Hi @karaca,
Are you using the internal DCDC converter to supply the power domains of the RT1172, or are you using external supplies? Where and how are you measuring the current consumptions? Are you not able to debug the boards with the current design you have? What about using UART for SDP instead of USB? Have you seen a specific link with the affected MCUs? Perhaps they are from the same batch? Or maybe a pattern like the faulty boards were the first 30 boards to be assembled, or last 30?
This issue could be due to a number of things, perhaps even a combination of smaller issues. We would require further information in order to better pinpoint the problem, as it currently still has many broad possibilities. Any more information, no matter how small, about the corrupt boards would be very helpful.
BR,
Edwin.