I got all xxxxxxx on hash board today. Rebooted and came online fine. At the time hw errors we're in the 70k on all 3 boards. Maybe it simply hit temp limit and shutdown? Now the board that shutdown is considerably lower in hw errors. My l3+ runs for weeks without a single hw error. I get the feeling bitmain rushed these and overclock them.
If you underclock to 575 your HW errors will practically vanish.
Very helpful! Thank you.