Visible to Intel only — GUID: GUID-928AF07C-0BBF-4BDE-AEEE-9539BE367D7B
Convolution
Inner Product
Matrix Multiplication
RNN
Batch Normalization
Binary
Concat
General
Execution Arguments
Implementation Details
Implementation Limitations
Performance Tips
Example
Eltwise
Group Normalization
Layer Normalization
Local Response Normalization (LRN)
Pooling
PReLU
Resampling
Shuffle
Softmax
Sum
Reorder
Reduction
Abs
AbsBackward
Add
AvgPool
AvgPoolBackward
BatchNormForwardTraining
BatchNormInference
BatchNormTrainingBackward
BiasAdd
BiasAddBackward
Clamp
ClampBackward
Concat
Convolution
ConvolutionBackwardData
ConvolutionBackwardWeights
ConvTranspose
ConvTransposeBackwardData
ConvTransposeBackwardWeights
Dequantize
Divide
DynamicDequantize
DynamicQuantize
Elu
EluBackward
End
Exp
GELU
GELUBackward
HardSigmoid
HardSigmoidBackward
HardSwish
HardSwishBackward
Interpolate
InterpolateBackward
LayerNorm
LayerNormBackward
LeakyReLU
Log
LogSoftmax
LogSoftmaxBackward
MatMul
Maximum
MaxPool
MaxPoolBackward
Minimum
Mish
MishBackward
Multiply
Pow
PReLU
PReLUBackward
Quantize
Reciprocal
ReduceL1
ReduceL2
ReduceMax
ReduceMean
ReduceMin
ReduceProd
ReduceSum
ReLU
ReLUBackward
Reorder
Round
Select
Sigmoid
SigmoidBackward
SoftMax
SoftMaxBackward
SoftPlus
SoftPlusBackward
Sqrt
SqrtBackward
Square
SquaredDifference
StaticReshape
StaticTranspose
Subtract
Tanh
TanhBackward
TypeCast
Wildcard
enum dnnl_alg_kind_t
enum dnnl_normalization_flags_t
enum dnnl_primitive_kind_t
enum dnnl_prop_kind_t
enum dnnl_query_t
enum dnnl::normalization_flags
enum dnnl::query
struct dnnl_exec_arg_t
struct dnnl_primitive
struct dnnl_primitive_desc
struct dnnl::primitive
struct dnnl::primitive_desc
struct dnnl::primitive_desc_base
enum dnnl_rnn_direction_t
enum dnnl_rnn_flags_t
enum dnnl::rnn_direction
enum dnnl::rnn_flags
struct dnnl::augru_backward
struct dnnl::augru_forward
struct dnnl::gru_backward
struct dnnl::gru_forward
struct dnnl::lbr_augru_backward
struct dnnl::lbr_augru_forward
struct dnnl::lbr_gru_backward
struct dnnl::lbr_gru_forward
struct dnnl::lstm_backward
struct dnnl::lstm_forward
struct dnnl::rnn_primitive_desc_base
struct dnnl::vanilla_rnn_backward
struct dnnl::vanilla_rnn_forward
Visible to Intel only — GUID: GUID-928AF07C-0BBF-4BDE-AEEE-9539BE367D7B
Concat
General
The concat primitive concatenates 
tensors over concat_dimension (here designated 
) and is defined as (the variable names follow the standard Naming Conventions):
where 
.
The concat primitive does not have a notion of forward or backward propagation. The backward propagation for the concatenation operation is simply an identity operation.
Execution Arguments
When executed, the inputs and outputs should be mapped to an execution argument index as specified by the following table.
Primitive input/output |
Execution argument index |
|---|---|
|
DNNL_ARG_MULTIPLE_SRC |
|
DNNL_ARG_DST |
Implementation Details
General Notes
The
memory format can be either specified by a user or derived by the primitive. The recommended way is to allow the primitive to choose the most appropriate format.The concat primitive requires all source and destination tensors to have the same shape except for the concat_dimension. The destination dimension for the concat_dimension must be equal to the sum of the concat_dimension dimensions of the sources (i.e.
). Implicit broadcasting is not supported.
Data Types Support
The concat primitive supports arbitrary data types for source and destination tensors according to the Data Types page. However, it is required that all source tensors are of the same data type (but not necessarily matching the data type of the destination tensor).
Data Representation
The concat primitive works with arbitrary data tensors. There is no special meaning associated with any logical dimensions.
Post-Ops and Attributes
Type |
Operation |
Description |
Res |
|---|---|---|---|
Attribute |
Scales the corresponding input tensor by the given scale factor(s). |
Only one scale per tensor is supported. Input tensors only. |
Implementation Limitations
The primitive works with several memory formats, such as plain formats dnnl_nchw, dnnl_nhwc, and blocked formats dnnl_nChw16c, dnnl_nCdhw8c that appear in convolutions. The primitive does not support non-blocked formats that are typically used in prepacked weights, such as:
Winograd format dnnl_format_kind_opaque,
RNN format dnnl_format_kind_opaque, or
In some cases for blocked format with attached compensation that is used in s8s8 convolutions (see Nuances of int8 Computations).
Refer to Data Types for limitations related to data types support.
GPU
Only tensors of 6 or fewer dimensions are supported.
Performance Tips
Whenever possible, avoid specifying the destination memory format so that the primitive is able to choose the most appropriate one.
The concat primitive is highly optimized for the cases in which all source tensors have same memory format and data type matches the destination tensor data type. For other cases, more general but slower code is working. Consider reordering sources to the same data format before using the concat primitive.
Example
This C++ API example demonstrates how to create and execute a Concat primitive.
Key optimizations included in this example:
Identical source (src) memory formats.
Creation of optimized memory format for destination (dst) from the primitive descriptor