Function: Get_device_utilization_rate

aclError aclrtGetDeviceUtilizationRate(int32_t deviceId, aclrtUtilizationInfo *utilizationInfo)

utilization_info, ret = acl.rt.get_device_utilization_rate (device_id)

Queries the usage of the Cube, Vector, and AI CPU on the device.

device_id: int, device ID.

Must be in the range of [0, Device count – 1]. Call acl.rt.get_device_count to obtain the device count.

utilization_info: dict, usage information dictionary of the aclrtUtilizationInfo type. For details, see aclrtUtilizationInfo.

ret: int, error code.

• 0 indicates success.
• Other values indicate failure.

• For Ascend virtual instances, this API cannot be called to query the usage, and the returned value has no meaning.
• When profiling is enabled, this API cannot be called to query the usage, and the returned value has no meaning.
• Querying the memory usage of a device is a reserved function and is not supported in the current version. If this API is called to query the memory usage, the returned value is -1.
• There is no vector on the Atlas 200/300/500 Inference Product. When this API is called to query the vector usage, the returned value is -1.
• There is no vector on the Atlas Training Series Product. When this API is called to query the vector usage, the returned value is -1.