RowsPerBlock);
+
+        if(block_row_id > kargs.num_rows)
+            return;
+
+        index_t block_os_inp = __builtin_amdgcn_readfirstlane(block_row_id * kargs.stride_input);
+        index_t block_os_out = __builtin_amdgcn_readfirstlane(block_row_id * kargs.stride_output);
+        index_t num_rows_rem = __builtin_amdgcn_readfirstlane(kargs.num_rows - block_row_id);
+
+        const auto input_window = [&]() {
+            const InputType* p_input =
+                reinterpret_cast<const InputType*>(kargs.p_input) + block_os_inp;
+
+            auto tmp = make_naive_tensor_view<address_space_enum::global>(
+                p_input,
+                make_tuple(num_rows_rem, kargs.num_experts),
+                make_tuple(kargs.stride_input, 1),
+                number<Problem::VectorSize>{},
+                number<1>{});
+
+            auto view = pad_tensor_view(
+                tmp,
+                make_tuple(number<Problem::RowsPerBlock>{}, number<Problem::Experts>{}),
+                sequence<0, 1>{}); // out-most dim no need pad(leverage oob)
+
+            return make_tile_window(
+                view,
+                make_tuple(number<Problem::RowsPerBlock>{}, number<Problem::Experts>{}),
+                {0, 0});
+        }();
+
+        auto output_window = [&]() {
+            WeightType* p_output = reinterpret_cast<WeightType*>(kargs.p_output) + block_os_out;
+            auto tmp             = make_naive_tensor_view<address_space_enum::global>(
+                p_output,
+                make_tuple(num_rows_rem, kargs.topk),
+                make_tuple(kargs.stride_output, 1),
+                number<Problem::VectorSize>{},
+                number<1>{});
+            auto view =
+                pad_tensor_view(tmp,
+                                make_tuple(number<Problem::RowsPerBlock>{}, number<1>{}),
+                                sequence<0, 0>{}); // 1. out-most dim no need pad(leverage oob)
+                                                   // 2. we loop over topk 1-1, no need padding
+            return make_tile_window(
+                view, make_tuple(number<Problem::RowsPerBlock>{}, number<1>{}), {0, 0});
+        }();
+
+        auto indices_window = [&]() {
+            IndexType* p_indices = reinterpret_cast<IndexType*>(kargs.p_indices) + block_os_out;
+            auto tmp             = make_naive_tensor_view<address_space_enum::global>(
+                p_indices,
+                make_tuple(num_rows_rem, kargs.topk),
+                make_tuple(kargs.stride_output, 1),
+                number<Problem::VectorSize>{},
+                number<1>{});
+            auto view =
+                pad_tensor_view(tmp,
+                                make_tuple(number<Problem::RowsPerBlock>{}, number<1>{}),
+                                sequence<0, 0>{}); // 1. out-most dim no need pad(leverage oob)
+                                                   // 2. we loop over topk 1-1, no need padding
+            return make_tile_window(
+                view, make_tuple(number<Problem::RowsPerBlock>{}, number<1>{}), {0, 0});
+        }();
+
+        Pipeline{}(input_window,
+                   output_window,
+                   indices_window,
+                   kargs.num_rows,
+                   kargs.num_experts,
+                   kargs.topk,
+                   block_row_id);
+    }
+};
+} // namespace ck_tile

include/ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_pipeline.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include "ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_policy.hpp"
+#include <string>
+#include <type_traits>
+
+#ifndef TOPK_SOFTMAX_USE_RAW_TILE_WINDOW
+#define TOPK_SOFTMAX_USE_RAW_TILE_WINDOW 0
+#endif
+
+namespace ck_tile {
+
+template <typename Problem_, typename Policy_ = TopkSoftmaxWarpPerRowPolicy>
+struct TopkSoftmaxWarpPerRowPipeline
+{
+    // TODO: this kernel only support warp per row
+    using Problem    = remove_cvref_t<Problem_>;
+    using Policy     = remove_cvref_t<Policy_>;
+    using WeightType = typename Problem::WeightType;
+
+    template <typename InputWindow, typename OutputWindow, typename IndexWindow>
+    CK_TILE_DEVICE auto operator()(const InputWindow& input_window,
+                                   OutputWindow& out_window,
+                                   IndexWindow& idx_window,
+                                   index_t rows,
+                                   index_t experts,
+                                   index_t k,
+                                   index_t block_row_id)
+    {
+#if TOPK_SOFTMAX_USE_RAW_TILE_WINDOW
+        auto inp_win = make_tile_window_linear_raw(
+            input_window, Policy::template MakeInputDistribution<Problem>(), sequence<0, 1>{});
+#else
+        auto inp_win = make_tile_window_linear(
+            input_window, Policy::template MakeInputDistribution<Problem>(), sequence<0, 1>{});
+#endif
+        auto out_win = make_tile_window_linear(out_window.get_bottom_tensor_view(),
+                                               out_window.get_window_lengths(),
+                                               out_window.get_window_origin(),
+                                               Policy::template MakeOutputDistribution<Problem>());
+        auto idx_win = make_tile_window_linear(idx_window.get_bottom_tensor_view(),
+                                               idx_window.get_window_lengths(),
+                                               idx_window.get_window_origin(),
+                                               Policy::template MakeOutputDistribution<Problem>());
+
+        auto softmax = Policy::template GetSoftmax<Problem>();
+        auto topk    = Policy::template GetTopk<Problem>();
+
+        const index_t grid_rows_per_loop = gridDim.x * Problem::RowsPerBlock;
+
+        while(1)
+        {
+#if TOPK_SOFTMAX_USE_RAW_TILE_WINDOW
+            __builtin_amdgcn_sched_barrier(0);
+            auto x =
+                load_tile_raw(inp_win, number<-1>{}, bool_constant<true>{}, bool_constant<true>{});
+            buffer_load_fence(number<0>{});
+            __builtin_amdgcn_sched_barrier(0);
+#else
+            auto x = load_tile(inp_win);
+#endif
+            // cast and pad input data
+            auto w = [&]() {
+#if 0
+                auto w_ = cast_tile<WeightType>(x);
+
+                constexpr auto span_2d = decltype(w_)::get_distributed_spans();
+                sweep_tile_span(span_2d[number<0>{}], [&](auto idx0) {
+                    sweep_tile_span(span_2d[number<1>{}], [&](auto idx1) {
+                        constexpr auto i_j_idx = make_tuple(idx0, idx1);
+                        const auto x_indices   = get_x_indices_from_distributed_indices(
+                            w_.get_tile_distribution(), i_j_idx);
+                        const auto current_expert = x_indices.at(number<1>{});
+                        // set to -INF if OOB so that later softmax can work properly
+                        w_(i_j_idx) = current_expert >= experts ? -numeric<WeightType>::infinity()
+                                                                : w_(i_j_idx);
+                    });
+                });
+                return w_;
+#else
+                auto w_  = make_static_distributed_tensor<WeightType>(x.get_tile_distribution());
+                auto w_f = [&](auto idx) {
+                    w_(idx) = type_convert<WeightType>(x(idx));
+                    const auto x_indices =
+                        get_x_indices_from_distributed_indices(w_.get_tile_distribution(), idx);
+                    const auto current_expert = x_indices.at(number<1>{});
+                    w_(idx) =
+                        current_expert >= experts ? -numeric<WeightType>::infinity() : w_(idx);
+                };
+                tile_sweeper ts{w_, w_f};
+                ts();
+                return w_;
+#endif
+            }();
+
+            // softmax
+            auto y = softmax(w);
+
+            topk(y, out_win, idx_win, k);
+
+            // check exit
+            if constexpr(Problem::LaunchType == 0)
+            {
+                break;
+            }
+            else
+            {
+                block_row_id += grid_rows_per_loop;
+                if(block_row_id >= rows)
+                    break;
+            }
+
+            move_tile_window(inp_win, {grid_rows_per_loop, number<0>{}});
+            move_tile_window(out_win, {grid_rows_per_loop, number<0>{}});
+            move_tile_window(idx_win, {grid_rows_per_loop, number<0>{}});
+        }
+    }
+};
+} // namespace ck_tile

include/ck_tile/ops/topk_softmax/pipeline/topk_softmax_warp_per_row_policy.hpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#pragma once
+
+#include "ck_tile/core.hpp"
+#include "ck_tile/ops/softmax.hpp"
+#include "ck_tile/ops/topk.hpp"
+
+namespace ck_tile {
+
+struct TopkSoftmaxWarpPerRowPolicy
+{
+    template <typename Problem>
+    CK_TILE_HOST_DEVICE static constexpr auto MakeInputDistribution()
+    {
+        // TODO: Y dim must have one dim that is not reduced
+        return make_static_tile_distribution(
+            tile_distribution_encoding<
+                sequence<1>,
+                tuple<sequence<Problem::IssuesPerCol,
+                               Problem::WarpsPerBlock,
+                               Problem::RowsPerWarpPerColIssue>,
+                      sequence<Problem::IssuesPerRow, Problem::LanesPerRow, Problem::VectorSize>>,
+                tuple<sequence<1>, sequence<1, 2>>,
+                tuple<sequence<1>, sequence<2, 1>>,
+                sequence<1, 2, 2>,
+                sequence<0, 0, 2>>{});
+    }
+
+    template <typename Problem>
+    CK_TILE_HOST_DEVICE static constexpr auto MakeOutputDistribution()
+    {
+        return make_static_tile_distribution(
+            tile_distribution_encoding<sequence<Problem::LanesPerRow>, // repeat this one
+                                       tuple<sequence<Problem::IssuesPerCol,
+                                                      Problem::WarpsPerBlock,
+                                                      Problem::RowsPerWarpPerColIssue>,
+                                             sequence<1>>, // each row write out single element
+                                       tuple<sequence<1>, sequence<1, 0>>,
+                                       tuple<sequence<1>, sequence<2, 0>>,
+                                       sequence<1, 2>,
+                                       sequence<0, 0>>{});
+    }
+
+    template <typename Problem>
+    CK_TILE_HOST_DEVICE static constexpr auto GetSoftmax()
+    {
+        using softmax_problem = BlockSoftmax2DProblem<typename Problem::WeightType>;
+        return BlockSoftmax2D<softmax_problem>{};
+    }
+
+    template <typename Problem>
+    CK_TILE_HOST_DEVICE static constexpr auto GetTopk()
+    {
+        using topk_problem = BlockTopkStream2DProblem<typename Problem::WeightType,
+                                                      typename Problem::IndexType,
+                                                      Problem::LanesPerRow>;
+        // Note: replicate is LanesPerRow
+        return BlockTopkStream2D<topk_problem>{};
+    }
+};
+} // namespace ck_tile