As I tend to work with some Cisco 4506 and similiar switches - equipped with WSX45SUP6LE / WS-X45-SUP6L-E Supervisor Engines, I had the unfortunate delight to seem those rebooting at random.
Cisco did put out an entry in their bug tool ("Sup6LE reloads silently (Reset State: 00000201) / watchdog CISR0: 0x80") with the Bug Tracker ID CSCtf85481 - however, of the multiple accounts I had, I only found it on one of those. Other accounts did refuse my access to this file. As we confronted our Cisco Tech Support prior to finding the bug ourself, there was "no known error" which represented itself in rebooting the switch at random - so they said.
How do you find out it your board is one of those faulty ones?
a) If the serial number is JAE1422xxxx or higher (e.g., JAE1445xxxx, JAE1729xxx, etc) they board should not be affected. If it is in the range provided - check b)
b) If the "Hardware Revision" is 1.3 or higher, it is not affected. [sh idprom supervisor]
c) If either a) or b) are not true. Then the board may be affected by this bug. Contact TAC.
Another sign of the error is the "Last reload reason: Unknown reason" output of show version
It seems like the first batch of Supervisor 6L-E Engines were faulty on a HW level, so only a replacement will help. The engines will start rebooting once a while after a certain amount of time - with the time between each rebooting getting shorter.