ReinterpretCast

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	√
Atlas training product	√

Function Usage

Reinterprets the current tensor as a new type specified by the user. The converted tensor has the same address and content as the original tensor, and the tensor size (number of bytes) remains unchanged.

Prototype

template <typename CAST_T> 
__aicore__ inline LocalTensor<CAST_T> ReinterpretCast() const

Parameters

**Table 1** Template parameters
Parameter	Description
CAST_T	New type specified by the user.

Returns

Returns the Tensor after reinterpretation.

Restrictions

None

Examples

// Example
// input_local is of the int32_t type and contains 16 elements (64 bytes).
for (int32_t i = 0; i < 16; ++i) {
    inputLocal.SetValue(i, i); // Assign i to the ith position in inputLocal.
}


// Call ReinterpretCast to reinterpret input_local to the int16_t type.
AscendC::LocalTensor<int16_t> interpreTensor = inputLocal.ReinterpretCast<int16_t>();
// The result of this example is as follows. The data of the two is the same and the same address is used in the physical memory. The data is reinterpreted based on different types.
// inputLocal:0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
// interpreTensor:0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 0 9 0 10 0 11 0 12 0 13 0 14 0 15 0

Parent topic: LocalTensor