DML_ELEMENT_WISE_SUBTRACT_OPERATOR_DESC structure (directml.h)
Subtracts each element of BTensor from the corresponding element of ATensor, placing the result into the corresponding element of OutputTensor.
f(a, b) = a - b
This operator supports in-place execution, meaning that OutputTensor is permitted to alias one of the input tensors during binding.
struct DML_ELEMENT_WISE_SUBTRACT_OPERATOR_DESC {
const DML_TENSOR_DESC *ATensor;
const DML_TENSOR_DESC *BTensor;
const DML_TENSOR_DESC *OutputTensor;
};
ATensor
Type: const DML_TENSOR_DESC*
A tensor containing the left-hand side inputs.
BTensor
Type: const DML_TENSOR_DESC*
A tensor containing the right-hand side inputs.
OutputTensor
Type: const DML_TENSOR_DESC*
The output tensor to write the results to.
This operator was introduced in DML_FEATURE_LEVEL_1_0
.
ATensor, BTensor, and OutputTensor must have the same DataType, DimensionCount, and Sizes.
Tensor | Kind | Supported dimension counts | Supported data types |
---|---|---|---|
ATensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, INT16, INT8, UINT64, UINT32, UINT16, UINT8 |
BTensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, INT16, INT8, UINT64, UINT32, UINT16, UINT8 |
OutputTensor | Output | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, INT16, INT8, UINT64, UINT32, UINT16, UINT8 |
Tensor | Kind | Supported dimension counts | Supported data types |
---|---|---|---|
ATensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, UINT64, UINT32 |
BTensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, UINT64, UINT32 |
OutputTensor | Output | 1 to 8 | FLOAT32, FLOAT16, INT64, INT32, UINT64, UINT32 |
Tensor | Kind | Supported dimension counts | Supported data types |
---|---|---|---|
ATensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT32, UINT32 |
BTensor | Input | 1 to 8 | FLOAT32, FLOAT16, INT32, UINT32 |
OutputTensor | Output | 1 to 8 | FLOAT32, FLOAT16, INT32, UINT32 |
Tensor | Kind | Supported dimension counts | Supported data types |
---|---|---|---|
ATensor | Input | 4 | FLOAT32, FLOAT16, INT32, UINT32 |
BTensor | Input | 4 | FLOAT32, FLOAT16, INT32, UINT32 |
OutputTensor | Output | 4 | FLOAT32, FLOAT16, INT32, UINT32 |
Tensor | Kind | Supported dimension counts | Supported data types |
---|---|---|---|
ATensor | Input | 4 | FLOAT32, FLOAT16 |
BTensor | Input | 4 | FLOAT32, FLOAT16 |
OutputTensor | Output | 4 | FLOAT32, FLOAT16 |
Requirement | Value |
---|---|
Header | directml.h |