Do Processors "wear out"?

[QuietIce]

I have two Storms on the same loop, one installed in June '06, another installed about a year later, and both using AS5. When I first started experiencing (what I think of as) over-heating problems I looked at CoreTemp and found my older Opty's temps had gone up 3-4°C. At the same time the newer installation had hardly moved, maybe 1-2°C above normal. During the last 6 months or so that spread has gotten wider.

So to answer your question - yes, I think AS5 has a breakdown point. But keep in mind, everything was fine for almost 3 years, 0 issues, 0 change in temps. Since most people rebuild their rigs more often than that it's just not an issue as far as I'm concerned. I'll continue to use AS5 because I now know it IS good for 3 years ...

 

[Careface]

Ice, do you really believe your AS5 goes bad over time?

It definitely does - I've been using this current application for ~1.5years now, and I took the HSF off the other day for a laugh, and it was all.. hard to explain. It had "veins" where there was no AS5, like it had dried up (obviously it hadn't). Also, its not due to a misapplication of AS5 either.

On a more related note, I've had this chip for nearly 5 years now, and it's been oc'd and under full load since the day I got it, on 24/7 except for exceptional circumstances, and I've had the same issues as you. Diminishing ocs, general instability.. etc.

 

[Angry]

Really old tubes of AS tend to separate as well. Ive got some thats on it way out I think, its became super ultra gooey....I let a friend use some and when he went to pull the heatsink off, it pulled his AM2 cpu out of its socket.

I need to pick up another tube of it....but this stuff still works, its just super "gooey".

 

[AMDGuy]

I downloaded and installed OCCT. Right off the bat I noticed that the 3.3V rail is reporting at 1.52V. I assume that's a problem?

I don't see any bulging caps.

So an anyone answer this question? What voltage should I be seeing on the 3.3 rail in OCCT?

 

[AMDGuy]

So an anyone answer this question? What voltage should I be seeing on the 3.3 rail in OCCT?

Scratch that. Looks like OCCT is seeing the chipset voltage as 3.3V. I used Epox Thunderprobe to double check the voltages and all look OK except for the 5VSB, but would that be important since it's only for standby?

+3.3V is showing 3.3V
Chipset is showing 1.5V
+5V shows 5.0V
Vdimm shows 2.6V
Vbat shows 3.1V
5VSB shows 4.3V

I took the power supply out and opened it up to check the caps, etc. I didn't find any bulging caps and I blew it out with some compressed air to clean it up good.

I've upped the DIMM voltage on the crucial memory to 2.7V rather than 2.6. I noticed it fluctuating from 2.52 to 2.6 in the bios.

I've set the system back to 2200Mhz. I'll run OCCT for an hour and see what happens. If all goes well I'll jump it up to 2400Mhz.

 

[QuietIce]

Isn't that what keeps the CMOS active when the computer is off (without using the battery)?
Or am I way out of date here ...?

 

[Niku-Sama]

i thought 3.3 did that now

 

[AMDGuy]

This motherboard is an Epox EP9NPA+ Ultra (4 years old) so the 5VSB probably is the standby not 3.3V.

Anyway, something's definitely going on. I'm not sure if the processor has had it or the PSU is going out, but either way the system won't even POST at 2400Mhz anymore. I tried 2300, no POST as well.

This thing used to would run along without a tick at 2600Mhz.

I've got it set to 2250 right now with a +.75 Vcore and running OCCT again to see what happens.

All things point to the CPU being the problem, but if it's not getting good clean voltage I guess the PSU could be the culprit. However, all the monitoring software I've run doesn't show any obvious PSU issues.

OH, one other very ODD thing. If I power the system off. When I hit the power button to turn it back on, it takes about 30 seconds or so before the system shows any signs of having been powered on. No lights not nothing, then all of a sudden boom, it starts up. That's the only indication I've found of a PSU problem. Anyone else experienced something like that to have input?



I may just have to live with it at a slower speed until I build the new system.

 

[QuietIce]

That power-on issue makes it sound more and more like the PSU is going bad. Just hope it doesn't take out the system before you get the new rig built.

Any chance you can buy the new PSU now???

i thought 3.3 did that now

 

[Niku-Sama]

This motherboard is an Epox EP9NPA+ Ultra (4 years old) so the 5VSB probably is the standby not 3.3V.

uhh....heh i we are talking about things way older than 4 years

 

[AMDGuy]

That power-on issue makes it sound more and more like the PSU is going bad. Just hope it doesn't take out the system before you get the new rig built.

Any chance you can buy the new PSU now???

Yeah, I can get a new PSU now, just didn't wanna spend the $100 on one and have the same issues.

 

[QuietIce]

Not just A PSU - buy whatever PSU you'll be using in the new rig. Then it won't matter if the PSU doesn't solve the problem, you'll still have it for the new build ...

 

[AMDGuy]

Not just A PSU - buy whatever PSU you'll be using in the new rig. Then it won't matter if the PSU doesn't solve the problem, you'll still have it for the new build ...

Quick update...on Black Friday I went ahead and ordered new components for an i5 system, and as part of that got the Corsair 650TX PSU.

Put it in my old system last night and it solved my problems.

Bottom line, the Antec PSU was either over taxed, bad, or both. Now I can take my time getting my data off the old system before moving the new one.

 

[dgb303]

I seemed to have a similar problem in my old 939 system also. I couldn't ever seem to get my Athlon XP over 2.4Ghz without problems. It's aged, so I used the opportunity to justify a new i5 system.