From: jasee on

"Dave {Reply Address In.Sig}" <noone$$@llondel.org> wrote in message
news:5ol2h7-99g.ln1(a)llondel.org...
> Simon Brooke wrote:
>
>>
>>
>> On Mon, 12 Jul 2010 11:24:23 +0100, jasee wrote:
>>
>>> Network
>>> cards don't wear out, of course hard disks do, but also cpu have a
>>> limited life
>>
>> Uhhh???!?
>>
>> I have CPUs in this house which date back to 1979, and are still working
>> just fine. In fact of the 19 pre-1986 machines in my collection, all
>> still work perfectly, including the one MS-DOS machine (an Apricot).
>>
>> Limited life? Well, probably. I suspect some of them will die before I
>> do. But not very limited!
>>
> I had a Pentium CPU that died. It can probably be dated by the speed
> (166MHz) and had been stressed due to the fact that it had fan failure to
> the extent that it had a scorch mark on the package. It worked for about a
> year after that incident though. I remember losing a Z80 once as well, but
> apart from that they've been pretty reliable. I've got an 8051 that's
> managed 130,000 hours of near-continuous use and is still going.

At one point (can't remember when) if you took the heat sink off the cpu
simply blew up, then I think Intel first fitted a thermal device so it
switched off then AMD followed suit.
Of course they all have MTBF which is temperature related and not a straight
line!


From: Simon Brooke on
On Thu, 15 Jul 2010 09:13:26 +0100, jasee wrote:

> "Dave {Reply Address In.Sig}" <noone$$@llondel.org> wrote in message
> news:5ol2h7-99g.ln1(a)llondel.org...
>> Simon Brooke wrote:
>>
>>
>>>
>>> On Mon, 12 Jul 2010 11:24:23 +0100, jasee wrote:
>>>
>>>> Network
>>>> cards don't wear out, of course hard disks do, but also cpu have a
>>>> limited life
>>>
>>> Uhhh???!?
>>>
>>> I have CPUs in this house which date back to 1979, and are still
>>> working just fine. In fact of the 19 pre-1986 machines in my
>>> collection, all still work perfectly, including the one MS-DOS machine
>>> (an Apricot).
>>>
>>> Limited life? Well, probably. I suspect some of them will die before I
>>> do. But not very limited!
>>>
>> I had a Pentium CPU that died. It can probably be dated by the speed
>> (166MHz) and had been stressed due to the fact that it had fan failure
>> to the extent that it had a scorch mark on the package. It worked for
>> about a year after that incident though. I remember losing a Z80 once
>> as well, but apart from that they've been pretty reliable. I've got an
>> 8051 that's managed 130,000 hours of near-continuous use and is still
>> going.
>
> At one point (can't remember when) if you took the heat sink off the cpu
> simply blew up, then I think Intel first fitted a thermal device so it
> switched off then AMD followed suit.
> Of course they all have MTBF which is temperature related and not a
> straight line!

True. But, they don't 'wear out' in anything remotely approaching the
timescale you implied.



--

;; Semper in faecibus sumus, sole profundam variat