Asrock P67 Extreme4 UEFI & Drivers thread [Extreme6 & Pro3 owners welcome]

[banwell]

What the difference between additional turbo voltage and offset? Fine tuning??

[Ket]

Turbo voltage is a option that lets the user specify how much extra voltage the CPU uses when it goes into turbo mode. It should be fine left on "Auto".

[dnottis]

Got my new B3 Pro 3 today. Its got the awful 1.80 bios on there, can I go back to 1.70 on this with a B3 board? Shouldn't be anything different with the bioses from b2 to b3 right?

[dnottis]

Guess not. Tried it, any OC causes BSOD. God I got screwed in this swap. Time for Asus board maybe.

[aoch88]

@Ket

I've updated to UEFI 1.61 and the offset mode doesn't work correctly. I used to do 47x100 on 1.60 with offset mode but with the new 1.61, it just hangs after initializing the disks. I need "fixed voltage" to get it running. How can we feedback to Asrock? I think the board is good except for crappy BIOS-es they make.

[Ket]

Guess not. Tried it, any OC causes BSOD. God I got screwed in this swap. Time for Asus board maybe.

Go through the UEFI with a fine toothed comb and give me as much detail as possible about the bugs you encounter and let me know what the latest UEFI version was the bug wasn't present in.

@Ket

I've updated to UEFI 1.61 and the offset mode doesn't work correctly. I used to do 47x100 on 1.60 with offset mode but with the new 1.61, it just hangs after initializing the disks. I need "fixed voltage" to get it running. How can we feedback to Asrock? I think the board is good except for crappy BIOS-es they make.

UEFI 1.61 is buggy. I discovered that myself with settings I knbow to be stable but IBT "failed". Went back to UEFI 1.6A and all is well. Try 1.6A.

[dnottis]

Go through the UEFI with a fine toothed comb and give me as much detail as possible about the bugs you encounter and let me know what the latest UEFI version was the bug wasn't present in.

I get a BSOD every boot now with this board. 0x003b error. I think I'm gonna just order an Asus and be done with this.

Flashed back to 1.80 and once I try overclocking it just won't even get into Windows 7 now. I still lose the option for load line calibration as soon as I change multiplier to manual (same thing b2 board was doing). Really annoying me at this point.

[aoch88]

UEFI 1.61 is buggy. I discovered that myself with settings I knbow to be stable but IBT "failed". Went back to UEFI 1.6A and all is well. Try 1.6A.

This is really a pain I also notice that after the PC sleeps (30 mins or more idle), the moment I wake it up using my keyboard, it doesn't come back up. I had to reset the whole thing. I guess dnottis is right and maybe Asus is indeed a better board overall.

[dnottis]

I don't even know what to do... anytime I raise the volts above 1.344v I get a BSOD at Windows 7. I was running 4.6 and 4.8 without an issue on the b2. It was running 4.6 earlier but kept BSODing on me during Prime.. So I went back to 1.70 bios and have been having issues ever since. I can get to 4.0 @ 1.34v and it's ok, but higher crashes. Gonna try to just reflash back to 1.80.

1.80 on the Asrock and the 1.80 shipped on the b3 board should be the same though. I've got the logo back on there when I flashed from 1.80 off the website earlier.

[aoch88]

Are you on offset or fixed Vcore mode? Perhaps you can try using fixed Vcore to see if it helps. I'm looking at Asus Sabertooth now. M4E would be a good choice but it's way toooooooooo expensive.

[papatsonis]

..
papatsonis; try using offset mode with LLC level 2.
..

It won't work that way, as long as you cant alter the VIDs requested from CPU, the gap between 1c VID and 4c/8t VID will always be the same (0.07v in my case), "lowering" LLC lvl will only make it worse.. the voltage with LLC2 will be rock-steady in both cases (1c, 4c) so if you use the minimum needed voltage for 4c load stability.. you'll likely have a nice BSOD if you stress it 1thread. In fact the solution to this is the exact opposite.., i had to use LLC5, the reason is simple, when the cpu is loaded 1c, besides the lower VID, the vrm load is low, so the Vcore delta 1c idle-load, is much smaller than the Vcore Delta 4c/8t idle-load. an arithmetic example
1c VID 1.2710
2c VID 1.3060
4c VID 1.3461
offset -0.05volt LLC5
-----------
1c Load, (with added offset, 1.2210 VID), ~1.172v
2c Load, ( >> 1.2560 VID), ~1.184v
4c Load, ( >> 1,2961 VID), ~1.20v

[Ket]

This is really a pain I also notice that after the PC sleeps (30 mins or more idle), the moment I wake it up using my keyboard, it doesn't come back up. I had to reset the whole thing. I guess dnottis is right and maybe Asus is indeed a better board overall.

UEFI 1.61 is BETA, can't expect all the non finished UEFIs to be perfect Just flash in 1.6A. The bug you describe also isn't Asrocks fault its a fundamental firmware bug in the P67 and intel are working to fix it.

I don't even know what to do... anytime I raise the volts above 1.344v I get a BSOD at Windows 7. I was running 4.6 and 4.8 without an issue on the b2. It was running 4.6 earlier but kept BSODing on me during Prime.. So I went back to 1.70 bios and have been having issues ever since. I can get to 4.0 @ 1.34v and it's ok, but higher crashes. Gonna try to just reflash back to 1.80.

1.80 on the Asrock and the 1.80 shipped on the b3 board should be the same though. I've got the logo back on there when I flashed from 1.80 off the website earlier.

I would love to get my hands on your board to see what its doing. Right now I would suggest ripping the board apart and lapping the HS bases and replacing the stock TIM, then replacing the mounting mechanism with some screws, washers and nuts. It sounds like something on the board is overheating. Particularly questionable would be the MOSFET area. For what its worth though 99% of people seem to be regretting swapping their B2 board for a B3 regardless of the manufacturer. They seem to be having more problems and I'm yet to see a B3 that clocks better than a B2. Lastly, I would download the first release UEFI (DOS version), load UEFI defaults then flash the first release to the board. I did this with my Extreme4 and a lot of weird issues went away when I re-updated the UEFI.

I can't help push through UEFI fixes for the Pro3 unless somebody gives me detailed bug descriptions

[dnottis]

Are you on offset or fixed Vcore mode? Perhaps you can try using fixed Vcore to see if it helps. I'm looking at Asus Sabertooth now. M4E would be a good choice but it's way toooooooooo expensive.

Tried offset and fixed. No love

[dnottis]

I would love to get my hands on your board to see what its doing. Right now I would suggest ripping the board apart and lapping the HS bases and replacing the stock TIM, then replacing the mounting mechanism with some screws, washers and nuts. It sounds like something on the board is overheating. Particularly questionable would be the MOSFET area. For what its worth though 99% of people seem to be regretting swapping their B2 board for a B3 regardless of the manufacturer. They seem to be having more problems and I'm yet to see a B3 that clocks better than a B2. Lastly, I would download the first release UEFI (DOS version), load UEFI defaults then flash the first release to the board. I did this with my Extreme4 and a lot of weird issues went away when I re-updated the UEFI.

I can't help push through UEFI fixes for the Pro3 unless somebody gives me detailed bug descriptions

Yea, I shut it off and unplugged power. At work now for about 6 more hours. If it's overheating issue I'm thinking it might work when I get home after it's been off and cold for a bit. I'm just wondering if the b3 shipping bios is any different than the 1.80 on their website though.

[Ket]

Its extremely unlikely theres been any changes to the UEFI for B3 boards. The changes that were made to the P67 only needed to be made on the hardware level, not software.

[Sync0r]

I can confirm that the board will do 3 x 5870's, although my PSU can not. Corsair TX 950 on order

[dnottis]

Ket -

did you have problems that were fixed by changing the tim and stuff on your board?

[Garrett]

Strange... when I choose 47 MP it will boot into Windows but freeze after a few seconds on the desktop... MP 48 and higher result in Windows not booting at all... any ideas?

[Poohead]

Strange... when I choose 47 MP it will boot into Windows but freeze after a few seconds on the desktop... MP 48 and higher result in Windows not booting at all... any ideas?

Increase core voltage and enable PLL overvolt.

[Falkentyne]

dnottis:
Did you clear the CMOS after flashing?

I had tons of problems if I tried just starting the board after a flash, and using settings without a cmos clear first.

About the B2 vs B3 swap:
Seems to be all over the map. No problems at all on my UD5. Saw some Asus people say that the stability was exactly the same, one who said he was able to LOWER the vcore by 0.015v on overclock.net, and many saying less stable. But I don't think anyone who said it was less stable, even tried measuring voltages with a DMM. You'd need readings from both boards to make it mean anything.

But I saw these same complaints when people were just flashing different bioses on the B2 boards. The hardware is identical besides the PCH, right?

What someone really needs to do is take a DMM to the board and smack the vcore readings and see if the boards are really outputting the same voltages...

[Falkentyne]

Ket,
Why don't you apply for a job at Asrock? You'd probably be accepted right away and then you wouldn't even have to make modded bioses as you'd be making the official or betas....plus you'd get bios tools...

[dnottis]

dnottis:
Did you clear the CMOS after flashing?

I had tons of problems if I tried just starting the board after a flash, and using settings without a cmos clear first.

About the B2 vs B3 swap:
Seems to be all over the map. No problems at all on my UD5. Saw some Asus people say that the stability was exactly the same, one who said he was able to LOWER the vcore by 0.015v on overclock.net, and many saying less stable. But I don't think anyone who said it was less stable, even tried measuring voltages with a DMM. You'd need readings from both boards to make it mean anything.

But I saw these same complaints when people were just flashing different bioses on the B2 boards. The hardware is identical besides the PCH, right?

What someone really needs to do is take a DMM to the board and smack the vcore readings and see if the boards are really outputting the same voltages...

I did. I used the instant flash feature. Going to try maybe the one in Windows. Used that with my B2 and never had a problem. I left the power unplugged and out of the PSU while I've been at work so it should completely discharge it, gonna reset cmos when I get home. If I still have issues, I'll reflash it again, clear everything and see. It was running ok with the shipped B3 just crapped out during prime. now I can't even get into windows with the volts up over 1.35v... I'm hoping the shipped bios isnt differnt from the one online... otherwise I'm gonna need to get that bios again...

[Ket]

Ket -

did you have problems that were fixed by changing the tim and stuff on your board?

Not direct stability problems, but by replacing the TIM, lapping the PCH heatsink and replacinf all pushpins with screws and nuts thermal efficiency went right up and even under load the board doesn't get very hot in critical areas where it used to get quite toasty. Well worth the time spent IMO.

dnottis:
Did you clear the CMOS after flashing?

I had tons of problems if I tried just starting the board after a flash, and using settings without a cmos clear first.

About the B2 vs B3 swap:
Seems to be all over the map. No problems at all on my UD5. Saw some Asus people say that the stability was exactly the same, one who said he was able to LOWER the vcore by 0.015v on overclock.net, and many saying less stable. But I don't think anyone who said it was less stable, even tried measuring voltages with a DMM. You'd need readings from both boards to make it mean anything.

But I saw these same complaints when people were just flashing different bioses on the B2 boards. The hardware is identical besides the PCH, right?

What someone really needs to do is take a DMM to the board and smack the vcore readings and see if the boards are really outputting the same voltages...

The only change with the B3 revision is something done to a electronic gate in the chipset. I think intel may of removed the gate completely or made the gate thicker which might lead to other issues.

Ket,
Why don't you apply for a job at Asrock? You'd probably be accepted right away and then you wouldn't even have to make modded bioses as you'd be making the official or betas....plus you'd get bios tools...

I can't make modded UEFIs.. no tools available for modders. all UEFIs on the first page are what I've aquired from various sources. If Asrock want to employ me to keep them in check with bug reports and so on, then I'm not stopping them from approaching me.

I did. I used the instant flash feature. Going to try maybe the one in Windows. Used that with my B2 and never had a problem. I left the power unplugged and out of the PSU while I've been at work so it should completely discharge it, gonna reset cmos when I get home. If I still have issues, I'll reflash it again, clear everything and see. It was running ok with the shipped B3 just crapped out during prime. now I can't even get into windows with the volts up over 1.35v... I'm hoping the shipped bios isnt differnt from the one online... otherwise I'm gonna need to get that bios again...

To make sure everything goes smootly as possible try these steps;

Remove power cord, CMOS battery, and set CMOS jumper to "clear"
Press power button 2-3 times
Flash to the "first release" UEFI. Don't use instant update or Windoze flasher - download and run the DOS executable.
Load UEFI defaults, save and exit
re-update to what you found to be the best UEFI

[dnottis]

Well the board is just bad. The board can't hold voltage, the back VRMs dont even get warm so I'm assuming they aren't even working. I set 1.40v for fun and when it got to Windows CPUZ was showing 1.408 dropping to 1.32v lmao. I get BSODs 0x0101 basically saying needs more vcore.

I stripped the VRM coolers, mounted with screws and used some Ceramique... same thing. Freaking huge waste of time. Wish I hadn't mailed back my b2 that worked well.

Got screwed. Newegg will get an earful tomorrow and they better send me out a new one asap.

[aoch88]

@Ket

I've reverted to BIOS 1.60A and things are much better now I'm a happier man and I hope the board won't give me funny problems again.