>Tuesday, September 22, 1998 8:41 PM Drew Dunn wrote:

>Re: [PCBUILD] Celeron Prosessor
>I hear that the 300 and 333MHz Celeron "A" overclocks quite a bit, to the
>tune of 400MHz.  I also hear that when the CPU fails, you have a nice
>plastic paperweight on your desk, along with the expense of buying a new
>chip.  Like I've said time and again, overclocking is just playing with
>fire.  The chip WILL fail.  Maybe not tomorrow, but probably before you're
>done with it.  While I don't have any scientific evidence to back it up,
>three years of anecdotal experience of *not* providing warranty replacements
>for fried, overclocked CPUs certainly tells me something.

Keeping this debate on the anecdotal experience level, I have never
had an overclocked cpu fail, simply because you notice that it gets
too unstable long before it suffers catastrophic physical damage. So,
I would assume there are two probable causes when an overclocked
cpu dies: It was left unattended for a very long time, so the overheating
was not witnessed, or the fan failed, maybe even while you were at lunch.
I knew of one to die while the guy was on vacation, because his
co-workers needed access to his files and evey time it locked up
they would power it down for an hour or so.  None of them realized
why that worked, they were just treating the symptom.  Three days
later, and after dozens of power on cycles, the cpu failed.  Turns out
the cpu fan was running at about half speed from worn bearings.
My rules are: install good cooling, regularly inspect the cooling and
keep it well maintained, and turn the PC off when you leave.  Of course
a regularly clocked cpu can fail under similar circumstances, so my
view is that the risk isn't significantly different for an overclocked cpu.
Tom Turak

                                  -----
       **Need help with PCBUILD mailing list? Send an Email to:**
        Bob Wright<[log in to unmask]> or Drew Dunn<[log in to unmask]>