Axpy

Applicability

Product	Supported
Atlas 350 Accelerator Card	√
Atlas A3 training product / Atlas A3 inference product	√
Atlas A2 training product / Atlas A2 inference product	√
Atlas 200I/500 A2 inference product	√
Atlas inference product AI Core	√
Atlas inference product Vector Core	x
Atlas training product	√

Function Usage

Adds the product of each element of the source operand (src) and a scalar to the corresponding element in the destination operand (dst). The formula is as follows.

$\text{[math]}$

Prototype

Computation of the first n pieces of data of a tensor

        
             template <typename T, typename U>
__aicore__ inline void Axpy(const LocalTensor<T>& dst, const LocalTensor<U>& src, const U& scalarValue, const int32_t& count)

High-dimensional tensor sharding computation

Bitwise mask mode

          
               template <typename T, typename U, bool isSetMask = true>
__aicore__ inline void Axpy(const LocalTensor<T>& dst, const LocalTensor<U>& src, const U& scalarValue, uint64_t mask[], const uint8_t repeatTime, const UnaryRepeatParams& repeatParams)

Contiguous mask mode

          
               template <typename T, typename U, bool isSetMask = true>
__aicore__ inline void Axpy(const LocalTensor<T>& dst, const LocalTensor<U>& src, const U& scalarValue, uint64_t mask, const uint8_t repeatTime, const UnaryRepeatParams& repeatParams)

Parameters

**Table 1** Template parameters
Parameter	Description
T	Data type of the destination operand. For details about data type restrictions of the destination and source operands, see Table 3. For the Atlas training product , the supported data types are half and float. For the Atlas inference product AI Core, the supported data types are half and float. For the Atlas A2 training product / Atlas A2 inference product , the supported data types are half and float. For the Atlas A3 training product / Atlas A3 inference product , the supported data types are half and float. For the Atlas 200I/500 A2 inference product , the supported data types are half and float. For the Atlas 350 Accelerator Card, the supported data types are half, float, bfloat16_t, uint64_t, and int64_t.
U	Data type of the source operand. For the Atlas training product , the supported data types are half and float. For the Atlas inference product AI Core, the supported data types are half and float. For the Atlas A2 training product / Atlas A2 inference product , the supported data types are half and float. For the Atlas A3 training product / Atlas A3 inference product , the supported data types are half and float. For the Atlas 200I/500 A2 inference product , the supported data types are half and float. For the Atlas 350 Accelerator Card, the supported data types are half, float, bfloat16_t, uint64_t, and int64_t.
isSetMask	Indicates whether to set mask inside the API. true: sets mask inside the API. false: sets mask outside the API. Developers need to use the SetVectorMask API to set the mask value. In this mode, the mask value in the input parameter of this API must be set to the placeholder MASK_PLACEHOLDER.

**Table 2** Parameters
Parameter	Input/Output	Meaning
dst	Output	Destination operand. The type is LocalTensor, and TPosition can be VECIN, VECCALC, or VECOUT. The start address of LocalTensor must be 32-byte aligned.
src	Input	Source operand. The type is LocalTensor, and TPosition can be VECIN, VECCALC, or VECOUT. The start address of LocalTensor must be 32-byte aligned.
scalarValue	Input	Source operand (scalar). The data type of scalarValue must be the same as that of src.
count	Input	Number of elements involved in the computation.
mask/mask[]	Input	mask controls the elements that participate in computation in each iteration. Bitwise mode: controls the elements that participate in computation by bit. If a bit is set to 1, the corresponding element participates in the computation. If a bit is set to 0, the corresponding element is masked from the computation. The mask value is an array. The array length and the value range of the array elements are related to the operand data type. When the operand is 16-bit, the array length is 2, with mask[0] and mask[1] each in the range [0, 2⁶⁴ – 1], and they cannot both be 0 at the same time. When the operand is 32-bit, the array length is 1, with mask[0] in the range (0, 2⁶⁴ – 1]. When the operand is 64-bit, the array length is 1, with mask[0] in the range (0, 2³² – 1]. For example, if mask = [0, 8] and 8 = 0b1000, only the fourth element participates in computation. Contiguous mode: indicates the number of contiguous elements that participate in computation. The value range is related to the operand data type. The maximum number of elements that can be processed in each iteration varies according to the data type. When the operand is 16-bit, mask ∈ [1, 128]. When the operand is 32-bit, mask ∈ [1, 64]. When the operand is 64-bit, mask ∈ [1, 32].
repeatTime	Input	Number of iteration repeats. The Vector Unit reads 256 bytes of contiguous data for computation each time. To read the complete data for processing, the unit needs to read the input data in multiple repeats. repeatTime indicates the number of iterations. For details about this parameter, see High-dimensional Sharding APIs.
repeatParams	Input	Parameters that control the operand address strides. They are of the UnaryRepeatParams type (see UnaryRepeatParams), and contain parameters such as the address stride of the operand for the same Data Block between adjacent iterations and the address stride of the operand between different Data Blocks in a single iteration. For details about the address stride of the operand between adjacent iterations, see repeatStride. For details about the address stride of the operand between different data blocks in a single iteration, see dataBlockStride.

**Table 3** Data type restrictions
src Data Type	scalar Data Type	dst Data Type	PAR	Availability
half	half	half	128	Atlas training product Atlas A2 training product / Atlas A2 inference product Atlas A3 training product / Atlas A3 inference product Atlas inference product AI Core Atlas 200I/500 A2 inference product Atlas 350 Accelerator Card
float	float	float	64	Atlas training product Atlas A2 training product / Atlas A2 inference product Atlas A3 training product / Atlas A3 inference product Atlas inference product AI Core Atlas 200I/500 A2 inference product Atlas 350 Accelerator Card
half	half	float	64	Atlas training product Atlas A2 training product / Atlas A2 inference product Atlas A3 training product / Atlas A3 inference product Atlas inference product AI Core Atlas 200I/500 A2 inference product Atlas 350 Accelerator Card
int64_t	int64_t	int64_t	64	Atlas 350 Accelerator Card
uint64_t	uint64_t	uint64_t	64	Atlas 350 Accelerator Card
bfloat16_t	bfloat16_t	bfloat16_t	128	Atlas 350 Accelerator Card

Returns

None

Restrictions

For details about the operand address alignment requirements, see General Address Alignment Restrictions.
For details about the constraints on operand address overlapping, see General Address Overlapping Restrictions.

For high-dimensional tensor sharding computation APIs, if the data type of src and the scalar is half and that of dst is float, the number of source operand elements in an iteration must be the same as that of the destination operand. Therefore, the first four data blocks are selected for computation in each iteration. This restriction needs to be considered when Repeat Stride, mask, and address overlapping are configured.
For the Atlas 350 Accelerator Card, uint64_t and int64_t support only the APIs that compute the first n pieces of data in a tensor.

Examples

These examples show only part of the code used in the computation.

Example of high-dimensional tensor sharding computation (contiguous mask mode)

        
             // repeatTime = 4, mask = 128, 128 elements one repeat, 512 elements total
// srcLocal, scalar, and dstLocal are all of half type.
// dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
// dstRepStride, srcRepStride = 8, no gap between repeats 
AscendC::Axpy(dstLocal, srcLocal, (half)2.0, 128, 4,{ 1, 1, 8, 8 });

// srcLocal and scalar are of half type. dstLocal is of float type.
// repeatTime = 8, mask = 64, 64 elements one repeat, 512 elements total
// dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
// dstRepStride = 8, srcRepStride = 4, no gap between repeats 
AscendC::Axpy(dstLocal, srcLocal, (half)2.0, 64, 8,{ 1, 1, 8, 4 }); // Select the first four data blocks of the source operand for computation in each iteration.

Example of high-dimensional tensor sharding computation (bitwise mask mode)

        
             uint64_t mask[2] = { 0xFFFFFFFFFFFFFFFF, 0xFFFFFFFFFFFFFFFF };
// repeatTime = 4, 128 elements one repeat, 512 elements total, half type
// dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
// dstRepStride, srcRepStride = 8, no gap between repeats
AscendC::Axpy(dstLocal, srcLocal, (half)2.0, mask, 4,{ 1, 1, 8, 8 });

Example of computing the first n pieces of data of a tensor

        
             AscendC::Axpy(dstLocal, src0Local, (half)2.0, 512);// half type

Result example:

Input (src0Local):
[1. 2. 3. 4. 5. 6. ... 512.]
Input (scalarValue): 2.0
Intermediate output (dstLocal):
[0. 0. 0. 0. 0. 0. ... 0.]
Final output (dstLocal) after Axpy is complete:
[2. 4. 6. 8. 10. 12. ... 1024.]

Parent topic: Compound Computation