Xavier sometimes crash and then reboot

Hi
I have two Xavier, one is 16GB Xavier, other is 32GB Xavier. I put my code on sdcard. My code needs to convert 7 yolov3 models to tensorRT engines and use one pytorch classify model. I have checked the memory is enough and fans runs well.

I run my code in docker environment on 32GB Xavier. The machine sometimes will reboot. The same code runs in docker environment on 16GB Xavier. My code work well. I try to make the enviroment same on both machines. I flash os and use same docker images. I also changed the power supply, sdcard, and wall sockets. My 32GB Xavier is still have problem. Sometime works well, sometimes reboot.

This is my log file when I run my code in docker and then crash:
On Jetpack 4.4.1
log_shutdown.tar (14.2 MB) log_shutdown_3.tar (7.9 MB)

This is my log file when I run my code in OS(not in docker)
Jetpack4.4.1_crash2.tar (14.3 MB)

Jetpack4.4_OS_shutdown_allNight_from18:00.tar (14.8 MB)

Hi,
I borrow another 32GB Xaiver. I run my code on new Xavier all night successfully. I think my 32GB Xavier has some hardware problem.

HI, xwater8

We are sorry about this.
Please check our RMA procedures below:

Thanks.