Replace multiD with multiABD

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_bias_relu_perchannel_quantization_int8.cpp

@@ -2,7 +2,7 @@
 // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 using InDataType           = int8_t;
 using WeiDataType          = int8_t;
@@ -33,7 +33,7 @@ template <ck::index_t NDimSpatial,
           typename RequantScaleLayout,
           typename OutLayout>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_bias_relu_perlayer_quantization_int8.cpp

@@ -2,7 +2,7 @@
 // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 using InDataType       = int8_t;
 using WeiDataType      = int8_t;
@@ -31,7 +31,7 @@ template <ck::index_t NDimSpatial,
           typename BiasLayout,
           typename OutLayout>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_perchannel_quantization_int8.cpp

@@ -2,7 +2,7 @@
 // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 using InDataType           = int8_t;
 using WeiDataType          = int8_t;
@@ -31,7 +31,7 @@ template <ck::index_t NDimSpatial,
           typename RequantScaleLayout,
           typename OutLayout>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/40_conv2d_fwd_quantization/conv2d_fwd_xdl_perlayer_quantization_int8.cpp

@@ -2,7 +2,7 @@
 // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #include "common.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 using InDataType       = int8_t;
 using WeiDataType      = int8_t;
@@ -26,7 +26,7 @@ static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecializatio
 
 template <ck::index_t NDimSpatial, typename InLayout, typename WeiLayout, typename OutLayout>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/62_conv_fwd_activ/convnd_fwd_activ_common.hpp

@@ -11,7 +11,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 #include "ck/library/utility/algorithm.hpp"
 #include "ck/library/utility/check_err.hpp"
@@ -47,7 +47,7 @@ static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecializatio
 
 template <typename OutElementOp>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/62_conv_fwd_activ/convnd_fwd_xdl_scaleadd_scaleadd_relu_fp16.cpp

@@ -9,7 +9,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 #include "ck/library/utility/algorithm.hpp"
 #include "ck/library/utility/check_err.hpp"
@@ -47,7 +47,7 @@ static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecializatio
 
 template <typename OutElementOp>
 using DeviceGroupedConvNDFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

example/62_conv_fwd_activ/multi_AB/convnd_fwd_activ_multi_ab_common.hpp

@@ -9,7 +9,7 @@
 #include "ck/ck.hpp"
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
-#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_xdl_cshuffle.hpp"
+#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
 
 #include "ck/library/utility/algorithm.hpp"
 #include "ck/library/utility/check_err.hpp"
@@ -44,7 +44,7 @@ template <typename DataType,
           typename InElementOp,
           typename WeiElementOp>
 using DeviceGroupedConvNDMultiABFwdInstance =
-    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleD_Xdl_CShuffle<
+    ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
         NDimSpatial,
         InLayout,
         WeiLayout,

include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_dl_multiple_d_nhwc_kyxc_nhwk.hpp

@@ -15,7 +15,7 @@
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
 #include "ck/tensor_operation/operator_transform/transform_conv_fwd_to_gemm.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/gemm_specialization.hpp"
 #include "ck/tensor_operation/gpu/device/matrix_padder.hpp"
 #include "ck/tensor_operation/gpu/grid/gridwise_gemm_dl_multiple_d.hpp"
@@ -216,7 +216,7 @@ template <index_t NDimSpatial,
           index_t CThreadTransferSrcDstVectorDim,
           index_t CThreadTransferDstScalarPerVector>
 struct DeviceGroupedConvFwdDlMultipleD_NHWC_KYXC_NHWK
-    : public DeviceGroupedConvFwdMultipleD<NDimSpatial,
+    : public DeviceGroupedConvFwdMultipleABD<NDimSpatial,
                                              ALayout,
                                              BLayout,
                                              DsLayout,

include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_multiple_r_xdl_cshuffle.hpp

@@ -1090,7 +1090,7 @@ struct DeviceGroupedConvFwdMultipleDMultipleR_Xdl_CShuffle
         auto str = std::stringstream();
 
         // clang-format off
-        str << "DeviceGroupedConvFwdMultipleD_Xdl_CShuffle"
+        str << "DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle"
             << "<"
             << BlockSize << ", "
             << MPerBlock << ", "

include/ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_d_wmma_cshuffle.hpp

@@ -15,7 +15,7 @@
 #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
 #include "ck/tensor_operation/gpu/device/convolution_forward_specialization.hpp"
 #include "ck/tensor_operation/operator_transform/transform_conv_fwd_to_gemm.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/gemm_specialization.hpp"
 #include "ck/tensor_operation/gpu/device/matrix_padder.hpp"
 #include "ck/tensor_operation/gpu/grid/gridwise_gemm_multiple_d_wmma_cshuffle.hpp"
@@ -92,7 +92,7 @@ template <index_t NDimSpatial,
           LoopScheduler LoopSched         = make_default_loop_scheduler(),
           ck::PipelineVersion PipelineVer = ck::PipelineVersion::v1>
 struct DeviceGroupedConvFwdMultipleD_Wmma_CShuffle
-    : public DeviceGroupedConvFwdMultipleD<NDimSpatial,
+    : public DeviceGroupedConvFwdMultipleABD<NDimSpatial,
                                              ALayout,
                                              BLayout,
                                              DsLayout,

library/include/ck/library/tensor_operation_instance/gpu/grouped_convolution_forward_scaleadd_ab.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 ScaleAdd    = ck::tensor_operation::element_wise::ScaleAdd;
 #ifdef CK_ENABLE_BF16
 // grouped conv3d forward multi AB scaleadd, NDHWGC/GKZYXC/NDHWGK
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_bf16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<>,
@@ -40,7 +40,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_bf16_ins
 
 #ifdef CK_ENABLE_FP16
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_f16_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<>,
@@ -56,7 +56,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_f16_inst
 
 #ifdef CK_ENABLE_FP32
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_f32_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<>,
@@ -72,7 +72,7 @@ void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_f32_inst
 
 #ifdef CK_ENABLE_INT8
 void add_device_grouped_conv3d_fwd_xdl_scaleadd_ab_ndhwgc_gkzyxc_ndhwgk_int8_instances(
-    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleD<3,
+    std::vector<std::unique_ptr<DeviceGroupedConvFwdMultipleABD<3,
                                                                 NDHWGC,
                                                                 GKZYXC,
                                                                 ck::Tuple<>,
@@ -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,
@@ -111,7 +111,8 @@ struct DeviceOperationInstanceFactory<ck::tensor_operation::device::DeviceGroupe
     ck::tensor_operation::element_wise::PassThrough,
     ComputeType>>
 {
-    using DeviceOp = DeviceGroupedConvFwdMultipleD<NumDimSpatial,
+    using DeviceOp =
+        DeviceGroupedConvFwdMultipleABD<NumDimSpatial,
                                         InLayout,
                                         WeiLayout,
                                         DLayouts,

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,