**Official** Asrock Z68 Extreme4 thread. Share your experiences, OCs & tech issues!

[Ket]

Well, its certainly a strange problem. Both the P67 and Z68 I have behave themselves. The third point I suggested is deffinately worth a shot though. Changing the LLC level will give just a touch more vcore to the CPU. One thing I did notice when testing out the P67 and Z68 is that the P67 defaults to 1.18v vcore, while the Z68 defaults to 1.16v which makes me think Asrock may of reduced the potency of their LLC a tad with the Z68 as a lot of people got unexplainable BSODs with the P67 if setting LLC to "Level 1". Best guess is that LLC setting pushed voltages too far out of spec and hence the BSODs. Now it seems like it could be the opposite problem, if LLC potency was reduced a tad because of the BSODs when setting "Level 1" on the P67 board, now the BSODs on the Z68 may be because the "Level 5" LLC setting isn't quite giving enough juice for people using the IGPU.

The simple answer could also always be theres a bug in intels chipset / IGPU driver, but I thought I would include the long glamorus techie theory as well

[qbazdz]

Great... somehow I managed to delete all my typing from just a second ago... great start on my posting here...

Well... what I was saying is...

Love the board! my chip is quite good overclocker, so managed to get 5.24GHz in a suicide high Vcore run to get the best SuperPi result


I seem to have a strange issue, or two really.
First, the PC doesnt wake up from sleep on the first try. I had a stability run of BIONIC and forgot to modify the power profile and the PC went to sleep. I powered it back on and it did something and the board had a FF, if I remember correctly, on the display.
I powered it down and on again and I got a Resuming Windows screen and it booted from sleep! strange... I had that before but have no idea why...

2nd thing is the idle CPU clock... Im trying to find a stable, lowest Vcore set up for 4.8GHz day to day usage, and it doesnt seem to be clocking down on IDLE.
Don't remember the settings now, but I am sure that the speed step tech is ON.

Any help will be much appreciated

Oh... that is all on the BIOS 1.5

[Ket]

The first issue sounds like a unstable OC, especially if you were at 5.2GHz when it happened. Second issue, unless you inadvertently disabled any of the power saving stuff the CPU would downclock, also, if you still had anything like BOINC running, the CPU would stay full speed

[qbazdz]

I only now run it @4.8 so all the issues were then.
I'll report the settings later today.

Bionic was definitely closed and off but I modified some stuff as per your screenshots... (posted earlier)...

[Ket]

If you changed some stuff based on my screenshots, its very possible you turned throttling of the CPU off so it would run full speed all the time. Because I run BOINC theres just no need for me to have power saving options enabled

[qbazdz]

So what power saving options can be enabled and not disturb the o/c potential?

[Ket]

OC potential won't be disturbed by turning off power saving features. Don't turn off Intel Speedstep Technology in OC Tweaker though. The CPU will run your OC speed but once any stress is put on the CPU it will revert to stock speeds.

[Ket]

Updated link on first page with new driver package

[Ghostleader]

Some new Asrock 1155 socket board is coming soon

A1.jpg

A2.jpg

Source

[AAbenson]

LOL thats some great stuff coming from Asrock
was thinking about getting a ASUS R.O.G Maximus IV GENE-Z but now im gonna wait out this Asrock release and see how the Extr7 and Extr4 are doing and what prices they will have.
power phases are the same count on both boards so im not sure if getting the Extr7 will make that big a difference over the 4 version

[jpm]

Some new Asrock 1155 socket board is coming soon

official spec sheet from asrock is here:
http://www.asrock.com/download/e-cat...me4%20Gen3.pdf

doing a quick diff vs. the Z68 Extreme4 spec sheet, this is what i see to be changed, besides the obvious addition of PCIe 3.0:
* V8+4 power phase design vs. the non-Gen3 V8
* "premium gold" caps vs. the non-Gen3 "duracaps"
* PCIE5 operates at full x16, as compred to the current x4 operation
* claimed support for DirectX 11 when using Ivy Bridge CPU, although i presume this is not a motherboard limitation and will apply to the non-Gen3 as well
* somewhat bizarrely, the non-Gen3 supports "THX TruStudio PRO" whereas the new Gen3 board only supports "THX TruStudio", this despite having the same exact Realtek chip - although i'm quite sure there won't be a single person on the planet that cares about this difference

that looks to be it. all the on-board chips (marvell, etron, realtek, broadcom) are unchanged. i can't say i'm regretting picking up the non-Gen3 with such an incremental upgrade on the horizon.

[qbazdz]

Love the heatsinks and color combo on the Extreme7! fantastic stuff ASRock!

[EzyRyder]

Updated link on first page with new driver package

Ket, what changed in the package?

[Ket]

Newer drivers

ED - If you can believe it, despite UEFI 1.5 for the Z68 only coming out a few weeks ago Asrock already have another UEFI! Version 1.6 Clicky!

[qbazdz]

This is just driving me insane!!!
Ive been stability testing for a past few hours 4.8GHz @1.370V. All went ok, moved to 1.365V, tested... all ok. So moved to 1.36V and got the BSOD. Ok. went back to 1.365V to test more and now I cant get it to boot! Stupid thing!

Ive been thinking about the Load Line Calibration a bit and can some correct me if Im wrong please.. According to the graph, depending on the different level we get less or more Vcore droop when CPU gets more load. Less droop = better stability => Level 1 should be the best. The main reason for the Vdrop is for the CPU not to exceed the TDP limit when loaded.
I still noticed though that when downclocked the Vcore seems a bit too high and wondered if it was possible to only have higher Vcore for multipliers higher than 37...?

Has anyone mastered the offset voltage on this board?

Great news about the bios. Any release notes for it?

[Ket]

Do a full CMOS reset and redo your OC, that should fix your issue. I've found LLC level 2 to work best so you might want to try that.

[Anusha]

i wonder what the updated raid rom means. more performance? XD

[qbazdz]

Full CMOS reset? you mean by the back button, 15 sec, no PS etc.?

Tried that already and still no luck... Will keep trying though.

What sort of voltages people run @4.8GHz anyway... I though my 1.37V was pretty low anyway...

[Ket]

By full CMOS reset I mean;

Set CMOS jumper to "clear"
Remove CMOS battery
Remove PSU cord
Press power button 2-3 times (discharges capacitors)
Wait 1 min

Astonishing how people don't even know what a full CMOS reset is anymore, it makes me sad

[qbazdz]

"Jumper method" means more to me than full CMOS reset :P

[Ket]

This is a post I'll be putting in the P67 thread as well. Basically, everybody is hellbent on keeping their SB chip as cool as possible for some reason so I conducted some of my own tests and done a lot of research. Here is what you can expect to get cooling-wise depending on how much cash you throw at cooling your CPU and whats really going on with these coolers.

CoolIT ECO ALC - £45 Amazon.co.uk
Antec Kuhler H2O 620 - £46 kikatek.co.uk
Corsair H50 - £52 box.co.uk
Corsair H60 - £49.55 ILGS.co.uk (thats right, a better model is cheaper than the H50... go figure)
Corsair H70 - £73.39 component warehouse.co.uk
Corsair CWCH80 - £70.18 ILGS.co.uk (yep.. again a better model is cheaper. Methinks this pricing strategy is deliberate)
Antec Kuhler H2O 920 - £80.54 ILGS.co.uk

So, cooling-wise whats really going on here? Firstly I'm going to stress the difference in cooling between the "worst" and "best" liquid CPU coolers is 8-10c. This seems like a lot at first glance.. but is it? Well, here is how that list breaks down;

ECO ALC > Antec H2O 620 > Corsair H50 - There is nothing seperating ANY of these coolers. The minors differences (3c at best) can easily be chalked up to the type of TIM and fan the manufacturer has chosen to use. Something easily overcome with some proper TIM like MX4 alone nevermind a better fan as well.

H60 > H70 - You guessed it, theres nothing seperating these coolers either. Again (surprise, surprise) about 3c at best. Which also (again) can be put down to TIM and the fan used.

Corsair H80 (CWCH80) > Antec H2O 920 - A double phat radiator and these suckers still can't get a significant gain over any of the earlier coolers. These coolers make up the 8-10c difference.

By now you are all probably wondering what I'm getting at, its this; Anybody who owns one of these coolers before plumping out perhaps £80 on a new and "better" liquid CPU cooler instead opt to buy 1 or 2 (whichever is your preference) new fans (the coolermaster sickleflow fans are excellent, 70CFM @ low dba) and some high performance TIM like MX4. Remount your existing cooler with these upgrades and then look up some reviews and compare your new temps to the liquid cooler you had in mind as a replacement. I'm betting your existing cooler is now performing within a few degrees of that cooler you were about to splash some cash on

Some of you might think "Oh but wait! I can do the same thing to the more expensive cooler too." This is correct, but A: The more expensive cooler is likely going to already be using fan(s) equally as good as the new fan(s) you have just got meaning you won't really see any (if any) gains. Lastly, B: You would hope the more expensive cooler is also using better TIM which again means you probably wouldn't see much gain by replacing the TIM on the more expensive cooler.

Give this a go readers, I think you'll be surprised.

[Wendy]

i wonder what the updated raid rom means. more performance? XD

The SATA option ROM is now version 10.6.0.1091, which I hope fixes the problems that SATAIII SSD users were getting.
Here is the change log from Intel.
Click

[Puff Daddy]

new member here, i've been following this thread religiously and it's been very helpful. however, there is one issue i still can't solve - getting the onboard analog audio to work.

the Realtek drivers install fine and device manager indicates it to be working properly but no sound come out of the speakers, even if the Realtek sound manager indicates the audio is playing. i noticed the rear analog out can never be enabled in the realtek manager, only the front analog out.

i tried disabling the IGD in the Bios and that didn't work as it has for others who had a similar problem.

i have tried using the drivers from the Asrock site as well as the one from Realtek with no success.

Does the ATI HDMI Audio in Device manager have anything to do with it?

[Anusha]

Not having any problems here

Untitled.jpg

[Puff Daddy]

Odd. In the BIOS I have Onboard Audio Enabled and Front Panel to Auto. If Front Panel is set to Disabled, both Analog section goes away entirely. What is your driver version?

But I should still able to get sound through the Front Panel Analog and nothing plays in the speaker. With my Crossfire configuration, I rather not add a discrete soundcard to allow proper airflow around the Graphics cards.