Uniqueness of HWID on nRF5340

[neo_here]

Hi, we have a product in production and we are using the HWID which is accessible via this API

https://docs.nordicsemi.com/bundle/ncs-latest/page/nrf/libraries/others/hw_id.html


We have seen couple instances of duplicate HWIDs upto this point and wanted to know if these HWID can repeat or how unique are they? 

If they are not totally unique is there any other information that can be accessed which is unique to every silicon of the nRF5340 ? 

[Einar Thorsrud]

Hi,

The hardware ID comes from the DEVICEID in FICR, and is 64 bit that is randomly generated in prodcution. It is not guaranteed to be unique, but given the size of the number you should normally never see collisions. Do you use the full 64 bit when you see collisions?

[neo_here]

Hey yes, we use the the HW_ID_LEN provided by hw_id.h

We're also trying to figure out if it might be process error instead as well, but wanted to know about the chances of seeing collisions.

[Emil Lenngren]

Do you mean 16 bytes?

[neo_here]

Yes, just caught the typo

[Einar Thorsrud]

Which HW ID are you using? Have you set it explicitly yourself, or can you check the .config? If you are using HW_ID_LIBRARY_SOURCE_DEVICE_ID which should be default, can you double check that the value you get match wat you get from the DEVICEID registers? (note that the library gives you a string representation, and not the raw value).