Maximum safe voltage for Phenom 2 x6 1075T under water?

[almighty15]

My chip will run 4Ghz Intel Burn test stable at 1.44v ( Load Vcore ) all day long, Now I'm greedy and want more.

I can hit 4.1Ghz Intel Burn Test Stable with 1.56v ( Load Vcore )

I want 4.2Ghz but as the voltage is at the 1.55v safe limit is it safe to continue?

It's cooled by a thermochil PA120.3 with 66cfm fans in push-pull and it's a CPU only loop.

Also will I damage my Crosshair 4 extreme by running voltages that high?

[charged3800z24]

I try to keep mine udner 1.5v for 24/7 regaurdless of water or air. Temp will obvious come into play some. I have seen some run higher for 24/7 use, but I don't recommend it.

[BeepBeep2]

I'd stick to about 1.525v on water. I've ran up to 1.6v for periods of time but this is NOT recommended.

[JumpingJack]

You might spend some more time looking in the OC threads and such, I can only relay my experiences. I ran an 1090T at ~ 1.51 V under water to get to 4.2 GHz, lasted quite a while, but mostly idling, total load time (benching, testing, etch) was I would guess 18 hours or so, then became unstable. My prior most stable OC was 4.0 GHz at ~ 1.45 as I recall, but after 4.2 GHz @ 1.51 volts went south, was no longer stable at 4.0 GHz @ 1.45, new stable clock is 3.8 Ghz at 1.45. Another odd thing is that my CPU reports 0 deg C regardless of software (AOD, Everest, HWmon, coretemp), prior to going unstable temps would read (this CPU never reported correct/logical temps, but it at least responded).

My advice, stay below 1.5V on Thuban.

[Ccaution]

... Another odd thing is that my CPU reports 0 deg C regardless of software....

core unlocker disabled?

@OP
1.55 is on the edge for 24/7@h2o, have you finutuned every voltage to the minimum possible there?

[JumpingJack]

core unlocker disabled?

@OP
1.55 is on the edge for 24/7@h2o, have you finutuned every voltage to the minimum possible there?

Yeah, I have tried with and without the ADC, NB frequencies, etc.

If I take a gander at the Thuban OC database, I had an entry there when I was first ramping up. At 4.0 GHz I was actually stable at 1.48 V, on air, which was a bit higher than some at the same clock.

I think it was just luck of the draw. I probably goosed it a bit too much trying to get back to 4.2, went as high as 1.58 almost 1.6 it was after a few attempts there that the thermal reporting went to 0 C. So that actually may have been the tipping point.

Nonetheless, still a good chip, happy with 3.8.

[almighty15]

core unlocker disabled?

@OP
1.55 is on the edge for 24/7@h2o, have you finutuned every voltage to the minimum possible there?

Not really, I'm such a noob as I have not got a clue what most of them do except for V-core!

My biggest issue is overclocking and getting the memory stable, Not got a flying clue on what 90% of the timings meen!

[Ccaution]

@almighty15
Get yourself the latest bios and start messing around with the vcore & cpu_nb voltage to understand how that affects stability. Turn turbo off to avoid overvolting. Feel free to ask for further info

@JumpingJack
Maaaan, it's so hard to believe you degraded a Χ6 @ 1.51 @ water! Have you tried the chip on another board? Where you sure about the past stability? Did you ever actually mm'd the board to see if it overvolts a lot? Something went very bad there...

[DedEmbryonicCe1]

My 1090T has been running on 1.5v under water (no Cool N' Quiet) for the full year since launch without any degradation.

[Manicdan]

cool n quiet does reduce alot of the risks, since your cpu is probably idling or using so little voltage most of the time anyway. mine is set to .925v

at stock when gaming if your turbo clocks are used, your already past 1.45v, even if its just on 3 cores. so i think 1.55v is ok, just as long as your not using it as a 24/7 cruncher.

right now im at 1.475v for 4ghz on all cores (ignore my sig, its way outdated), and i did some testing last night, idling its 18c on coretemp, load its 37c, when playing games its only 23c, since they use no where near the full power 6 cores has to offer. so while my limit might be 4ghz, i bet i can run 4.2ghz at the same voltage in games since temps will be lower

[almighty15]

The biggest problem I have is getting stability with a high FSB and a high RAM clock.

My old RAM was pathetic at overclocking so I changed it for some Kingston Hyper X that has an XMP profile for 1866Mhz or 2000Mhz.

The 1866Mhz profile is 9-11-9-27-1T at 1.65v

The 2000Mhz profile is 9-11-9-27-2T at 1.65v

My old RAM was only running at 1500Mhz 9-9-9-24-2T as that's all it could manage, I couldn't even tighten the timings a single bit!

Running stock settings with the 1866Mhz profile it's nice and stable and will happily sit through Mem test 86.

But as soon as I try for a 270+ FSB it fails mem test 86!

I'm a bit of noob when it comes memory clocking and in all honestly the only setting that I've only ever used while overclocking is the CPU Vcore.

I leave everything else on auto!

Also with my old RAM I was Intel Burn Test stable at 300 FSB for a 3Ghz north bridge.

With the new RAM I can't even get above 285 FSB

[almighty15]

Anyone? ?

[BeepBeep2]

Normal memory clock limits on the 3:10 divider are around 1850-1950 for stability...

HTT Ref clock around 285 will bing you to 1900+ on that divider, yes.

[JumpingJack]

@JumpingJack
Maaaan, it's so hard to believe you degraded a Χ6 @ 1.51 @ water! Have you tried the chip on another board? Where you sure about the past stability? Did you ever actually mm'd the board to see if it overvolts a lot? Something went very bad there...

Yep. I borked the chip no doubt, like I mentioned above, it does not report temps any more. It's not the board because I can plug in my Deneb X4 and hit the same OC I have always been able to on that chip (and I don't have another board to check out the X6).

Yeah, I am sure of stability, see this thread: http://www.xtremesystems.org/forums/...d.php?t=250607 I half bought the chip just so I could contribute to the forum.

When I was reaching 4.0 GHz at 1.48 I was on air, was both Prime and y-chruncher stable at that mark, then I went on water and got to 4.2 GHz at 1.51, ran there without much problem until one day, bam no temps then blue screened on something routine (CB R11.5), after that would post but would not boot to windows without either locking up or BSOD.

I can now boot at 4.0 GHz / 1.48 (just as in the thread above), but I am no longer prime or y-cruncher stable, prime will error within 30 min or less. No BSODs though, 3.8 at 1.43 or so is where I can now prime stable (over night).

I would have suspected the board as well if it weren't that the temps are no longer reporting.

I write it off as just the luck of the draw, some chips go gang busters, take 1.55 or higher and no problems hit 4.2/4.3, others people struggle to get over 4.0 GHz no matter the voltage.

[Manicdan]

^is it possible the board gave it too many volts and borked it that way?

[BeepBeep2]

I've had my chip up in 1.63v range on water for 1M no issues, 1.6 for 32m
I've run 1.57v load LLC on 1.53 idle a long time before no issues.
Chip still does 2 hour P95 Blend at 4 Ghz 1.416v

Still, I dont recommend this to anyone

[almighty15]

What's your cpu temp? Not core temp but actual cpu temp.

Mine was touching 70c while at 1.55v load.

[[XC] Oj101]

For what it's worth, I've had my Phenom 2 1055T at 1.525v for over six months under a TRUE Black with two Gentle Typhoons in push/pull without ill effect, so it's very much a case of YMMV

[JumpingJack]

^is it possible the board gave it too many volts and borked it that way?

It is certainly possible, I did most of my tweaking when I first bought the CPU (I worked diligently but quickly because I wanted to contribute to Particle's OC database at the time). But I was always apprehensive about going over 1.5V.

I have not MM'ed the board to check the actual delivered voltage, so it is also possible that the board simply is not reporting actual voltage correctly. I think this is very low probability.

When I lost 4.2 GHz stability and the thermal reporting went to 0 C always, I tried several attempts with increasing to 1.58 as well as bumping NB voltage another 0.05 V to no avail. It is possible I borked the 4.0 GHz stability with that 1.58 V mucking around.

Nevertheless, when I was originally contributing to the OC database I was on air and I noted that this particular chip was struggling with stability at slightly higher voltages than what others were reporting, c'est la vie that is just how it works.

I don't think there is anything wrong with the chip, but to be safe I would avoid anything higher than 1.5 V for 24x7 in the future. Just like I avoid anything higher than 1.4 V on my 990X.