Does a processor's performance degrade over time?

[Mark Oueis]

I was wondering if a CPU's performance would degrade over time and
use. Would it lose clock speed? Why would it do that?

Mark

[Robert Redelmeier]

Mark Oueis <markoueis@hotmail.com> wrote: I was wondering if a CPU's performance would degrade over time and use. Would it lose clock speed? Why would it do that?

CPUs don't degrade over time. Generally it works or it
doesn't. Partial failures are rare. Potentially an Intel
P4 chip might throttle if the HSF is too small/dirty.

Older MS-Windows machines frequently are slower than when new due
to bigger registries, increasing disk fragmentation, additional
process load, etc. But that is an OS maintenance issue.

-- Robert

[Greg]

Robert Redelmeier wrote: Mark Oueis <markoueis@hotmail.com> wrote: I was wondering if a CPU's performance would degrade over time and use. Would it lose clock speed? Why would it do that? CPUs don't degrade over time. Generally it works or it doesn't. Partial failures are rare. Potentially an Intel P4 chip might throttle if the HSF is too small/dirty. Older MS-Windows machines frequently are slower than when new due to bigger registries, increasing disk fragmentation, additional process load, etc. But that is an OS maintenance issue. -- Robert

There are known transistor degradation mechanisms such as gate-oxide
breakdown and hot-electron effects that slowly change transistor
performance. This can slowly degrade timings for signals across chip
and eventually cause the chip to not work. Or it can result in a bit
flip, if it happens in the chip's cache.

Ultimately, the chip will slowly become less *capable*. That is,
instead of a 3.4G chip being able to run at 3650Mhz it will slowly be
able to run at 3600, then 3550, then 3500, etc... The end user
wouldn't notice since the chip is multiplier locked at 3.4G until it is
becomes unable to run at that speed and then it would act unstably.

There are Q&R models that are used to predict failure rates. Chips are
stressed at high temperature and voltage to simulate years of usage.
This is why Intel offers only a 7 year warrantee on their chips.

-Greg

[Edmund]

On 9 Feb 2005 17:22:09 -0800, "Greg" <jgmillr1@yahoo.com> wrote:

re are Q&R models that are used to predict failure rates. Chips arestressed at high temperature and voltage to simulate years of usage.This is why Intel offers only a 7 year warrantee on their chips.-Greg

7?

http://support.intel.com/support/processors/sb/CS-020033.htm#2
Boxed Processor family
Intel offers a three (3) year limited warranty to original purchasers of
Intel boxed processors.

Ed

[Tony Hill]

On 9 Feb 2005 10:12:51 -0800, markoueis@hotmail.com (Mark Oueis)
wrote:
I was wondering if a CPU's performance would degrade over time anduse.

In a word, no. Or at least not if it's operating at all properly.
There are some situations where it might be slower, one that jumps to
mind is if the cooling mechanism on a P4 becomes severely degraded
(heatsink clogged with excessive dust or a fan that just isn't working
quite right),then the processor might start to overheat and slow
itself down to prevent damage. However this isn't the processor
"degrading" in any way, it's you're cooling solution that's degraded
and your processor is trying to compensate. Of course, such
situations are VERY rare, usually if you're cooling setup isn't
working right then you're computer will just crash.
Would it lose clock speed?

No
Why would it do that?

It wouldn't, though it might FEEL slower for a variety of reasons,
some technical and some more psychological. On the technical side,
the big problem is that processors have to do more with newer software
to implement the new features.

The psychological side is probably MUCH more important here though,
but I'm not sure that I'm the most qualified person to talk about it!


Of course, BY FAR the #1 reason why a computer will feel slower these
days is spyware. CHECK YOUR SYSTEM FOR SPYWARE! Those stupid flashy
adds that say "9 out of 10 systems are infected with spyware and
adware" are NOT bullshitting (err, at least not on that part, they are
bullshitting when they try to sell you some crap), the problem really
is that severe.

-------------
Tony Hill
hilla <underscore> 20 <at> yahoo <dot> ca

[Johannes H Andersen]

Mark Oueis wrote: I was wondering if a CPU's performance would degrade over time and use. Would it lose clock speed? Why would it do that? Mark

I probably feels that way, but it's usually due to disk fragmentation,
Windows register bloat and countless software installations that
automatically hook memory resources at boot time.