mkldnn_memory_desc_t Struct Reference

#include <mkldnn_types.h>

Collaboration diagram for mkldnn_memory_desc_t:

Public Attributes
mkldnn_primitive_kind_t	primitive_kind
int	ndims
mkldnn_dims_t	dims
mkldnn_data_type_t	data_type
mkldnn_memory_format_t	format
union {
mkldnn_blocking_desc_t   blocking
mkldnn_wino_desc_t   wino_desc
mkldnn_rnn_packed_desc_t   rnn_packed_desc
}	layout_desc

Detailed Description

Memory descriptor. The description is based on a number of dimensions, dimensions themselves, plus information about elements type and memory format. Additionally, contains format-specific descriptions of the data layout.

Member Data Documentation

mkldnn_blocking_desc_t mkldnn_memory_desc_t::blocking

Description of the data layout for memory formats that use blocking.

mkldnn_data_type_t mkldnn_memory_desc_t::data_type

Data type of the tensor elements.

mkldnn_dims_t mkldnn_memory_desc_t::dims

Dimensions in the following order:

• CNN data tensors: mini-batch, channel, spatial ({N, C, [[D,] H,] W})
• CNN weight tensors: group (optional), output channel, input channel, spatial ({[G,] O, I, [[D,] H,] W})
• RNN data tensors: time, mini-batch, channels ({T, N, C}) or layers, directions, states, mini-batch, channels ({L, D, S, N, C})
• RNN weight tensor: layers, directions, input channel, gates, output channels ({L, D, I, G, O}).

Note

The order of dimensions does not depend on the memory format, so whether the data is laid out in mkldnn_nchw or mkldnn_nhwc the dims for 4D CN data tensor would be {N, C, H, W}.

mkldnn_memory_format_t mkldnn_memory_desc_t::format

Memory format.

union { ... } mkldnn_memory_desc_t::layout_desc

int mkldnn_memory_desc_t::ndims

Number of dimensions

mkldnn_primitive_kind_t mkldnn_memory_desc_t::primitive_kind

The kind of primitive. Used for self-identifying the primitive descriptor. Must be mkldnn_memory.

mkldnn_rnn_packed_desc_t mkldnn_memory_desc_t::rnn_packed_desc

Tensor of packed weights for RNN.

mkldnn_wino_desc_t mkldnn_memory_desc_t::wino_desc

Tensor of weights for integer 8bit winograd convolution.

---

The documentation for this struct was generated from the following file:

• include/mkldnn/mkldnn_types.h