feat(dnn/cuda): add cutlass matmul using split k parallel

GitOrigin-RevId: 650209e35f

dnn/scripts/Makefile

@@ -9,9 +9,9 @@ ELEMWISE_IMPL := ../src/cuda/cond_take/kimpl \
 	../src/cuda/elemwise_multi_type/kimpl
 
 CUDA_CONV_IMPL := ../src/cuda/conv_bias/int8/kimpl ../src/cuda/conv_bias/int8_imma/kimpl ../src/cuda/batch_conv_bias/int8/kimpl 
-CUDA_MATMUL_KIMPL := ../src/cuda/matrix_mul/fp32_simt/kimpl
+CUDA_MATMUL_IMPL := ../src/cuda/matrix_mul/fp32_simt/kimpl
 
-all: ${PARAM_DEFS} ${ELEMWISE_IMPL} ${CUDA_CONV_IMPL} $(CUDA_MATMUL_KIMPL)
+all: ${PARAM_DEFS} ${ELEMWISE_IMPL} ${CUDA_CONV_IMPL} $(CUDA_MATMUL_IMPL)
 
 ../src/common/elemwise/each_mode.inl: gen_elemwise_each_mode.py
 	./$^ $@

dnn/src/cuda/matrix_mul/algos.cpp

@@ -37,6 +37,9 @@ MatrixMulForwardImpl::AlgoPack::AlgoPack() {
     for (auto&& algo : simt_float32) {
         all_algos.push_back(&algo);
     }
+    for (auto&& algo : simt_float32_split_k) {
+        all_algos.push_back(&algo);
+    }
 
     for (auto&& algo : all_algos) {
         m_all_algos_map.emplace(algo->info().desc, algo);
@@ -62,6 +65,23 @@ void MatrixMulForwardImpl::AlgoPack::fill_cutlass_algos() {
     simt_float32.emplace_back(AlgoParam{16, 32, 8, 16, 32, 8});
     simt_float32.emplace_back(AlgoParam{16, 64, 8, 16, 64, 8});
     simt_float32.emplace_back(AlgoParam{16, 128, 8, 16, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{64, 256, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{256, 64, 8, 64, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{32, 256, 8, 16, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{256, 32, 8, 64, 16, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{128, 128, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{128, 64, 8, 64, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{64, 128, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{128, 32, 8, 64, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{32, 128, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{64, 64, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{32, 64, 8, 32, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{64, 32, 8, 64, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{32, 32, 8, 32, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{8, 32, 8, 8, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{16, 32, 8, 16, 32, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{16, 64, 8, 16, 64, 8});
+    simt_float32_split_k.emplace_back(AlgoParam{16, 128, 8, 16, 64, 8});
 }
 
 MatrixMulForwardImpl::AlgoPack MatrixMulForwardImpl::sm_algo_pack;

dnn/src/cuda/matrix_mul/algos.h

@@ -43,6 +43,7 @@ public:
         CUDA_NAIVE,
         CUDA_BFLOAT16, 
         CUDA_FLOAT32_SIMT, 
+        CUDA_FLOAT32_SIMT_SPLIT_K, 
     };
     using Mapper = std::unordered_map<AlgorithmDesc, AlgoBase*>;
 
@@ -198,6 +199,31 @@ private:
     std::string m_name;
 };
 
+class MatrixMulForwardImpl::AlgoFloat32SIMTSplitK final : public AlgoBase {
+public:
+    using AlgoParam = MatrixMulForwardImpl::AlgoFloat32SIMT::AlgoParam;
+    AlgoFloat32SIMTSplitK(AlgoParam algo_param)
+            : m_algo_param{algo_param},
+              m_name{ssprintf("CUTLASS_FLOAT32_SIMT_SPLIT_K_%s",
+                              m_algo_param.to_string().c_str())} {}
+    bool is_available(const SizeArgs& args) const override;
+    size_t get_workspace_in_bytes(const SizeArgs& args) const override;
+    const char* name() const override { return m_name.c_str(); }
+    void exec(const ExecArgs& args) const override;
+    bool is_reproducible() const override { return true; }
+    MEGDNN_DECL_ALGO_TYPE(CUDA_FLOAT32_SIMT_SPLIT_K)
+
+    std::string param() const override {
+        std::string ret;
+        serialize_write_pod(m_algo_param, ret);
+        return ret;
+    }
+
+private:
+    AlgoParam m_algo_param;
+    std::string m_name;
+};
+
 class MatrixMulForwardImpl::AlgoPack : NonCopyableObj {
 private:
     AlgoBase::Mapper m_all_algos_map;
@@ -216,6 +242,7 @@ public:
     AlgoBFloat16 bfloat16;
 #endif
     std::vector<AlgoFloat32SIMT> simt_float32;
+    std::vector<AlgoFloat32SIMTSplitK> simt_float32_split_k;
     std::vector<AlgoBase*> all_algos;
 
     const AlgoBase::Mapper& all_algos_map() const { return m_all_algos_map; }

dnn/src/cuda/matrix_mul/cutlass_float32_simt_split_k.cpp

@@ -0,0 +1,76 @@
+/**
+ * \file dnn/src/cuda/matrix_mul/cutlass_float32_simt_split_k.cpp
+ * MegEngine is Licensed under the Apache License, Version 2.0 (the "License")
+ *
+ * Copyright (c) 2014-2020 Megvii Inc. All rights reserved.
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT ARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ * implied.
+ */
+
+#include "src/cuda/handle.h"
+#include "src/cuda/matrix_mul/algos.h"
+#include "src/cuda/matrix_mul/cutlass_matrix_mul_wrapper.cuh"
+#include "src/cuda/utils.h"
+
+using namespace megdnn;
+using namespace cuda;
+using namespace cutlass_wrapper;
+
+bool MatrixMulForwardImpl::AlgoFloat32SIMTSplitK::is_available(
+        const SizeArgs& args) const {
+    auto&& param = args.opr->param();
+    int m = args.layout_c.shape[0], n = args.layout_c.shape[1],
+        k = args.layout_a.shape[param.transposeA ? 0 : 1];
+    return args.opr->param().format == param::MatrixMul::Format::DEFAULT &&
+           args.layout_a.dtype == dtype::Float32() &&
+           args.layout_b.dtype == dtype::Float32() &&
+           args.layout_c.dtype == dtype::Float32() && k > std::max(m, n);
+}
+
+size_t MatrixMulForwardImpl::AlgoFloat32SIMTSplitK::get_workspace_in_bytes(
+        const SizeArgs& args) const {
+    size_t lda = args.layout_a.stride[0], ldb = args.layout_b.stride[0],
+           ldc = args.layout_c.stride[0];
+    auto&& param = args.opr->param();
+    int m = args.layout_c.shape[0], n = args.layout_c.shape[1],
+        k = args.layout_a.shape[param.transposeA ? 0 : 1];
+    GemmCoord problem_size{m, n, k};
+    int split_k_slices = k / std::max(m, n);
+    return cutlass_matrix_mul_float32_simt_get_workspace_size(
+            param.transposeA, lda, param.transposeB, ldb, ldc, problem_size,
+            1.f, 0.f,
+            GemmCoord{m_algo_param.threadblock_m, m_algo_param.threadblock_n,
+                      m_algo_param.threadblock_k},
+            GemmCoord{m_algo_param.warp_m, m_algo_param.warp_n,
+                      m_algo_param.warp_k},
+            split_k_slices);
+}
+
+void MatrixMulForwardImpl::AlgoFloat32SIMTSplitK::exec(
+        const ExecArgs& args) const {
+    size_t lda = args.tensor_a.layout.stride[0],
+           ldb = args.tensor_b.layout.stride[0],
+           ldc = args.tensor_c.layout.stride[0];
+    auto&& param = args.opr->param();
+    int m = args.tensor_c.layout.shape[0], n = args.tensor_c.layout.shape[1],
+        k = args.tensor_a.layout.shape[param.transposeA ? 0 : 1];
+    GemmCoord problem_size{m, n, k};
+    int split_k_slices = k / std::max(m, n);
+    auto&& stream = cuda_stream(args.opr->handle());
+    int* workspace = reinterpret_cast<int*>(args.workspace.raw_ptr);
+    return cutlass_matrix_mul_float32_simt(
+            args.tensor_a.ptr<dt_float32>(), param.transposeA, lda,
+            args.tensor_b.ptr<dt_float32>(), param.transposeB, ldb,
+            args.tensor_c.ptr<dt_float32>(), ldc, workspace, problem_size, 1.f,
+            0.f,
+            GemmCoord{m_algo_param.threadblock_m, m_algo_param.threadblock_n,
+                      m_algo_param.threadblock_k},
+            GemmCoord{m_algo_param.warp_m, m_algo_param.warp_n,
+                      m_algo_param.warp_k},
+            stream, split_k_slices);
+}
+
+// vim: syntax=cpp.doxygen

dnn/src/cuda/matrix_mul/cutlass_matrix_mul_wrapper.cu

@@ -18,6 +18,7 @@
 #if __CUDACC_VER_MAJOR__ > 9 || \
         (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
 #include "cutlass/gemm/device/gemm.h"
+#include "cutlass/gemm/device/gemm_splitk_parallel.h"
 #endif
 #include "src/common/opr_param_defs_enumv.cuh"
 #include "src/cuda/matrix_mul/cutlass_matrix_mul_wrapper.cuh"
@@ -62,14 +63,20 @@ void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_float32_simt(
         float* /* d_C */, size_t /* ldc */, int* /* workspace */,
         GemmCoord const& /* problem_size */, float /* alpha */,
         float /* beta */, const GemmCoord& /* threadblock_shape */,
-        const GemmCoord& /* warp_shape */, cudaStream_t /* stream */) {}
+        const GemmCoord& /* warp_shape */, cudaStream_t /* stream */,
+        int /* split_k_slices */) {}
 #else
 void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_float32_simt(
         const float* d_A, bool transpose_A, size_t lda, const float* d_B,
         bool transpose_B, size_t ldb, float* d_C, size_t ldc, int* workspace,
         GemmCoord const& problem_size, float alpha, float beta,
         const GemmCoord& threadblock_shape, const GemmCoord& warp_shape,
-        cudaStream_t stream) {
+        cudaStream_t stream, int split_k_slices) {
+    static constexpr int kEpilogueElementsPerAccess = 1;
+    using EpilogueOp = cutlass::epilogue::thread::LinearCombination<
+            float, kEpilogueElementsPerAccess, float, float>;
+    typename EpilogueOp::Params epilogue{alpha, beta};
+    if (split_k_slices == 1) {
 #define cb(threadblock_m_, threadblock_n_, threadblock_k_, warp_m_, warp_n_,   \
            warp_k_)                                                            \
     if (threadblock_shape.m() == threadblock_m_ &&                             \
@@ -93,29 +100,67 @@ void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_float32_simt(
                                                 workspace, problem_size,       \
                                                 epilogue, stream);             \
     }
-    static constexpr int kEpilogueElementsPerAccess = 1;
-    using EpilogueOp = cutlass::epilogue::thread::LinearCombination<
-            float, kEpilogueElementsPerAccess, float, float>;
-    typename EpilogueOp::Params epilogue{alpha, beta};
-    if (!transpose_A && !transpose_B) {
-        using LayoutA = cutlass::layout::RowMajor;
-        using LayoutB = cutlass::layout::RowMajor;
-        DISPATCH(cb)
-    } else if (!transpose_A && transpose_B) {
-        using LayoutA = cutlass::layout::RowMajor;
-        using LayoutB = cutlass::layout::ColumnMajor;
-        DISPATCH(cb)
-    } else if (transpose_A && !transpose_B) {
-        using LayoutA = cutlass::layout::ColumnMajor;
-        using LayoutB = cutlass::layout::RowMajor;
-        DISPATCH(cb)
+        if (!transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else if (!transpose_A && transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        } else if (transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else {
+            megdnn_assert(transpose_A && transpose_B);
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        }
+#undef cb
     } else {
-        megdnn_assert(transpose_A && transpose_B);
-        using LayoutA = cutlass::layout::ColumnMajor;
-        using LayoutB = cutlass::layout::ColumnMajor;
-        DISPATCH(cb)
+#define cb(threadblock_m_, threadblock_n_, threadblock_k_, warp_m_, warp_n_,   \
+           warp_k_)                                                            \
+    if (threadblock_shape.m() == threadblock_m_ &&                             \
+        threadblock_shape.n() == threadblock_n_ &&                             \
+        threadblock_shape.k() == threadblock_k_ &&                             \
+        warp_shape.m() == warp_m_ && warp_shape.n() == warp_n_ &&              \
+        warp_shape.k() == warp_k_) {                                           \
+        using ThreadBlockShape =                                               \
+                cutlass::gemm::GemmShape<threadblock_m_, threadblock_n_,       \
+                                         threadblock_k_>;                      \
+        using WarpShape = cutlass::gemm::GemmShape<warp_m_, warp_n_, warp_k_>; \
+        using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;            \
+        using Gemm = cutlass::gemm::device::GemmSplitKParallel<                \
+                float, LayoutA, float, LayoutB, float,                         \
+                cutlass::layout::RowMajor, float, cutlass::arch::OpClassSimt,  \
+                cutlass::arch::Sm50, ThreadBlockShape, WarpShape,              \
+                InstructionShape, EpilogueOp>;                                 \
+        return cutlass_matrix_mul_wrapper<Gemm>(                               \
+                d_A, lda, d_B, ldb, d_C, ldc, workspace, problem_size,         \
+                epilogue, stream, split_k_slices);                             \
     }
+        if (!transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else if (!transpose_A && transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        } else if (transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else {
+            megdnn_assert(transpose_A && transpose_B);
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        }
 #undef cb
+    }
 }
 #endif
 
@@ -127,7 +172,7 @@ size_t megdnn::cuda::cutlass_wrapper::
                 bool /* transpose_B */, size_t /* ldb */, size_t /* ldc */,
                 GemmCoord const& /* problem_size */, float /* alpha */,
                 float /* beta */, const GemmCoord& /* threadblock_shape */,
-                const GemmCoord& /* warp_shape */) {
+                const GemmCoord& /* warp_shape */, int /* split_k_slices */) {
     return 0;
 }
 #else
@@ -136,7 +181,12 @@ size_t megdnn::cuda::cutlass_wrapper::
                 bool transpose_A, size_t lda, bool transpose_B, size_t ldb,
                 size_t ldc, GemmCoord const& problem_size, float alpha,
                 float beta, const GemmCoord& threadblock_shape,
-                const GemmCoord& warp_shape) {
+                const GemmCoord& warp_shape, int split_k_slices) {
+    static constexpr int kEpilogueElementsPerAccess = 1;
+    using EpilogueOp = cutlass::epilogue::thread::LinearCombination<
+            float, kEpilogueElementsPerAccess, float, float>;
+    typename EpilogueOp::Params epilogue{alpha, beta};
+    if (split_k_slices == 1) {
 #define cb(threadblock_m_, threadblock_n_, threadblock_k_, warp_m_, warp_n_,   \
            warp_k_)                                                            \
     if (threadblock_shape.m() == threadblock_m_ &&                             \
@@ -169,30 +219,80 @@ size_t megdnn::cuda::cutlass_wrapper::
                                            split_k_slices};                    \
         return Gemm::get_workspace_size(arguments);                            \
     }
-    static constexpr int kEpilogueElementsPerAccess = 1;
-    static constexpr int split_k_slices = 1;
-    using EpilogueOp = cutlass::epilogue::thread::LinearCombination<
-            float, kEpilogueElementsPerAccess, float, float>;
-    typename EpilogueOp::Params epilogue{alpha, beta};
-    if (!transpose_A && !transpose_B) {
-        using LayoutA = cutlass::layout::RowMajor;
-        using LayoutB = cutlass::layout::RowMajor;
-        DISPATCH(cb)
-    } else if (!transpose_A && transpose_B) {
-        using LayoutA = cutlass::layout::RowMajor;
-        using LayoutB = cutlass::layout::ColumnMajor;
-        DISPATCH(cb)
-    } else if (transpose_A && !transpose_B) {
-        using LayoutA = cutlass::layout::ColumnMajor;
-        using LayoutB = cutlass::layout::RowMajor;
-        DISPATCH(cb)
+        if (!transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else if (!transpose_A && transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        } else if (transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else {
+            megdnn_assert(transpose_A && transpose_B);
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        }
+#undef cb
     } else {
-        megdnn_assert(transpose_A && transpose_B);
-        using LayoutA = cutlass::layout::ColumnMajor;
-        using LayoutB = cutlass::layout::ColumnMajor;
-        DISPATCH(cb)
+#define cb(threadblock_m_, threadblock_n_, threadblock_k_, warp_m_, warp_n_,   \
+           warp_k_)                                                            \
+    if (threadblock_shape.m() == threadblock_m_ &&                             \
+        threadblock_shape.n() == threadblock_n_ &&                             \
+        threadblock_shape.k() == threadblock_k_ &&                             \
+        warp_shape.m() == warp_m_ && warp_shape.n() == warp_n_ &&              \
+        warp_shape.k() == warp_k_) {                                           \
+        using ThreadBlockShape =                                               \
+                cutlass::gemm::GemmShape<threadblock_m_, threadblock_n_,       \
+                                         threadblock_k_>;                      \
+        using WarpShape = cutlass::gemm::GemmShape<warp_m_, warp_n_, warp_k_>; \
+        using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;            \
+        using Gemm = cutlass::gemm::device::GemmSplitKParallel<                \
+                float, LayoutA, float, LayoutB, float,                         \
+                cutlass::layout::RowMajor, float, cutlass::arch::OpClassSimt,  \
+                cutlass::arch::Sm50, ThreadBlockShape, WarpShape,              \
+                InstructionShape, EpilogueOp>;                                 \
+        using TensorRefA = cutlass::TensorRef<typename Gemm::ElementA const,   \
+                                              typename Gemm::LayoutA>;         \
+        using TensorRefB = cutlass::TensorRef<typename Gemm::ElementB const,   \
+                                              typename Gemm::LayoutB>;         \
+        using TensorRefC = cutlass::TensorRef<typename Gemm::ElementC const,   \
+                                              typename Gemm::LayoutC>;         \
+        using TensorRefD = cutlass::TensorRef<typename Gemm::ElementC,         \
+                                              typename Gemm::LayoutC>;         \
+        TensorRefA tensor_A{nullptr, Gemm::LayoutA{static_cast<int>(lda)}};    \
+        TensorRefB tensor_B{nullptr, Gemm::LayoutB{static_cast<int>(ldb)}};    \
+        TensorRefC tensor_C{nullptr, Gemm::LayoutC{static_cast<int>(ldc)}};    \
+        TensorRefD tensor_D{nullptr, Gemm::LayoutC{static_cast<int>(ldc)}};    \
+        typename Gemm::Arguments arguments{problem_size,  tensor_A, tensor_B,  \
+                                           tensor_C,      tensor_D, epilogue,  \
+                                           split_k_slices};                    \
+        return Gemm::get_workspace_size(arguments);                            \
     }
+        if (!transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else if (!transpose_A && transpose_B) {
+            using LayoutA = cutlass::layout::RowMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        } else if (transpose_A && !transpose_B) {
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::RowMajor;
+            DISPATCH(cb)
+        } else {
+            megdnn_assert(transpose_A && transpose_B);
+            using LayoutA = cutlass::layout::ColumnMajor;
+            using LayoutB = cutlass::layout::ColumnMajor;
+            DISPATCH(cb)
+        }
 #undef cb
+    }
 }
 #endif
 

dnn/src/cuda/matrix_mul/cutlass_matrix_mul_wrapper.cuh

@@ -26,19 +26,19 @@ void cutlass_matrix_mul_wrapper(
         typename Gemm::ElementC* d_C, size_t ldc, int* workspace,
         GemmCoord const& problem_size,
         typename Gemm::EpilogueOutputOp::Params const& epilogue,
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices = 1);
 
 void cutlass_matrix_mul_float32_simt(
         const float* d_A, bool transpose_A, size_t lda, const float* d_B,
         bool transpose_B, size_t ldb, float* d_C, size_t ldc, int* workspace,
         GemmCoord const& problem_size, float alpha, float beta,
         const GemmCoord& threadblock_shape, const GemmCoord& warp_shape,
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices = 1);
 
 size_t cutlass_matrix_mul_float32_simt_get_workspace_size(
         bool transpose_A, size_t lda, bool transpose_B, size_t ldb, size_t ldc,
         GemmCoord const& problem_size, float alpha, float beta,
-        const GemmCoord& threadblock_shape, const GemmCoord& warp_shape);
+        const GemmCoord& threadblock_shape, const GemmCoord& warp_shape, int split_k_slices = 1);
 
 }  // namespace cutlass_wrapper
 }  // namespace cuda

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x128x8_32x64x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x32x8_64x32x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_128x64x8_64x32x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<128, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x128x8_16x64x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x32x8_16x32x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_16x64x8_16x64x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 16, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 16, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 16, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x32x8_64x16x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 16, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_256x64x8_64x32x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<256, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<64, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x128x8_32x64x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 128, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 256, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 256, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 256, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x256x8_16x64x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 256, 8>;
+using WarpShape = cutlass::gemm::GemmShape<16, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_tt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x32x8_32x32x8_tt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 32, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_nn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_nn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_nt.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::RowMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_nt_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::RowMajor;
+using LayoutB = cutlass::layout::ColumnMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_tn.cu

@@ -5,6 +5,7 @@
 #pragma GCC diagnostic ignored "-Wunused-parameter"
 #pragma GCC diagnostic ignored "-Wstrict-aliasing"
 #pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
 #include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
 
 using LayoutA = cutlass::layout::ColumnMajor;
@@ -28,6 +29,7 @@ template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
         int* workspace, 
         cutlass::gemm::GemmCoord const& problem_size,   
         typename Gemm::EpilogueOutputOp::Params const& epilogue, 
-        cudaStream_t stream);
+        cudaStream_t stream, int split_k_slices);
+
 #pragma GCC diagnostic pop
 #endif

dnn/src/cuda/matrix_mul/fp32_simt/kimpl/matrix_mul_fp32_simt_32x64x8_32x64x8_tn_splitk_parallel.cu

@@ -0,0 +1,33 @@
+#if __CUDACC_VER_MAJOR__ > 9 || (__CUDACC_VER_MAJOR__ == 9 && __CUDACC_VER_MINOR__ >= 2)
+// generated by gen_cutlass_matrix_mul_kern_impls.py
+// ignore warning of cutlass
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wunused-parameter"
+#pragma GCC diagnostic ignored "-Wstrict-aliasing"
+#pragma GCC diagnostic ignored "-Wuninitialized"
+#pragma GCC diagnostic ignored "-Wmaybe-uninitialized"
+#include "src/cuda/matrix_mul/fp32_simt/matrix_mul_float_simt_cutlass_wrapper.cuinl"
+
+using LayoutA = cutlass::layout::ColumnMajor;
+using LayoutB = cutlass::layout::RowMajor;
+using ThreadBlockShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using WarpShape = cutlass::gemm::GemmShape<32, 64, 8>;
+using InstructionShape = cutlass::gemm::GemmShape<1, 1, 1>;
+using EpilogueOp = cutlass::epilogue::thread::LinearCombination<float, 1, float, float>;
+using Gemm = cutlass::gemm::device::GemmSplitKParallel<
+    float, LayoutA, 
+    float, LayoutB, 
+    float, cutlass::layout::RowMajor, float, 
+    cutlass::arch::OpClassSimt, cutlass::arch::Sm50, 
+    ThreadBlockShape, WarpShape, InstructionShape, EpilogueOp>;
+template void megdnn::cuda::cutlass_wrapper::cutlass_matrix_mul_wrapper<Gemm>(
+        const typename Gemm::ElementA* d_A, size_t lda, 
+        const typename Gemm::ElementB* d_B, size_t ldb,  
+        typename Gemm::ElementC* d_C, size_t ldc,  
+        int* workspace, 
+        cutlass::gemm::GemmCoord const& problem_size,   
+        typename Gemm::EpilogueOutputOp::Params const& epilogue, 
+        cudaStream_t stream, int split_k_slices);
+
+#pragma GCC diagnostic pop
+#endif