Ok, so I've very slowly been trying new things with my machine, going for slightly higher overclocks and such. I've been running at 4.2GHz and 1.23V for quite a long time, and it's been rock solid. I've tried 4.5 and 4.4, but with my current cooler, and the apparent voltage I need for those to be stable, I have always gone back to 4.2.
So last night I went to see if 4.3 was still in that "just a little more voltage" range.
I booted at 4.3 and 1.27V, and it passed the Standard IBT just fine, and I backed it down to 1.25V and it passed Standard IBT fine as well. I then proceeded to use the computer for quite a while, including 2 hours of multiplayer MW3. No problems at all. Before I went to bed, I put on P95 for the night on Blend, and it was fine when I went into bed 20 minutes later, but had BSODed by the morning. So, I bumped the voltage up by 0.01, since I knew it couldn't be TOO far from stable given the gaming the night before.
Ran 15 passes of IBT on very high, and it passed stability fine, but after 4 passes, it started doing what I've had happen on my 4.4 and 4.5 OCs as well...it starts slowing itself down. I was getting 114 GFLOPs for the first few runs, then it begain slowing, to around 106-108...and the all too familiar to me (now) pulse of my PWM fan on my CPU going up and down (along with temps dropping significantly during these pulses). It still shows 100% load, but it's not the FULL 100% load that gets the CPU so hot.
The thought is that it's some sort of error correction possibly. (though i'm wondering if it's maybe thermal?)
Anyway, I tried raising voltage again....same thing. And again...same thing. So now I'm running at 1.284V and it still does the IBT slowdown after a few passes. However, I'm now 0.03V higher than when I was solid on MW3 for hours.... So I'm wondering if I'm even misreading that...I mean, the benchmark still finishes and it says the system passed and is stable.....
So, do I need to keep raising voltage until this is solid right through, or should I just not worry about the IBT slowdown and find what's prime stable at the lowest voltage I can? I'm running P95 right now at 1.28V to see if that's good (and then depending on what you guys say, I'd try and go lower if it's stable here)....so far no issues, though it's only been 10 minutes. Core temps are 68-72C right now (hitting 80C at this voltage in IBT, though, and 60s max during normal load situations). I'm not upgrading my cooler now, as I like the smaller heatsink vs the monster ones (and I don't know if something like a 212 would even fit in my case), I'm just looking for the fastest OC I can get with a low voltage, so that I'm not risking chip degradation, as I tend to keep my systems for 3-4 years.
Oh, I'm also using offset voltage for this, as I absolutely want it to run at lower voltage when at idle. Right now I'm at -0.02...last night I was at -0.06, but that's what failed P95.
So last night I went to see if 4.3 was still in that "just a little more voltage" range.
I booted at 4.3 and 1.27V, and it passed the Standard IBT just fine, and I backed it down to 1.25V and it passed Standard IBT fine as well. I then proceeded to use the computer for quite a while, including 2 hours of multiplayer MW3. No problems at all. Before I went to bed, I put on P95 for the night on Blend, and it was fine when I went into bed 20 minutes later, but had BSODed by the morning. So, I bumped the voltage up by 0.01, since I knew it couldn't be TOO far from stable given the gaming the night before.
Ran 15 passes of IBT on very high, and it passed stability fine, but after 4 passes, it started doing what I've had happen on my 4.4 and 4.5 OCs as well...it starts slowing itself down. I was getting 114 GFLOPs for the first few runs, then it begain slowing, to around 106-108...and the all too familiar to me (now) pulse of my PWM fan on my CPU going up and down (along with temps dropping significantly during these pulses). It still shows 100% load, but it's not the FULL 100% load that gets the CPU so hot.
The thought is that it's some sort of error correction possibly. (though i'm wondering if it's maybe thermal?)
Anyway, I tried raising voltage again....same thing. And again...same thing. So now I'm running at 1.284V and it still does the IBT slowdown after a few passes. However, I'm now 0.03V higher than when I was solid on MW3 for hours.... So I'm wondering if I'm even misreading that...I mean, the benchmark still finishes and it says the system passed and is stable.....
So, do I need to keep raising voltage until this is solid right through, or should I just not worry about the IBT slowdown and find what's prime stable at the lowest voltage I can? I'm running P95 right now at 1.28V to see if that's good (and then depending on what you guys say, I'd try and go lower if it's stable here)....so far no issues, though it's only been 10 minutes. Core temps are 68-72C right now (hitting 80C at this voltage in IBT, though, and 60s max during normal load situations). I'm not upgrading my cooler now, as I like the smaller heatsink vs the monster ones (and I don't know if something like a 212 would even fit in my case), I'm just looking for the fastest OC I can get with a low voltage, so that I'm not risking chip degradation, as I tend to keep my systems for 3-4 years.
Oh, I'm also using offset voltage for this, as I absolutely want it to run at lower voltage when at idle. Right now I'm at -0.02...last night I was at -0.06, but that's what failed P95.
Last edited:
