Degrading overclock?

[Tomash]

OK, here's the deal.
I have an Opteron 165, DFI LandParty Ultra-D, 2x 1GB OCZ Platinium DDR500.
When I first bought the system, I OCed it to 2.7Ghz, and it was stable. For about few weeks.
Then, suddenly after one restard, the system wouldn't get past the second boot screen. It just loaded most of that page, then locked up, and rebooted after a few seconds. I've fixed it by lowering the OC to 2.6Ghz. Then, after another couple of weeks 2.6GHz became unusable, and now I'm at ~2.4Ghz.
What's causing it, and how do I fix it?
GOing from 2.7Ghz to 2.6 wasn't that big of a deal for me, but not I'd rather be back at 2.6Ghz (or ideally at 2.7) instead of current 2.4.
Help?

PS
Temps weren't and aren't a problem. I have a Scythe Ninja, and the CPU temp never gets above 40*C even after hours of gaming.

 

[Ricemarine]

what was the vcore at the time of 2.7 GHz?

 

[Fox5]

1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

 

[Tomash]

Originally posted by: Fox5
1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

1. The temperatures never (as in not a single time) rose above 40*C. After hours of gaming, they were around 35-37*C, with idle temps of 26*C.
2. When I had that 2.7GHz OC for a few week,s the Vcore was a 1.425v. I realize it was quite so, but it was stable. I could play games for hours without a single problem.

 

[Tomash]

Originally posted by: Ricemarine
what was the vcore at the time of 2.7 GHz?

1.425v according to the motherboard's settings and hardware monitor.

 

[Budman]

looks like you never stress tested your system (prime95) and now it's coming back to bite you in the azz.

seems like your system was never stable at 2.7ghz.

 

[Tomash]

Originally posted by: Budman
looks like you never stress tested your system (prime95) and now it's coming back to bite you in the azz.

seems like your system was never stable at 2.7ghz.

That is true. I've never ran a Prime95 stress test.
It was 100% stable in games, and it was enough for me.
Now, how do I get it back to the 2.7Ghz overclock?

 

[Amplifier]

Originally posted by: Tomash

Originally posted by: Budman
looks like you never stress tested your system (prime95) and now it's coming back to bite you in the azz.

seems like your system was never stable at 2.7ghz.

That is true. I've never ran a Prime95 stress test.
It was 100% stable in games, and it was enough for me.
Now, how do I get it back to the 2.7Ghz overclock?

i know running 2.7 is cool and all. but the dimishing returns on FPS start at around 2.2-2.4ghz depending on your GPU(s). the last time i reset my motherboard I didn't overclock for a week because I didn't notice much of a difference.

 

[dug777]

i'd suggest the weather is warming up for you crazy backwards yanks

I know my load temps change massively winter/summer here in Aus.

 

[Amaroque]

I had a Tbird that did that to me. It was a 1.33 @ 1.6. After about 6 months, I had to drop to 1.5. After about 8 months I had to drop to a stock 1.33. After 12 months, I couldn't even run it at 1.33, I had to drop down to 1.2.

Anyway, when it wouldn't run even at stock, I finally put it on a shelf and just bought a new board & CPU. Even though it was a retail CPU, with a three year warranty I didn't try to get a new one. (I did abuse it by OCing & overvolting). BTW, I run my systems 24/7 with 100% CPU load.

 

[F1shF4t]

Originally posted by: dug777
i'd suggest the weather is warming up for you crazy backwards yanks

I know my load temps change massively winter/summer here in Aus.

yea too true about the temps, in summer i got max of 52C at current settings (thats 35C ambient), right now it got cooler and its barely going above 40C

Buts its expected that a cpu will eventually start to degrade, its just for some reason ur one started early, or wasnt stable at those settings, and as soon as it started degrading a bit u started seeing a lot of errors.

 

[letdown427]

You weren't running a massive vDimm were you?

Have you tried putting the mem on a divider and going for 2.6 again? Could be the RAM...?

 

[JSFLY]

Jeez I hope that never happens to me!

 

[mtnd3vil]

I've noticed the same thing with my Intel 630 (3.0GHz). Brand new, with a 50mv bump I was able to hold 4.00GHz through every single version of 3dmark, sisoft sandra, and all the games I play. 4 months later I got several blue-screens and bumped it to 3.75 at the same voltage, that was good for a month and I started getting blue-screens again. Now I just keep it at 3.0/stock voltage.

I'm kind-of tired of overclocking. My next motherboard is going to be an intel mobo based on the 975 chipset. I don't even want the option to overclock anymore, it's not as advantagous as it used to be. The latest games don't seem to be CPU intensive at all.

Oblivion is the most hard-core video game money can buy and it holds my meager 3.0 at 50-80% during game play. so.....screw it, intel mobo all the way.

Austin

 

[m1ldslide1]

Originally posted by: Tomash

Originally posted by: Budman
looks like you never stress tested your system (prime95) and now it's coming back to bite you in the azz.

seems like your system was never stable at 2.7ghz.

That is true. I've never ran a Prime95 stress test.
It was 100% stable in games, and it was enough for me.
Now, how do I get it back to the 2.7Ghz overclock?

If you're looking for a stable overclock, you'll need to start the process over and prime95 it every step of the way (read zebo's overclocking guide here on anand). I think I replied to somebody yesterday with this, but I can boot over 3ghz just fine, but after trying to game for a few minutes the thing crashes. Therefore it isn't a stable overclock. I realize that you were able to game for a while at 2.7, but video games are NOT a stability test, and you can't expect to have a stable system just because you can play for a while at a certain OC.

/soapbox

 

[Amaroque]

Temps are only secondary. Voltage is the main killer of CPU's. This is why I don't overvolt my CPU's anymore. If I overvolt my CPU, I can boot windows at 3.1GHz. I can run games and such at 3.0GHz, but dual Prime95 will crash within 5 minutes.

Please read this thread. Very technical, and very informitave.

Originally posted by: pm
As to why 10% more voltage is much worse than 10% more temperature, well let's take the example of electromigration and look at it in detail.

Wires are made up of atoms all lined up. Electrons flow through these atoms. An electron is a very small thing, and atoms are a lot bigger. So the idea of an electron moving an atom around is a lot like someone trying to move a car (atom) by shooting a BB (electron) at it. Clearly to ever hope to move a car by shooting BB's at it, you would need a lot of BB's... a storm of BB's. But if you get enough, the car will move. Millions of BB's and that car will likely start getting pushed around. The temperature of the chip could be thought of as how slippery the road is. A little bit more slipperiness isn't going to help a BB move a car. It helps a little but not a lot. On the other hand, the voltage determines how many BB's you have, and it's not like 10% more voltage is 10% more BB's (electrons), you get a lot more than 10%. And worse than that, because increasing the voltage in a chip also increases the temperature (all things being equal, like same heatsink, same air temp, etc.), increasing the voltage is a double-whammy.

 

[Fox5]

Originally posted by: Tomash

Originally posted by: Fox5
1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

1. The temperatures never (as in not a single time) rose above 40*C. After hours of gaming, they were around 35-37*C, with idle temps of 26*C.
2. When I had that 2.7GHz OC for a few week,s the Vcore was a 1.425v. I realize it was quite so, but it was stable. I could play games for hours without a single problem.

1. When you overclock, the processor becomes more susceptible to high temperatures. Look at extreme LN2 overclocks where the processor will crash if it goes above 0 degrees.
2. It may have been stable, but that doesn't mean it didn't damage the processor. Overvolting can slowly cause the current to eat away at the silicon paths, until it eventually breaks through (even just electrically), limiting the overclockability of the chip or perhaps even rendering it inoperable. Especially with Athlon 64s/Opterons, since their integrated memory controller is extremely sensitive.
A system can be stable while the cpu is in the process of being damaged, but hasn't been damaged yet. You may never get your 2.7ghz overclock back without severe cooling or severely relaxed memory timings/clocks.

Or it could be option 1 again, and you need to reapply thermal paste or wait until the weather cools down. Once you overclock, the safe operating temperatures for a processor decrease.

Still, if you want to do safe overclocking, never overvolt (or at least not more than a 0.05V past the chip's stock operating voltage). Overvolting hurts a chip more than about anything else you can do to it. If you want to be really safe, don't exceed the maximum speed produced for the silicon you're using. (so I believe 2.4ghz for standard athlon 64s, 2.8ghz for the higher quality silicon)

 

[Tomash]

Originally posted by: letdown427
You weren't running a massive vDimm were you?

Have you tried putting the mem on a divider and going for 2.6 again? Could be the RAM...?

I had ram at around 200Mhz the whole time. It's DDR500, so that shouldn't be a problem either.

 

[Tomash]

Originally posted by: m1ldslide1

Originally posted by: Tomash

Originally posted by: Budman
looks like you never stress tested your system (prime95) and now it's coming back to bite you in the azz.

seems like your system was never stable at 2.7ghz.

That is true. I've never ran a Prime95 stress test.
It was 100% stable in games, and it was enough for me.
Now, how do I get it back to the 2.7Ghz overclock?

If you're looking for a stable overclock, you'll need to start the process over and prime95 it every step of the way (read zebo's overclocking guide here on anand). I think I replied to somebody yesterday with this, but I can boot over 3ghz just fine, but after trying to game for a few minutes the thing crashes. Therefore it isn't a stable overclock. I realize that you were able to game for a while at 2.7, but video games are NOT a stability test, and you can't expect to have a stable system just because you can play for a while at a certain OC.

/soapbox

Yeah, I think I;'ll have to reset everything and start from the start.
But, in its defense, the system was 100% stable, until it wasn't, LOL. I never crashed (other than in Oblivion until I sorted out mods incompatibility issues).

 

[Tomash]

Originally posted by: Fox5

Originally posted by: Tomash

Originally posted by: Fox5
1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

1. The temperatures never (as in not a single time) rose above 40*C. After hours of gaming, they were around 35-37*C, with idle temps of 26*C.
2. When I had that 2.7GHz OC for a few week,s the Vcore was a 1.425v. I realize it was quite so, but it was stable. I could play games for hours without a single problem.

1. When you overclock, the processor becomes more susceptible to high temperatures. Look at extreme LN2 overclocks where the processor will crash if it goes above 0 degrees.
2. It may have been stable, but that doesn't mean it didn't damage the processor. Overvolting can slowly cause the current to eat away at the silicon paths, until it eventually breaks through (even just electrically), limiting the overclockability of the chip or perhaps even rendering it inoperable. Especially with Athlon 64s/Opterons, since their integrated memory controller is extremely sensitive.
A system can be stable while the cpu is in the process of being damaged, but hasn't been damaged yet. You may never get your 2.7ghz overclock back without severe cooling or severely relaxed memory timings/clocks.

Or it could be option 1 again, and you need to reapply thermal paste or wait until the weather cools down. Once you overclock, the safe operating temperatures for a processor decrease.

Still, if you want to do safe overclocking, never overvolt (or at least not more than a 0.05V past the chip's stock operating voltage). Overvolting hurts a chip more than about anything else you can do to it. If you want to be really safe, don't exceed the maximum speed produced for the silicon you're using. (so I believe 2.4ghz for standard athlon 64s, 2.8ghz for the higher quality silicon)

That's the thing, man. I never overheated (below 40*C at all times), and I wasn't overvolting it either (1.425v then, 1.475v now).

 

[Fox5]

Originally posted by: Tomash

Originally posted by: Fox5

Originally posted by: Tomash

Originally posted by: Fox5
1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

1. The temperatures never (as in not a single time) rose above 40*C. After hours of gaming, they were around 35-37*C, with idle temps of 26*C.
2. When I had that 2.7GHz OC for a few week,s the Vcore was a 1.425v. I realize it was quite so, but it was stable. I could play games for hours without a single problem.

1. When you overclock, the processor becomes more susceptible to high temperatures. Look at extreme LN2 overclocks where the processor will crash if it goes above 0 degrees.
2. It may have been stable, but that doesn't mean it didn't damage the processor. Overvolting can slowly cause the current to eat away at the silicon paths, until it eventually breaks through (even just electrically), limiting the overclockability of the chip or perhaps even rendering it inoperable. Especially with Athlon 64s/Opterons, since their integrated memory controller is extremely sensitive.
A system can be stable while the cpu is in the process of being damaged, but hasn't been damaged yet. You may never get your 2.7ghz overclock back without severe cooling or severely relaxed memory timings/clocks.

Or it could be option 1 again, and you need to reapply thermal paste or wait until the weather cools down. Once you overclock, the safe operating temperatures for a processor decrease.

Still, if you want to do safe overclocking, never overvolt (or at least not more than a 0.05V past the chip's stock operating voltage). Overvolting hurts a chip more than about anything else you can do to it. If you want to be really safe, don't exceed the maximum speed produced for the silicon you're using. (so I believe 2.4ghz for standard athlon 64s, 2.8ghz for the higher quality silicon)

That's the thing, man. I never overheated (below 40*C at all times), and I wasn't overvolting it either (1.425v then, 1.475v now).

I'm not sure what the stock voltage on 90nm athlon 64s is. Isn't 1.4v the stock voltage, thus making 1.425 and 1.475 both overvolted?
And heat tolerances decrease when you overclock.

 

[Cooler]

Overclocking is not free and now you have to pay the piper.

 

[Tomash]

Originally posted by: Fox5

Originally posted by: Tomash

Originally posted by: Fox5

Originally posted by: Tomash

Originally posted by: Fox5
1. You were just on the verge of the temperatures your cpu could handle those speeds at, and degrading thermal paste and increasing spring heat affected it.
2. You damaged your processor with heat or raising the vcore too high (more likely the latter), and it wasn't able to sustain those speeds and now will never reach them again, at least not without serious cooling.

1. The temperatures never (as in not a single time) rose above 40*C. After hours of gaming, they were around 35-37*C, with idle temps of 26*C.
2. When I had that 2.7GHz OC for a few week,s the Vcore was a 1.425v. I realize it was quite so, but it was stable. I could play games for hours without a single problem.

1. When you overclock, the processor becomes more susceptible to high temperatures. Look at extreme LN2 overclocks where the processor will crash if it goes above 0 degrees.
2. It may have been stable, but that doesn't mean it didn't damage the processor. Overvolting can slowly cause the current to eat away at the silicon paths, until it eventually breaks through (even just electrically), limiting the overclockability of the chip or perhaps even rendering it inoperable. Especially with Athlon 64s/Opterons, since their integrated memory controller is extremely sensitive.
A system can be stable while the cpu is in the process of being damaged, but hasn't been damaged yet. You may never get your 2.7ghz overclock back without severe cooling or severely relaxed memory timings/clocks.

Or it could be option 1 again, and you need to reapply thermal paste or wait until the weather cools down. Once you overclock, the safe operating temperatures for a processor decrease.

Still, if you want to do safe overclocking, never overvolt (or at least not more than a 0.05V past the chip's stock operating voltage). Overvolting hurts a chip more than about anything else you can do to it. If you want to be really safe, don't exceed the maximum speed produced for the silicon you're using. (so I believe 2.4ghz for standard athlon 64s, 2.8ghz for the higher quality silicon)

That's the thing, man. I never overheated (below 40*C at all times), and I wasn't overvolting it either (1.425v then, 1.475v now).

I'm not sure what the stock voltage on 90nm athlon 64s is. Isn't 1.4v the stock voltage, thus making 1.425 and 1.475 both overvolted?
And heat tolerances decrease when you overclock.

1.4 is stock, I think.

 

[Duvie]

You guys may all be looking at the wrong issue...I have seen ocing take its effects on the mobo and it power regulations....I have seen this effect to place the cpu on a new mobo and have it go right back to similar OC...

I cannot see 1.425v set in mobo causing this....Temps seem reasonable as well....

I would look at PSU as well...they do degrade...what are the specs of the PSU???


You could do a complete strip down...rebuild...new paste...and rebuild part by part with new install and see if things go back to normal...

 

[Tomash]

Originally posted by: Duvie
You guys may all be looking at the wrong issue...I have seen ocing take its effects on the mobo and it power regulations....I have seen this effect to place the cpu on a new mobo and have it go right back to similar OC...

I cannot see 1.425v set in mobo causing this....Temps seem reasonable as well....

I would look at PSU as well...they do degrade...what are the specs of the PSU???


You could do a complete strip down...rebuild...new paste...and rebuild part by part with new install and see if things go back to normal...

It's a 500W Antec TruePower 2.0