![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Yesterday I completed my new rig:
Antec 900
Corsair CMPSU-TX650 PS
ASUS P5Q PRO
Intel C2D E8400 E0 (overclocked to 3.6GHz)
ZeroTherm Zen FZ120 cooler
4GB DDR2-800 RAM
Radeon HD4870
WD 640GB SATA HD
Vista Home Premium 64-bit
My goal when I created my shopping list was to overclock the E8400 to 3.6GHz with a 1:1 ratio. That's worked out pretty well so far... used Prime95 and IntelBurnTest for hours overnight with no errors.
What surprised me was I didn't have to add any VCORE to the E8400 to get the overclock. Yes, I've read in lots of forums where people are taking them to 4.0GHz and above with air cooling but I figured for sure I'd have to tweak voltage up a little to maintain a 20% overclock... nope. So here I was, expecting to be up the whole night getting to 3.6GHz tweak by little tweak and it took all of 3 minutes.
The problem was I was hitting sustained temps of 58°C while running 10+ loops of IntelBurnTest at full tilt on both cores. Others have reported temps as high as 70°C+ on the same CPU but I don't like anything even close to 60°C.
Screen shot of original IntelBurnTest, 3.6GHz, 1.3v (default)
Screen shot of CPU Hardware Monitor after IntelBurnTest, 3.6GHz, 1.3v, 58°C
Screen shot of CPU-Z, 3.6GHz, 1.3v
So I just had to do something with all that remaining time...
Using CPU-Z, Hardware Mon, Prime95, IntelBurnTest and 3DMark '06 I decided to see how *low* I could get the voltage and still be stable. Starting at the default 1.3v and 58°C, I kept adjusting down and verifying... adjusting down and verifying... adjusting down and verifying and finally reached 1.01v at 3.6GHz before I caught a single error. So I bumped it back up three steps to 1.10v at 3.6GHz and re-ran 5 loops of IntelBurnTest and 5 loops of 3DMark '06 clean again... rebooted, did both tests again and still with no errors. Now my CPU temp gets no higher than 45°C (a 13°C drop)
Screen shot of CPU Hardware Mon, 3.6GHz, 1.1v, 45°C max after IntelBurnTest
Screen shot of CPU-Z, 3.6GHz, 1.1v
Granted I haven't burned it in for 24+ hours yet but that's the plan for today. Based on testing so far I've been clean down to 1.05v at 3.6GHz but since I saw an error at 1.01v I bumped it 3 rungs back up my testing ladder to feel safer. My question is: will I be running any sort of risk carrying a 20% overclock with very low voltage? I've never read or heard any such thing but as the tag above says, I'm a n00bie.
Antec 900
Corsair CMPSU-TX650 PS
ASUS P5Q PRO
Intel C2D E8400 E0 (overclocked to 3.6GHz)
ZeroTherm Zen FZ120 cooler
4GB DDR2-800 RAM
Radeon HD4870
WD 640GB SATA HD
Vista Home Premium 64-bit
My goal when I created my shopping list was to overclock the E8400 to 3.6GHz with a 1:1 ratio. That's worked out pretty well so far... used Prime95 and IntelBurnTest for hours overnight with no errors.
What surprised me was I didn't have to add any VCORE to the E8400 to get the overclock. Yes, I've read in lots of forums where people are taking them to 4.0GHz and above with air cooling but I figured for sure I'd have to tweak voltage up a little to maintain a 20% overclock... nope. So here I was, expecting to be up the whole night getting to 3.6GHz tweak by little tweak and it took all of 3 minutes.
The problem was I was hitting sustained temps of 58°C while running 10+ loops of IntelBurnTest at full tilt on both cores. Others have reported temps as high as 70°C+ on the same CPU but I don't like anything even close to 60°C.
Screen shot of original IntelBurnTest, 3.6GHz, 1.3v (default)
Screen shot of CPU Hardware Monitor after IntelBurnTest, 3.6GHz, 1.3v, 58°C
Screen shot of CPU-Z, 3.6GHz, 1.3v
So I just had to do something with all that remaining time...
Using CPU-Z, Hardware Mon, Prime95, IntelBurnTest and 3DMark '06 I decided to see how *low* I could get the voltage and still be stable. Starting at the default 1.3v and 58°C, I kept adjusting down and verifying... adjusting down and verifying... adjusting down and verifying and finally reached 1.01v at 3.6GHz before I caught a single error. So I bumped it back up three steps to 1.10v at 3.6GHz and re-ran 5 loops of IntelBurnTest and 5 loops of 3DMark '06 clean again... rebooted, did both tests again and still with no errors. Now my CPU temp gets no higher than 45°C (a 13°C drop
)Screen shot of CPU Hardware Mon, 3.6GHz, 1.1v, 45°C max after IntelBurnTest
Screen shot of CPU-Z, 3.6GHz, 1.1v
Granted I haven't burned it in for 24+ hours yet but that's the plan for today. Based on testing so far I've been clean down to 1.05v at 3.6GHz but since I saw an error at 1.01v I bumped it 3 rungs back up my testing ladder to feel safer. My question is: will I be running any sort of risk carrying a 20% overclock with very low voltage? I've never read or heard any such thing but as the tag above says, I'm a n00bie.