Merge branch 'uif2-initial' into uif2-migraphx

library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_scaleadd_relu.hpp

@@ -7,7 +7,7 @@
 #include <memory>
 
 #include "ck/ck.hpp"
-#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_abd.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
 
@@ -24,7 +24,7 @@ using ScaleAddScaleAddRelu = ck::tensor_operation::element_wise::ScaleAddScaleAd
 #ifdef CK_ENABLE_BF16
 // grouped conv3d forward, NDHWGC/GKZYXC/NDHWGK
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhwgk_bf16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<NDHWGK, NDHWGK>,
@@ -40,7 +40,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhw
 
 #ifdef CK_ENABLE_FP16
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhwgk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<NDHWGK, NDHWGK>,
@@ -56,7 +56,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhw
 
 #ifdef CK_ENABLE_FP32
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhwgk_f32_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<NDHWGK, NDHWGK>,
@@ -72,7 +72,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhw
 
 #ifdef CK_ENABLE_INT8
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_scaleadd_relu_ndhwgc_gkzyxc_ndhwgk_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<NDHWGK, NDHWGK>,
@@ -96,7 +96,7 @@ template <ck::index_t NumDimSpatial,
           typename DDataTypes,
           typename OutDataType,
           typename ComputeType>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -112,7 +112,7 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ComputeType>>
 {
     using DeviceOp =
-        DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DLayouts,

library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perchannel_quantization.hpp

@@ -7,7 +7,7 @@
 
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
-#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_abd.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
 
 #include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
@@ -20,7 +20,7 @@ namespace instance {
 // grouped conv2d forward, NHWGC/GKYXC/NHWGK
 void add_device_conv2d_dl_bias_perchannel_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_GK_Tuple,
@@ -35,7 +35,7 @@ void add_device_conv2d_dl_bias_perchannel_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_dl_bias_relu_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_GK_Tuple,
@@ -51,7 +51,7 @@ void add_device_conv2d_dl_bias_relu_perchannel_quantization_int8_instances(
 
 void add_device_conv2d_dl_bias_tanh_perchannel_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_GK_Tuple,
@@ -67,7 +67,7 @@ void add_device_conv2d_dl_bias_tanh_perchannel_quantization_int8_instances(
 #endif
 void add_device_conv2d_xdl_bias_perchannel_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_GK_Tuple,
@@ -82,7 +82,7 @@ void add_device_conv2d_xdl_bias_perchannel_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_xdl_bias_relu_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_GK_Tuple,
@@ -98,7 +98,7 @@ void add_device_conv2d_xdl_bias_relu_perchannel_quantization_int8_instances(
 
 void add_device_conv2d_xdl_bias_tanh_perchannel_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_GK_Tuple,
@@ -123,7 +123,7 @@ template <ck::index_t NumDimSpatial,
           typename DsDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -137,7 +137,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Add_Activation_Mul2_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DsLayout,
@@ -193,7 +194,7 @@ template <ck::index_t NumDimSpatial,
           typename DsDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -207,7 +208,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Add_Mul2_Activation_Mul_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DsLayout,

library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_bias_forward_perlayer_quantization.hpp

@@ -7,7 +7,7 @@
 
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
-#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_abd.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
 
 #include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
@@ -20,7 +20,7 @@ namespace instance {
 // grouped conv2d forward, NHWGC/GKYXC/NHWGK
 void add_device_conv2d_dl_bias_perlayer_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -35,7 +35,7 @@ void add_device_conv2d_dl_bias_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_dl_bias_relu_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_Tuple,
@@ -50,7 +50,8 @@ void add_device_conv2d_dl_bias_relu_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_dl_bias_tanh_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -66,7 +67,7 @@ void add_device_conv2d_dl_bias_tanh_perlayer_quantization_int8_instances(
 #endif
 void add_device_conv2d_xdl_bias_perlayer_quantization_int8_instances(
     std::vector<
-        std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -81,7 +82,7 @@ void add_device_conv2d_xdl_bias_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_xdl_bias_relu_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_Tuple,
@@ -96,7 +97,8 @@ void add_device_conv2d_xdl_bias_relu_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_xdl_bias_tanh_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -121,7 +123,7 @@ template <ck::index_t NumDimSpatial,
           typename DsDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -135,7 +137,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Add_Activation_Mul_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DsLayout,
@@ -191,7 +194,7 @@ template <ck::index_t NumDimSpatial,
           typename DsDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -205,7 +208,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Add_Mul_Activation_Mul_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DsLayout,

library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perchannel_quantization.hpp

@@ -7,7 +7,7 @@
 
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
-#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_abd.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
 
 #include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
@@ -19,7 +19,8 @@ namespace instance {
 #ifdef DL_KERNELS
 // grouped conv2d forward, NHWGC/GKYXC/NHWGK
 void add_device_conv2d_dl_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -34,7 +35,7 @@ void add_device_conv2d_dl_perchannel_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_dl_relu_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_Tuple,
@@ -49,7 +50,8 @@ void add_device_conv2d_dl_relu_perchannel_quantization_int8_instances(
         instances);
 #endif
 void add_device_conv2d_xdl_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         GK_Tuple,
@@ -64,7 +66,7 @@ void add_device_conv2d_xdl_perchannel_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_xdl_relu_perchannel_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 GK_Tuple,
@@ -88,7 +90,7 @@ template <ck::index_t NumDimSpatial,
           typename DsDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -102,7 +104,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Activation_Mul2_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         GK_Tuple,

library/include/ck/library/tensor_operation_instance/gpu/quantization/grouped_convolution_forward_perlayer_quantization.hpp

@@ -7,7 +7,7 @@
 
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
-#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_d.hpp"
+#include "ck/tensor_operation/gpu/device/device_grouped_conv_fwd_multiple_abd.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
 
 #include "ck/library/tensor_operation_instance/device_operation_instance_factory.hpp"
@@ -19,7 +19,8 @@ namespace instance {
 #ifdef DL_KERNELS
 // grouped conv2d forward, NHWGC/GKYXC/NHWGK
 void add_device_conv2d_dl_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         Empty_Tuple,
@@ -34,7 +35,7 @@ void add_device_conv2d_dl_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_dl_relu_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 Empty_Tuple,
@@ -49,7 +50,8 @@ void add_device_conv2d_dl_relu_perlayer_quantization_int8_instances(
         instances);
 #endif
 void add_device_conv2d_xdl_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<
+        std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                         NHWGC,
                                                         GKYXC,
                                                         Empty_Tuple,
@@ -64,7 +66,7 @@ void add_device_conv2d_xdl_perlayer_quantization_int8_instances(
         instances);
 
 void add_device_conv2d_xdl_relu_perlayer_quantization_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 Empty_Tuple,
@@ -86,7 +88,7 @@ template <ck::index_t NumDimSpatial,
           typename WeiDataType,
           typename OutDataType,
           typename Activation>
-struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD<
+struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD<
     NumDimSpatial,
     InLayout,
     WeiLayout,
@@ -100,7 +102,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     Activation_Mul_Clamp<Activation>>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_bf16_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_bf16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<1,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<1,
                                                                 GNWC,
                                                                 GKXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f16_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<1,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<1,
                                                                 GNWC,
                                                                 GKXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f32_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_f32_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<1,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<1,
                                                                 GNWC,
                                                                 GKXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv1d_fwd/xdl/device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_int8_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv1d_fwd_xdl_gnwc_gkxc_gnwk_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<1,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<1,
                                                                 GNWC,
                                                                 GKXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f16_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f32_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv2d_fwd_dl_gnhwc_gkyxc_gnhwk_f32_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f16_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/dl/device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f32_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 
 void add_device_grouped_conv2d_fwd_dl_nhwgc_gkyxc_nhwgk_f32_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 NHWGC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1p0_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1p0_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1s1p0_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_1x1s1p0_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_oddc_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_f16_oddc_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1p0_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1p0_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1s1p0_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_1x1s1p0_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,

library/src/tensor_operation_instance/gpu/grouped_conv2d_fwd/wmma/device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_instance.cpp

@@ -10,7 +10,7 @@ namespace device {
 namespace instance {
 // Compilation parameters for in[g, n, hi, wi, c] * wei[g, k, y, x, c] = out[g, n, ho, wo, k]
 void add_device_grouped_conv2d_fwd_wmma_gnhwc_gkyxc_gnhwk_i8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<2,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<2,
                                                                 GNHWC,
                                                                 GKYXC,
                                                                 Empty_Tuple,