• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

ASUS P8P67 PRO

Hmm. Have you seen others with this problem on the EVO? The Pro stays reasonably consistent with the right tweaks.

I could see how you could say the level of flux is reasonable, just remember I'm comparing it to a board that had zero flux. Something did just occur to me though. The P6X was only that rock solid up to 1.25v and I've never run my EVO at that low a voltage. So maybe if I took it down to that it wouldn't fluctuate either.

Well, frmo what's in the manual, I am led to belive that if you are not tweaking the DIGI VRM, maybe the heatsinks aren't needed. But as soon as you tweak it to provide more current, those heatsinks must be in place, as stated in the manual over and over and over again.

Well that's the thing, just about anyone overclocking following the Asus guide their reps have posted around on forums would have been messing with vrm settings. If there's any chance that accounted for the relatively small number of dead boards and instability at high clocks I've seen on some threads it would be an interesting result.
 
I'm pretty sure both the EVO and this board share the same VRM design, and as such, there's not going to be muc hdifference between the VRM and how it deals with load. If aything, the larger heatsink array on the EVO should keep teh VRM a bit cooler, as there is only the same PCIe buffer chip under the EVO's extra heatsink. The EVO has a couple of other added features too, of course.

I can't comment on board deaths. I do fell however, that because ASUS mentioned VRM cooling so amny times on the manual, if care isn't taken to ensure it is cooled properly, and this leads to the board dying, the onus is not on ASUS to fix the problem. From waht I can tell ,teh cooling is more than adequate, and the flex seen around the top edge of the board has been present on just about every P67 board I've seen, and is most likely due to the socket retention mechanism.

Any product is going to have failures "in the field". A few months back it was reported that jsut under 10% is about the industry standard, so anything less than that, in my books, is pretty good. I can't see this being that large of an issue, or it would be an issue for everyone else, as well as ASUS.
 
10%? That seems high. I'd of expected maybe 3% for motherboards. Is that DOA or any failure within the warranty period?
 
10%? That seems high. I'd of expected maybe 3% for motherboards. Is that DOA or any failure within the warranty period?

I don't remember exactly, but I think within warranty.

Ok, i found the orginal thing I'm talking about. This is based on store failure rates, will have to dig into it further.


http://www.hardware.fr/articles/810-1/taux-pannes-composants.html

The failure rates reported are on parts sold from 1 October 2009 and April 1, 2010, for returns created before October 2010, six months to 1 year of operation and on models with a minimum sample of 100 sales, and the figures in parentheses from a previous report last April


Anyway, most rates are below 10%, far below. But on my end, in manufacturing, we order 10% more than needed in supplies, for just such events. Do we end up with extras? You bet.

Maybe that 10% number is in my head for a different reason. Can't place it at the moment, to be honest, but anything less seems OK to me? I mean, sure, it sucks when you get bad stuff. But it does happen to us all.
 
I haven't touched base with MSI, Zotac, Sapphire, Asrock, Biostar, Foxconn, or eVGA as of yet. Hopefully we can get them on "board" :laugh: some time in the future.




No problem. Thanks for the feedback! ;)

If you keep up with reviews like this.. You shouldn't have a problem with them getting on "board"

Good review.

:toast:
 
I haven't touched base with MSI, Zotac, Sapphire, Asrock, Biostar, Foxconn, or eVGA as of yet. Hopefully we can get them on "board" :laugh: some time in the future.




No problem. Thanks for the feedback! ;)

Off topic for a second. I have a contact at Biostar that might be helpful, if you are interested in reviewing their boards. :)
 
Nice review...I have the EVO version and talk about being noob all over again with the new BIOS layout:wtf: Will have to get used to it I guess.
 
Nice review...I have the EVO version and talk about being noob all over again with the new BIOS layout:wtf: Will have to get used to it I guess.

Its not really much different than a traditional BIOS. you can still use the same keyboard commands you always have (95% of the time thats what I do lol) and the pages/subpages are laid out similar to how they used to be.
 
Its not really much different than a traditional BIOS. you can still use the same keyboard commands you always have (95% of the time thats what I do lol) and the pages/subpages are laid out similar to how they used to be.

I was going to say the same, but you know, the addition of the VRM options, as well as the CPU-specific options can be a little daunting. But, those options make a difference when overclocking, for sure, so I'm very glad it's all there.
 
the addition of the VRM options, as well as the CPU-specific options can be a little daunting.

Daunting is an understatement, I'm a pretty savvy guy when it comes to playing around BIOSes but not this one. I will eventually get used to it so I'm playing with it whenever chance I get


But, those options make a difference when overclocking, for sure, so I'm very glad it's all there.

I will soon find that out once I start OCing;)
 
If you run into any issues, let me know, I might have a few tricks for ya. ;)

I'm playing as well still...maybe I can get lucky and get 5.0ghz stable 100%.
 
There is some big problem on the fan controller on this board, the controller only supports PWM (4pins) fans, if you are running anything 3 pin, there isn't any voltage control support, it will always run at full speed.
 
That's not really a problem, and is actually pretty common. I use full speed fans on the noctua cooler, and they are 3-pin, so didn't notice. I'll check and see how it works.
 
There is some big problem on the fan controller on this board, the controller only supports PWM (4pins) fans, if you are running anything 3 pin, there isn't any voltage control support, it will always run at full speed.

Actually that is only true on the CPU header the chassis headers still control 3pin fans. (I just hooked my CPU fan into chassis1 and a low speed into my CPU header although you could just disable CPU fan error warnings :) )
 
Well I been playing with my new Asus P8P67 Deluxe. I got PC Mark 7 to run stable at 4.5Ghz at 1.3v. Soon as I started Black Ops it crashed though.

I don't really understand about all the VRM and fixed frequency mode.

Can you post some tips please, i'd like to get mine stable at 4.5Ghz.
 
OK, so there is several different options you can tweak.

First, of course, is loadline calibration. Loadline calibration is an effort to keep voltages constant, by providing more current, as the votlage drops based on how much current the CPU consumes.

As you increase CPU frequency, the amount of current used by the CPU to switch states increases(which is why temps increase with speed, even though volts may stay the same).

The VRM charges the capacitors at a set frequency. if we increase the frequency that they are charged, the VRM can supply more current, and the cpu can use more before voltage fluctuations are noticed. If the frequency between the VRM and the CPU doesn't mesh well, it can cause stability issues.

You also have the option of customizing the VRM so that the phases are either all always on, or they are enabled dynamically based on load. When it is set dynamically, the time between phases going from off to on states may lead to teh VRM not supplying enough for a fraction of a second, which may be enough to break stability.

Because it deals more with how much power the CPU comsumes, there really isn't real specific adjustments I can suggest, as it will vary from CPU to CPU, as the VRM does react a bit differently with each of the CPUs I have on hand here.

I look at it this way...the voltage you give is the wave the current rides to the CPU. Instability is rocks that make the wave crash, and not continue, so we must adjust the wave's form so that it go over the rocks. Depending on how far from the coast(stock) the waves must go, they might need to be taller, or they might need to be more often, or they might need more force. In the past, we've only had very little options to make the wave go farther...increasing the wave height(voltage), and the wave force(current), but the frequency, in most products, is not something we could adjust.

I apply this to my overclocking in the reviews, and try to use the same wave, 1.27v, to see how far, with that wave, things will go. If the VRM supplies more current, the wave goes farther, but at the same time it gives each board the same wave to work with, and the same rocks(the CPUs stock), so really isolates what's going on in the VRM. Hence the VRM power comsumption numbers in my reviews.

The parts in the VRM usually dictate the frequency of the VRM, but guys that use extreme cold have found that the higher frequency can allow for the CPU to go further. To me, this is why certain products have huge followings for extreme clocks...of course, it's not quite exactly os simple, but I think you get the point.

So you need to play with it...finesse it...settings that are too high may only offer additional heat, so of course, you can just push everything to the max, but you'll also have the extra heat to deal with.

If your votlage is not stable, of course, some of those waves may hit rocks, so the main thing to try, IMHO, is stabilizing that wave. The votlage SHOULD vary a bit under load, as that is the nature of the circuit, but not too much, or the wave behind will not reach as high, and stability may falter.


Of course, this is all very abstract...but if voltage is fairly constant, than you may want to look and increasing the VRM frequency, but really, error codes and how the system crashes should point you in the right direction. Maybe the CPU has too much heat, maybe the voltage needs to be higher, maybe it needs more current...

Over the next several weeks while working between reviews, I'll be posting more info specific to each board, and overclocking, and what I run into for issues, and how I overcame them, so perhaps we can take that journey together.

I have a review to finish right now, but will be done later today, so look for my posts this weekend, as the ASUS is the board I will be looking at this weekend, and perhaps I'll have better, more direct, info for ya.
 
Thanks for the reply. I found a good article here on overclocking on asus motherboards:

http://hardforum.com/showthread.php?t=1578110

I have also noticed that speedstep does not seem to work even though it is enabled in the bios. My i2600k is currently stable at 4.3 Ghz with no voltage tweaks. It does not idle back within windows. Any ideas why this is happening?
 
Keep up the nice motherboard reviews cadaveca:toast:
 
Thanks for the reply. I found a good article here on overclocking on asus motherboards:

http://hardforum.com/showthread.php?t=1578110

I have also noticed that speedstep does not seem to work even though it is enabled in the bios. My i2600k is currently stable at 4.3 Ghz with no voltage tweaks. It does not idle back within windows. Any ideas why this is happening?

Yeah, those tips on ASUS boards were provided by ASUS staff and are all over the place. There is some very good tips in there.

All power-saving features should be enabled in BIOS as well as in Windows. I am not sure, exactly why EIST isn't working for you, but you could try a couple of things;

RESET CMOS, and re-enter you OC settings. BE careful to not disable any power-saving features, and OC using the TUrbo multipliers.

Choose CPU power settings that do not force full clocks in Windows power settings.
 
Thanks for the reply. I found a good article here on overclocking on asus motherboards:

http://hardforum.com/showthread.php?t=1578110

I have also noticed that speedstep does not seem to work even though it is enabled in the bios. My i2600k is currently stable at 4.3 Ghz with no voltage tweaks. It does not idle back within windows. Any ideas why this is happening?

On my latest bios the c-states were set to AUTO not ENABLED. As Cadaveca stated, setting the C-states to enabled sorted out my problems with speedstep not working correctly :)
 
Wow, very impressive board and for the price, I Cannot believe the performance. That one capacitor looked a bit crooked due to the heatsink. Love the included accessories as well. Great OCing and I am a huge fan of the UEFI bioses coming out.
 
What really makes the bord so good, in my eyes, is that it manages to score so high, yet sips the power while doing so.


When I started overclocking, you needed a soldering iron, and a reasonably steady hand. The board's VRM design is such that there really isn't any need for any of that any more, as is evident by the auto-OC software result.
 
EIST setting in BIOS should not effect anything, speedstep is now controlled by DIP2. ASUS recommends leaving it on though.

Also disabling C6 power states will remove the ability to multiclock within windows. Not an issue for many, just if it happens you know why :)



Your 5GHz autoOC result is insane. Best I maanged was 4.8. and that was with manually setting voltages before hitting autotune
 
And that result was with the "worse" chip.

I did not set voltages; VRM adjustments were made to set droop to what other products were doing only.

Mind you, that chip does do 4.5 GHz @ 1.27v, if I manually set voltages, it would have been higher, probably. The point was to NOT do anything manually, and see what the software does, anyway.

It's any interesting subject, because in the end, that speed wasn't fully stable. The chip can do it, with better cooling.
 
Back
Top