GetTensorC

Applicability

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

Function Usage

After the Iterate operation is complete, this API is called to obtain the result matrix block and transfer data from the L0C to the GM. This API is used together with the Iterate API to obtain the result matrix after the Iterate operation is complete.

Prototype

template <bool sync = true>
__aicore__ inline void GetTensorC(const AscendC::GlobalTensor<DstT> &output, uint8_t enAtomic = 0, bool enSequentialWrite = false)

Parameters

**Table 1** Template parameters
Parameter	Description
sync	Reserved parameter, which is imperceptible to users.

**Table 2** API parameters
Parameter	Input/Output	Description
output	Input	Moves the computation result to the GM address of the global memory.
enAtomic	Input	Reserved parameter, which is imperceptible to users.
enSequentialWrite	Input	Reserved parameter, which is imperceptible to users.

Returns

None

Restrictions

The GetTensorC API must be called after the Iterate API to complete the convolution backpropagation. The calling sequence is as follows:

while (Iterate()) {   
    GetTensorC(); 
}

Examples

while (gradInput_.Iterate()) {   
    gradInput_.GetTensorC(gradInputGm_[offsetC_]); 
}

Parent topic: Conv3DBackpropInput Kernel APIs