Merge remote-tracking branch 'origin/develop' into cpu_avx2

library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp

@@ -18,13 +18,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp

@@ -17,13 +17,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv2d_fwd/device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp

@@ -17,13 +17,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv2d_fwd_bias_relu/device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp

@@ -18,19 +18,19 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 using AddRelu     = ck::tensor_operation::element_wise::AddRelu;
 
-static constexpr auto MemorySet = ck::InMemoryDataOperationEnum_t::Set;
+static constexpr auto MemorySet = ck::InMemoryDataOperationEnum::Set;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 static constexpr auto ConvFwdOddC =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::OddC;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::OddC;
 
 // arbitrary conv
 using device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv2d_fwd_bias_relu_add/device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp

@@ -19,16 +19,16 @@ using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 using AddReluAdd  = ck::tensor_operation::element_wise::AddReluAdd;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 static constexpr auto ConvFwdOddC =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::OddC;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::OddC;
 
 // arbitrary conv
 using device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv2d_fwd_bias_relu_atomic_add/device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instance.cpp

@@ -18,10 +18,10 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 using AddRelu     = ck::tensor_operation::element_wise::AddRelu;
 
-static constexpr auto InMemoryAtomicAdd = ck::InMemoryDataOperationEnum_t::AtomicAdd;
+static constexpr auto InMemoryAtomicAdd = ck::InMemoryDataOperationEnum::AtomicAdd;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 using device_conv2d_fwd_xdl_c_shuffle_bias_relu_atomic_add_nhwc_kyxc_nhwk_f16_instances = std::tuple<
     // clang-format off

library/src/tensor_operation_instance/gpu/conv3d_fwd/device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp

@@ -18,13 +18,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_bf16_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv3d_fwd/device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp

@@ -18,13 +18,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_f16_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv3d_fwd/device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp

@@ -17,13 +17,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_f32_instances = std::tuple<

library/src/tensor_operation_instance/gpu/conv3d_fwd/device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_int8_instance.cpp

@@ -17,13 +17,13 @@ using S = ck::Sequence<Is...>;
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 
 static constexpr auto ConvFwdDefault =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
 
 static constexpr auto ConvFwd1x1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Pad0;
 
 static constexpr auto ConvFwd1x1S1P0 =
-    ck::tensor_operation::device::ConvolutionForwardSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionForwardSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_fwd_xdl_ndhwc_kzyxc_ndhwk_int8_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv1d_bwd_data_xdl_nwc_kxc_nwk_bf16_instance.cpp

@@ -9,7 +9,7 @@ namespace tensor_operation {
 namespace device {
 namespace device_conv2d_bwd_data_instance {
 
-using BF16 = ushort;
+using BF16 = bhalf_t;
 using F32  = float;
 
 template <ck::index_t... Is>
@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv1d_bwd_data_xdl_nwc_kxc_nwk_bf16_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f16_instance.cpp

@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f16_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f32_instance.cpp

@@ -16,10 +16,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f32_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv1d_bwd_data_xdl_nwc_kxc_nwk_int8_instance.cpp

@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv1d_bwd_data_xdl_nwc_kxc_nwk_int8_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp

@@ -9,7 +9,7 @@ namespace tensor_operation {
 namespace device {
 namespace device_conv2d_bwd_data_instance {
 
-using BF16 = ushort;
+using BF16 = bhalf_t;
 using F32  = float;
 
 template <ck::index_t... Is>
@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f16_instance.cpp

@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f16_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f32_instance.cpp

@@ -16,10 +16,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f32_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_int8_instance.cpp

@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_int8_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp

@@ -9,7 +9,7 @@ namespace tensor_operation {
 namespace device {
 namespace device_conv2d_bwd_data_instance {
 
-using BF16 = ushort;
+using BF16 = bhalf_t;
 using F32  = float;
 
 template <ck::index_t... Is>
@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_bf16_instances =

library/src/tensor_operation_instance/gpu/convnd_bwd_data/device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp

@@ -17,10 +17,10 @@ using S = ck::Sequence<Is...>;
 
 using PassThrough = ck::tensor_operation::element_wise::PassThrough;
 static constexpr auto ConvBwdDataDefault =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Default;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Default;
 
 static constexpr auto ConvBwdDataFilter1x1Stride1Pad0 =
-    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization_t::Filter1x1Stride1Pad0;
+    ck::tensor_operation::device::ConvolutionBackwardDataSpecialization::Filter1x1Stride1Pad0;
 
 // Compilation parameters for in[n, hi, wi, c] * wei[k, y, x, c] = out[n, ho, wo, k]
 using device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f16_instances =