aclrtReduceAsync

Product	Supported
Ascend 950PR/Ascend 950DT	√
Atlas A3 training products/Atlas A3 inference products	√
Atlas A2 training products/Atlas A2 inference products	√
Atlas 200I/500 A2 inference products	√
Atlas inference products	√
Atlas training products	√

Performs a Reduce operation, including SUM, MIN, and MAX. This is an asynchronous API.

aclError aclrtReduceAsync(void *dst, const void *src, uint64_t count, aclrtReduceKind kind, aclDataType type, aclrtStream stream, void *reserve)

Parameter	Input/Output	Description
dst	Input	Pointer to the destination memory address.
src	Input	Pointer to the source memory address.
count	Input	Size of the source memory, in bytes.
kind	Input	Operation type. For details about the type definition, see aclrtReduceKind.
type	Input	Data type. For details about the type definition, see aclDataType.
stream	Input	Stream for executing the Reduce operation task. For details about the type definition, see aclrtStream.
reserve	Input	Reserved. Currently, the value is fixed at NULL.

0 on success; otherwise, failure. For details, see aclError.

dst and src must be on the same device as stream.

For the Ascend 950PR/Ascend 950DT, the following data types are supported: int8, int16, int32, uint32, fp16, fp32, and bf16.

For the Atlas A3 training products/Atlas A3 inference products, the following data types are supported: int8, int16, int32, fp16, fp32, and bf16.

For the Atlas A2 training products/Atlas A2 inference products, the following data types are supported: int8, int16, int32, fp16, fp32, and bf16.

For the Atlas 200I/500 A2 inference products, the following data types are supported: int8, int16, int32, fp16, and fp32.

For the Atlas inference products, the following data types are supported: fp32, fp16, and int16.

For the Atlas training products, only the fp32 data type is supported.

Parent topic: Execution Control