In <[log in to unmask]> [log in to unmask] (Stan Sieler)
writes:
 
>Although the original author sounds like he/she was blowing smoke,
>some computers have actually been susceptible to hardware damage by
>loops...including, if memory serves:
>
>   - IBM Cadet and/or 401 ?
>      looped reading from same memory address could damage the core?
 
This is the only one I ever heard about (it had achieved "urban legend"
status by the time I became involved in computers).  Basically, old
"core" machines that executed a "halt" statement would physically
"halt" the processor, i.e. the current program counter register would
stop changing, but the hardware itself wouldn't stop.  The end result
is that the circuit responsible for fetching the current instruction
repeatedly accesses the same physical core element, eventually
overheating the element until it caught on fire.
 
Tom Emerson
[log in to unmask]