ASUS P8P67 Pro/Evo/Deluxe OC discussion thread

[deaffob]

You know that CPU Capability doesn't give you any stability? It doesn't matter if you put it at 100% or 140%, it won't give you any more stability.

[coolhandluke41]

You know that CPU Capability doesn't give you any stability? It doesn't matter if you put it at 100% or 140%, it won't give you any more stability.

do you actually have the answer to my problem or just asking?

[Mathemagician]

^^ What are you going to do with that 4.8GHz that you can't do with 4.2GHz for 24/7? Besides, that seams to be buggy at idle, but try the 130% too.

You have 3 options: @4.8GHz with a manual high idle/load vCore, or with low-idle offset mode that seams to be buggy at idle, and can crash on you anytime, or a @4-4.2GHz low-volt manual vCore. You choose yourself, but you have the 140% to try too, LOL.

Some of us use our PCs for more than Email and Word (and games). Though it can be a bit tricky to get stabilized, offset voltage and LLC are just what the doctor ordered for folks like me. Rock stable at 4.7, idles around 0.9v and runs under load at 1.38v when i need it. There is a good bit of overhead left, but I'm waiting for some more reliable information regarding safe 24/7 voltages...

I was getting x124 at idle/light load for a while, also. It was apparently due to insufficient voltage at lower loads. You can tune it out by easing up on the LLC and bumping the offset.

[Praz]

I was getting x124 at idle/light load for a while, also. It was apparently due to insufficient voltage at lower loads. You can tune it out by easing up on the LLC and bumping the offset.

What I was trying to say in my previous post. If load is stable but idle or off-idle is not increase the offset voltage. If the final load voltage is considerably more then necessary drop down to the next lower LLC level. Doing this may require bumping offset even more to bring the loaded voltage back to where stability is. This won't have a large effect on idle voltages because they increase considerably less then the loaded values per offset step.

[deaffob]

do you actually have the answer to my problem or just asking?

Neither. I was stating a fact that you won't get any more stability from changing the CPU Capability setting.

[crossg]

wow went from 1253 to 1204 and now the same settings that would cut power at 50x100 on superpi1m are now stable for superpi32m

EDIT: anyone else able to boot 1.5v+ but as soon as you start a bench it cuts power?

Just had this today at 1.53v+ and above, bench doesn't hang it just cuts power even SPi. Didn't matter what I did with the voltage. Ran Vantage and 3DMK11 Sli @ 52X with 1.488 and everything worked nicely. Maybe time to look at the Maximus board.

[coolhandluke41]

@orion23 weird bumped to +.04 under P95,Linux and the highest V i get is 1.416 from 1.408 (most of the time 1.408,@+.03 was 1.400~1.408)


@Praz thanks ,running stable so far

[Lu(ky]

Hey guys I been running BIOS 1053 and at one point I was @ 4.8GHz with 1.365V but since I moved my Western Digital VelociRaptor & DVD rom drive from sata II to sata III side I started getting cold boots problems again and now I can not even get into Windows 7 with any OC above 4.5GHz no matter what voltage. Not sure if I should jump to 1204 or 1253?

Also what BIOS are you guys running above?
Are you guys using BIOS for OCing or Asus AiSuite II?

Thanks

[bokoy]

New owner here with the deluxe... I'll be reading you

[Lu(ky]

I Well I decided to jump to 1204 first and I am glad I did. I was able to boot with picture below @ 5.41 1.45v blck @ 107.2 = memory at 2000MHz... I am stable at most setting like 3DMARK, AIDA64, a little of Prime95 for 30 minutes small FFTs and Blend, BUT!!! when I try to use SuperPI 1M computer shuts off within seconds? But at least I am finally at 5GHz I was getting very frustrated seeing everyone else get there so easily.. Getting tired have to go back to work tomorrow night I will miss the Superbowl..

[Sam_oslo]

Some of us use our PCs for more than Email and Word (and games). Though it can be a bit tricky to get stabilized, offset voltage and LLC are just what the doctor ordered for folks like me. Rock stable at 4.7, idles around 0.9v and runs under load at 1.38v when i need it. There is a good bit of overhead left, but I'm waiting for some more reliable information regarding safe 24/7 voltages...

I was getting x124 at idle/light load for a while, also. It was apparently due to insufficient voltage at lower loads. You can tune it out by easing up on the LLC and bumping the offset.

Yeah, there are different usage patterns. I use my SR-2 for heavy duty jobs like simulations and 3D-renders.
This setup is for emails, surfing and other low-load tasks, and it runs on idle mostly. Therefore, the idle and low-load stability is very important.

Yeah, increasing the Idle-voltage is the way to go, and easing LLC will help too, for sure. These new DIGI+ VRM-buzz may help too, and you guys have to give feedback what is working, or not, in real 24/7 usage.

[Praz]

I use my SR-2 for heavy duty jobs like simulations and 3D-renders.

Then you have the perfect tools to test stability with these CPUs. I can have a configuration that will pass Prime but not LinX, LinX but not Prime or pass both LinX and Prime. But fire up 3ds Max and not make it through an hour of rendering or the encoding of a DVD. I'm not even messing with LinX or Prime anymore for stability testing. If I can pass 3ds Max and HandBrake everything I throw at the system runs without issue.

[Gr1mR34p3r]

Prime95 has always seemed to do the best for me.

I guess Prime95 is better than IntelBurnTest for SB.

IBT 20x maximum stable, but BSOD in idle/light loads
46x100 -0.015V LLC Auto (1.240-1.256V)
48x100 +0.045V LLC Auto (1.296-1.328V)

Prime95 Blend
46x100 +0.030V LLC Auto (1.288-1.304V) - 8 hours stable
48x100 +0.090V LLC Auto (1.344-1.360V) - temperatures rise 83°+ to test it for long time, but +0.080 failed in 9min and +0.085 was stable for 15min+, so +0.090 should be stable

What I was trying to say in my previous post. If load is stable but idle or off-idle is not increase the offset voltage. If the final load voltage is considerably more then necessary drop down to the next lower LLC level. Doing this may require bumping offset even more to bring the loaded voltage back to where stability is. This won't have a large effect on idle voltages because they increase considerably less then the loaded values per offset step.

Have to try with higher LLC myself, Auto is like Regular on Offset and Ultra High on Manual vcore, so High for starters.

[MaStA]

I guess Prime95 is better than IntelBurnTest for SB.

IBT 20x maximum stable, but BSOD in idle/light loads
46x100 -0.015V LLC Auto (1.240-1.256V)
48x100 +0.045V LLC Auto (1.296-1.328V)

Prime95 Blend
46x100 +0.030V LLC Auto (1.288-1.304V) - 8 hours stable
48x100 +0.090V LLC Auto (1.344-1.360V) - temperatures rise 83°+ to test it for long time, but +0.080 failed in 9min and +0.085 was stable for 15min+, so +0.090 should be stable


Have to try with lower LLC myself, because these voltages are 0.02-0.03V higher than necessary. Auto is like Ultra High, so High for starters.

What type of cooling are you using to reach 83C+ at only 1.36v? I am using a Cooler Master Hyper 212+ with an added fan to pull and at that voltage I only reach like ~61C. It is a cheap cooler as well...I bought mine at Microcenter a few months back for $16.99

[Gr1mR34p3r]

It's Scythe Yasya, but with only pull fan at 1400RPM, because I have tall RAM heatsinks. If I also hold a push fan at 2000RPM in my hands there is only 2 degrees difference.

[Darkstormz]

Anyone got the solution for this problem...

When I try to use AS SSD Benchmark for my Intel X25-M G2, I will get a message "Display driver stopped resonding... and has recovered...blabla"
I use my Zotac GTX480 @ PCIe_x16_slot1.

When I move the card to PCIe_x16_slot2 (I only get 2.0 x8 mode) the AS SSD Benchmark will run without the display driver not responding message.
My SSD is hooked up @ Intel SATA 6.0 port 1.

[coolhandluke41]

here is my result
48x100 ( manual for stable 1.38v) +0.040V LLC Ultra High (1.024v idle---1.408~1.416V load) P95/Linux stable
@130% CPU capability aka (DIGI+VRM CPU Capability) stable idle/light load
@120% or 110%,100% BSOD in idle/light load
It's just my opinion but i think LLC and CCC have to be equal


Uploaded with ImageShack.us

[MaStA]

It's Scythe Yasya, but with only pull fan at 1400RPM, because I have tall RAM heatsinks. If I also hold a push fan at 2000RPM in my hands there is only 2 degrees difference.

Hmm...that unit has really good reviews so I would expect better temps than what you are getting out of it. Have you tried to remount it and look at how your last method of applying the TIM worked? I have tried several different methods to achieve the temps I currently have with my heatsink. I would also try to have a push and pull with the push being the strongest (most CFM generating) fan. If you can't fit the fan because of your ram sometimes you can mount the fan a little higher on the heatsink and have no adverse effects...but I'm not sure if that is possible with that heatsink or not.

[Sam_oslo]

here is my result
48x100 ( manual for stable 1.38v) +0.040V LLC Ultra High (1.024v idle---1.408~1.416V load) P95/Linux stable
@130% CPU capability aka (DIGI+VRM CPU Capability) stable idle/light load
@120% or 110%,100% BSOD in idle/light load
It's just my opinion but i think LLC and CCC have to be equal


Uploaded with ImageShack.us

That was my suspension too, that 130% in combination with DIGI+ VRM Phase&Duty-Control= Extrema both, may actually make the MB and power delivery responsive enough to keep it stable at idle/low-load.
This is interesting, and can make me move to higher 24/7 OC. Keep us updated how it's behaving in different low-load scenarios, in coming days.

[coolhandluke41]

That was my suspension too, that 130% in combination with DIGI+ VRM Phase&Duty-Control= Extrema both, may actually make the MB and power delivery responsive enough to keep it stable at idle/low-load.
This is interesting, and can make me move to higher 24/7 OC. Keep us updated how it's behaving in different low-load scenarios, in coming days.

testing @4.7 120% LLC high so far

[Sevenz]

to test the "light" load stability,test linx 20 pass on 1 thread.
in my tests i could do 25 pass linx avx on 4/8 threads but was failing linx 1 thread on the second pass so i had to up offset to pass linx 25 runs on 1 thread.
You can do it with prime too.

[dengyong]

Anyone know how to make these boards stop waking from S3 sleep when you move the usb mouse ?

[orion23]

That was my suspension too, that 130% in combination with DIGI+ VRM Phase&Duty-Control= Extrema both, may actually make the MB and power delivery responsive enough to keep it stable at idle/low-load.
This is interesting, and can make me move to higher 24/7 OC. Keep us updated how it's behaving in different low-load scenarios, in coming days.

Pretty stable for me here as well.

As of now, I'm sticking to Medium LLC with +120 offset for a load Vcore of 1.408 using Linx, and 1.386 using Prime95.
Idle, voltage drops as low as 1.1V

Using an Ultra high LLC, then the idle voltage drop actually goes somewhat lower than 1.1V. Here I need +.055 offset.
Had a couple of ramdom BSOD with these settings so I went with the medium LLC instead.

More testing to go....

BTW, this is using bios 1253 and windows 7 SP1. Haven't tried 1204 yet...

[nijel]

Anyone know how to make these boards stop waking from S3 sleep when you move the usb mouse ?

It's Windows thing:

[dengyong]

It's Windows thing:

Thanks.