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

Hey guys, I recently purchased an Asrock P67 Extreme6 and a Core-i5 2500K as part of a new system build a while back. I've been lurking in the thread for weeks (I've read every single page!)

My goal:
I want fixed OC @ 4.7ghz (or more), with all power-saving features off. I will not be browsing, checking e-mail, or anything else on the build; it is ONLY for performance. I'm posting this on my separate 24/7 machine. The Asrock P67 Extreme will be used for performance and ONLY performance.

Anyway, here's my problem:
I can't get stable/boot past 44x multi.

These settings work beautifully:
- All C-settings off
CPU ratio: 44
PLL Overvolt: Disabled
Speedstep: Disabled
Turbo Boost power limit: 250/250/1
Add. Turbo Voltage: Auto (can't disable)
Core Current Limit: 250
BCLK: 100.0
Spread Spectrum: Auto (should I disable this?)
Power Saving: Disabled
CPU Core Voltage: Fixed/1.375v
LLC: Level 3
(all of the following @ AUTO: here are reported values)
DRAM: 1.635v
PCH: 1.059v
PLL: 1.832v
VTT: 1.142v
VCCSA: 0.925v

These settings 100% stable in IBT, Prime95 (blend). BTW, temps are a non-issue; load barely touches 60C on medium fan setting w/ Tuniq Tower 120 Extreme.

- @ multi:45, I eventually fail Prime95 (blend)... Oddly, upping vcore does not seem to help (should I try the other voltages?).
- @ multi:46, I get black screen with flashing cursor after POST. Upping the vcore to 1.4 makes no difference. What the heck is going on?
What settings do you guys recommend?

My specs:
Core-i5 2500K (Tuniq Tower 120 Extreme)
Asrock P67 Extreme6 (B3)
2x2gb DDR-2133 on XMP profile (G.Skill F3-17000CL9D-4GBXL)
XFX Black Edition 750W
Windows 7 64-bit
-----------
Secondary issue (nowhere near as important)

I have a Toshiba CDRW/DVD drive installed on SATA port 2 (just above the ports 0/1), but for some reason it reads discs VERY VERY slowly. It also causes Intel Rapid Storage to hang for several minutes when Windows loads up. The drive works perfectly in other builds.
This sounds a little like the infamous SATA port issue, but my MB is a B3 model...so what's up?

Thanks Ket. That Rocketraid card, the 2640x4, caught my attention because it was not very expensive. Sub $100 if I remember correctly. However, I decided to pass because one cannot specify the stripe size on a stripe array. That was my reasoning for wanting a seperate raid card. I wonder if it's more the card itself than the P67 chipset or E4 board. Unfortunately, I don't have a raid card to test. :shrug:

Quote:

---

Originally Posted by Sailindawg

Thanks Ket. That Rocketraid card, the 2640x4, caught my attention because it was not very expensive. Sub $100 if I remember correctly. However, I decided to pass because one cannot specify the stripe size on a stripe array. That was my reasoning for wanting a seperate raid card. I wonder if it's more the card itself than the P67 chipset or E4 board. Unfortunately, I don't have a raid card to test. :shrug:

---

Yes, the stripe is fixed at 64kB.
The proper reason for buying this kind of the devices over the integrated controllers should be:
- you've got access to SMART attributes of each drive forming every array/logical volume.
- you can define separate arrays of different raid levels on the same set of disks.
- You got decent performance with raid-5, especially in the volume-degradated scenario (intel ICH10 is a joke).
- portability
- spare drives and spare areas
- mail notifications of errors and warnings
Do not even bother with buying a sub300$ hardware/fake RAID card without own cache for decent random performance. Only sequential.

As of the argument:
"if it was a problem with either Asrocks boards or the P67 chip itself there would be many people complaining about it."
Please try to find in the Internet any and I mean any person with Asrock P67 board using hardware raid card with or without success. Good luck. This is simply a niche market. Nobody reports because (almost)nobody uses.
Please try to find anyone using Asrock P67 with hardware PCI or PCIe DVBs(2) SAT card. Good luck too. The same applies.

Quote:

---

Originally Posted by blacksun1942

Hey guys, I recently purchased an Asrock P67 Extreme6 and a Core-i5 2500K as part of a new system build a while back. I've been lurking in the thread for weeks (I've read every single page!)

My goal:
I want fixed OC @ 4.7ghz (or more), with all power-saving features off. I will not be browsing, checking e-mail, or anything else on the build; it is ONLY for performance. I'm posting this on my separate 24/7 machine. The Asrock P67 Extreme will be used for performance and ONLY performance.

Anyway, here's my problem:
I can't get stable/boot past 44x multi.

These settings work beautifully:
- All C-settings off
CPU ratio: 44
PLL Overvolt: Disabled

---

Highlighted is your problem. Enable over volt and unlock higher multis. Its what its for...

Check here for some reading

http://www.clunk.org.uk/forums/overc...beginners.html

Quote:

---

Originally Posted by Strike@ps3

okay .. thanks for the quick reply. hmm.. Im using third party cooler... the the load temp were good.. using the OCCT and prime95 to test.. . never even went past 60c... mostly in the range of 54-58c for all for core ..

may I know which temp monitor software as of current is able to give the correct temp of these sandybridge..

I realized that the realtemp 3.60 was like outdated.. last year oct.

---

I use CoreTemp 0.99.8.

Quote:

---

Originally Posted by blacksun1942

Hey guys, I recently purchased an Asrock P67 Extreme6 and a Core-i5 2500K as part of a new system build a while back. I've been lurking in the thread for weeks (I've read every single page!)

My goal:
I want fixed OC @ 4.7ghz (or more), with all power-saving features off. I will not be browsing, checking e-mail, or anything else on the build; it is ONLY for performance. I'm posting this on my separate 24/7 machine. The Asrock P67 Extreme will be used for performance and ONLY performance.

Anyway, here's my problem:
I can't get stable/boot past 44x multi.

These settings work beautifully:
- All C-settings off
CPU ratio: 44
PLL Overvolt: Disabled
Speedstep: Disabled
Turbo Boost power limit: 250/250/1
Add. Turbo Voltage: Auto (can't disable)
Core Current Limit: 250
BCLK: 100.0
Spread Spectrum: Auto (should I disable this?)
Power Saving: Disabled
CPU Core Voltage: Fixed/1.375v
LLC: Level 3
(all of the following @ AUTO: here are reported values)
DRAM: 1.635v
PCH: 1.059v
PLL: 1.832v
VTT: 1.142v
VCCSA: 0.925v

These settings 100% stable in IBT, Prime95 (blend). BTW, temps are a non-issue; load barely touches 60C on medium fan setting w/ Tuniq Tower 120 Extreme.

- @ multi:45, I eventually fail Prime95 (blend)... Oddly, upping vcore does not seem to help (should I try the other voltages?).
- @ multi:46, I get black screen with flashing cursor after POST. Upping the vcore to 1.4 makes no difference. What the heck is going on?
What settings do you guys recommend?

My specs:
Core-i5 2500K (Tuniq Tower 120 Extreme)
Asrock P67 Extreme6 (B3)
2x2gb DDR-2133 on XMP profile (G.Skill F3-17000CL9D-4GBXL)
XFX Black Edition 750W
Windows 7 64-bit
-----------
Secondary issue (nowhere near as important)

I have a Toshiba CDRW/DVD drive installed on SATA port 2 (just above the ports 0/1), but for some reason it reads discs VERY VERY slowly. It also causes Intel Rapid Storage to hang for several minutes when Windows loads up. The drive works perfectly in other builds.
This sounds a little like the infamous SATA port issue, but my MB is a B3 model...so what's up?

---

- Enable PLL overvoltage
- Set LLC level 2
- Change PLL to 1.85v
- Change VCCSA to 1v
- VTT doesn't need to be that high
- Dial in 1.36v (fixed mode)

Now try booting 46x multi. You may just have a bum 2500k.

As for the other issue, try a different SATA cable and/or optical drive. Slow disc reading is ually the cause of a dirty laser, dying drive, or a bad cable.

Quote:

---

Originally Posted by piquadrat

As of the argument:
"if it was a problem with either Asrocks boards or the P67 chip itself there would be many people complaining about it."
Please try to find in the Internet any and I mean any person with Asrock P67 board using hardware raid card with or without success. Good luck. This is simply a niche market. Nobody reports because (almost)nobody uses.
Please try to find anyone using Asrock P67 with hardware PCI or PCIe DVBs(2) SAT card. Good luck too. The same applies.

---

1. It wasn't a argument.
2. Don't act like a :banana::banana::banana::banana: toward me, I'm not a Asrock rep just a mediatary passing on info I get from my contacts.
3. RAID cards being used isn't that abnormal, I'd think any decent RAID card (or most at least) and not the cheapest you can possibly find would work in most boards. If hardware is cheap, its usually cheap for a reason (lack of QC, little to no driver updates, little to no tech support, etc).

Why should vccsa be raised as high as 1v?

Quote:

---

Originally Posted by Ket

1. It wasn't a argument.
2. Don't act like a :banana::banana::banana::banana: toward me, I'm not a Asrock rep just a mediatary passing on info I get from my contacts.
3. RAID cards being used isn't that abnormal, I'd think any decent RAID card (or most at least) and not the cheapest you can possibly find would work in most boards. If hardware is cheap, its usually cheap for a reason (lack of QC, little to no driver updates, little to no tech support, etc).

---

Maybe that wasn't argument but sounded like that. Don't take it personally. I appreciate your work here and following the thread for a long time.
But we should not miss the point here. It seems that Asrock does not support raid controllers on their boards (cheap or expensive).
That's unusual. Did you hear about such precedence in the past from other mobo makers?
And actually why that isn't supported? And what that means for me as a customer, I can rma the board if I witness compatibility problems or can't?
That is very unusual. The board is equipped with 4x PCIe slot which do not loose bandwidth unlike competitors (Asus board requires draconian feature cutting to get 4x slot). And at the same time Asrock limiting that functionality only to graphic cards.
Definitely Asrock does not gain better reputation among professionals that way. It seems they think gamers hearts are sufficient.
Strange, that's all.

Quote:

---

Originally Posted by Ket

- Enable PLL overvoltage
- Set LLC level 2
- Change PLL to 1.85v
- Change VCCSA to 1v
- VTT doesn't need to be that high
- Dial in 1.36v (fixed mode)

Now try booting 46x multi. You may just have a bum 2500k.

As for the other issue, try a different SATA cable and/or optical drive. Slow disc reading is ually the cause of a dirty laser, dying drive, or a bad cable.

---

Thanks for the advice.
VTT was on "Auto"; I assume it was set that high so that it would stay within 0.5v of the DRAM (1.635v)....wouldn't it be "bad" if I manually changed it to be lower? Also does "Spread Spectrum" matter?

EDIT:
Tried your recommendations; black screen after POST @ 46x. :(:shrug:
45x with the same settings boots into Windows; testing stability now...
EDIT: 45x with above recommended settings eventually failed in Prime95 blend... :(

Quote:

---

Originally Posted by blacksun1942

Thanks for the advice.
VTT was on "Auto"; I assume it was set that high so that it would stay within 0.5v of the DRAM (1.635v)....wouldn't it be "bad" if I manually changed it to be lower? Also does "Spread Spectrum" matter?

EDIT:
Tried your recommendations; black screen after POST @ 46x. :(:shrug:
45x with the same settings boots into Windows; testing stability now...
EDIT: 45x with above recommended settings eventually failed in Prime95 blend... :(

---

If you get a black screen after POST, then either PLL overvoltage isn't enabled, your BIOS is corrupted or bugged and won't enable it, or you're not using a D2 stepping CPU.

Quote:

---

Originally Posted by Falkentyne

If you get a black screen after POST, then either PLL overvoltage isn't enabled, your BIOS is corrupted or bugged and won't enable it, or you're not using a D2 stepping CPU.

---

Nope, that's not correct.
There are many reports of this MB posting to black screen or black screen with flashing cursor when OC is too high. The system will automatically reboot, which allows me to simply go back into BIOS and readjust to enter Windows again.

My CPU is listed as "D2" stepping in CPUZ. My BIOS is not corrupt (I have reflashed just in case). PLL Overvoltage is enabled.
BTW: Currently retesting @ wimpy 44x multi; stable so far on Prime95 blend...

Quote:

---

Originally Posted by piquadrat

Maybe that wasn't argument but sounded like that. Don't take it personally. I appreciate your work here and following the thread for a long time.
But we should not miss the point here. It seems that Asrock does not support raid controllers on their boards (cheap or expensive).
That's unusual. Did you hear about such precedence in the past from other mobo makers?
And actually why that isn't supported? And what that means for me as a customer, I can rma the board if I witness compatibility problems or can't?
That is very unusual. The board is equipped with 4x PCIe slot which do not loose bandwidth unlike competitors (Asus board requires draconian feature cutting to get 4x slot). And at the same time Asrock limiting that functionality only to graphic cards.
Definitely Asrock does not gain better reputation among professionals that way. It seems they think gamers hearts are sufficient.
Strange, that's all.

---

Actually I can recall plenty of RAID card problems on Asus and Gigabyte boards not all that long ago (I think MSI too), about a year or so I think it was. in most circumstances I've heard of RAID cards that had problems worked fine after a firmware and / or driver update for the RAID card. Primarily, I'd say its the responsibility of the RAID card manufacturers but if they can't be arsed with the proper R&D, then what can anyone do :shrug:

Quote:

---

Originally Posted by blacksun1942

Thanks for the advice.
VTT was on "Auto"; I assume it was set that high so that it would stay within 0.5v of the DRAM (1.635v)....wouldn't it be "bad" if I manually changed it to be lower? Also does "Spread Spectrum" matter?

EDIT:
Tried your recommendations; black screen after POST @ 46x. :(:shrug:
45x with the same settings boots into Windows; testing stability now...
EDIT: 45x with above recommended settings eventually failed in Prime95 blend... :(

---

Yup.. I'd say likely a bum 2500k. Spread Spectrum should make a difference, particularly when the system is running stock. Best left disabled when OCing though. Going a little OT here but as most are unsure what Spread Spectrum actually does heres a little tech insight for everybody :D

SSCG as its referred to (Spread Spectrum Clock Generator) isolates signals helping to prevent Electro Magnetic Interference (EMI), or as some may simply call it, electronic crosstalk. Spread Spectrum isn't just limited to electronics either it also has useful acoustical purposes (which I'm unfamiliar with, a telecommunications engineer may be able to answer that part). Theoretically (to my mind anyway, I've never seen any real proof just white paper stuff) Spread Spectrum could help with a clearer audio output with on-board audio CODECs wired to one of the PCI-E lanes thanks to its ability to isolate crosstalk and "spread out" its signal. In practice, you are probably better off putting a small heatsink on the audio IC as that will prevent EMI more reliably.

The kicker with Spread Spectrum comes if you are a overcolcker, while that signal is spread out to help prevent interference, it means (I think, going back a few years in my memory now) that with the signal not being as focused it can reduce stability and overclocks. As I remember it, Spread Spectrum was introoduced in PCs back in the DDR1 days to help with memory Sine Waves in a attempt to get super high memory OCs stable. Did it ever really help? You'd have to ask somebody who was gunning for a WR back then and using a DFI Lanparty board. In modern boards, the biggest problem for OCers is with Spread Spectrum left enabled while trying to OC, it can cause data/clock misalignments.

Probably a good modern example of Spread Spectrum at work in electronic devices would be the use of a wireless adapter surrounded by typical household electricals like TVs or mobile phones. Most modern electricals have some kind of active Spread Spectrum built into their design to limit electrical and magnetic field strength, which when using a wireless adapter would have the visible effect of the adapter getting a better signal.

So in summary for those that didn't follow all that here it is bullet style;

- Spread Spectrum helps reduce EMI
- In PCs with audio going across the PCI-E lanes theoretically it may help with a cleaner audio signal
- Spread Spectrum is all benefits if a PC is left at stock
- Disable Spread Spectrum when OCing as it can lead to clock frequency / data misalignment; aka instability with OCs and data corruption.

I did offer a more breif and probably clearer explanation of Spread Spectrum here on XS a while ago, but I can't find it now plus I'm tired so my apologies if that new explanation is too broad.

As for VTT.. I have mine set to 1.1v, no problems. If you thinking about feeding the IMC to keep high memory clocks stable thats the PLL voltage. You can happily run 1.75v PLL if all you are looking for is 1600MHz and good timings, but going to perhaps 1866, and deffinately 2133+, you will need higher PLL voltages.

Quote:

---

Originally Posted by blacksun1942

Nope, that's not correct.
There are many reports of this MB posting to black screen or black screen with flashing cursor when OC is too high. The system will automatically reboot, which allows me to simply go back into BIOS and readjust to enter Windows again.

My CPU is listed as "D2" stepping in CPUZ. My BIOS is not corrupt (I have reflashed just in case). PLL Overvoltage is enabled.
BTW: Currently retesting @ wimpy 44x multi; stable so far on Prime95 blend...

---

Blacksun:
This is EXACTLY what happens on my gigabyte board when the OC is too high with PLL overvolt disabled.

System hangs on black screen with a flashing cursor at the top left, or just immediately reboots and says "OC failed (without a black screen).

What you described is exactly what happens if PLL Overvoltage is disabled, when you pass the chip's multiplier wall. Unless you need it enabled for 4.5 ghz already (but in that case, you would need it enabled just to pass 4.2 ghz, if the CPU's absolute wall were 4.5 ghz).

If you can boot up at 4.5 ghz and 4.6 ghz gives you a hard complete wall, even with 1.5v vcore, 1.1 VTT, etc, then PLL Overvoltage is not working.

Can you actually verify that PLL overvoltage is working?
If you disable PLL Overvoltage at 4.5 ghz, you can boot into windows, right?

But at 4.6 ghz you get a hard wall...that means there's a bug with PLL Overvoltage....
Or you could just have the worst CPU in the world that won't work at all even with PLL Overvoltage....
(I've only seen D1 steppings that do that).

(Ket please correct me if I'm wrong).

@Ket:
What does increasing the system agent vccsa voltage do? I thought it's best to leave it at the default....

Notching it up one just helps keep things stable really. I forget the specifics and I may of already posted this in this thread but;

DRAM Voltage:
[Intel Spec. Range: 1.5v - 1.58v]
Typical +10% electrical tolerence = 1.65v MAX Safe

PCH Voltage:
[Intel Spec. Range: Unknown]
Nominal Voltage: 1.059v
Typical +10% electrical tolerence = 1.16v

CPU PLL Voltage (VCCPL):
[Intel Spec. Range: 1.7v - 1.9v]
+5% electrical tolerence = 1.92v MAX safe
(no known prolonged tests) +10% tolerence = 2.01v MAX

VTT Voltage (VCCIO):
[Intel Spec. Range: 1.05v]
+5% electrical tolerence = 1.1v MAX safe
(no known prolonged tests) +10% tolerence = 1.15v MAX

VCCSA Voltage:
[Intel Spec. Range: 0.88v - 0.97v]
+10% electrical tolerence = (0.88) 0.97v = Do not exceed 1.016v!
(no known prolonged tests) +10% tolerence = (0.97) 1.067v = Do not exceed 1.1v!

Load Line Calibration (LLC):
It would seem LLC works in a slightly bizarre way on Sandy Bridge, causing 0x101 BSODs
if it's potency is too high. To remedy this simply set a more relaxed LLC option. I
recommend setting a mid-level LLC setting.

whelp first round of OCing on my new extreme6/2600K setup got me 4.8Ghz at 1.35v with lvl 3 LLC and nothing turned off except speedstep. I'd imagine I should be pretty happy with that. :up:

I have a P67 Extreme 6 and a 2500k OC'd @ 4.6Ghz. Everything works great except one thing that keeps bothering me. When I turn on the computer it will turn on for 2 seconds then power off, remain off for about 5 sec then power itself back on and then run through the POST process and boot to windows 7 like normal. This happens every time I shut my computer down and I have to turn it on again. Is this normal behavior for this board? I'm a little concerned with the wear on the parts.

Quote:

---

Originally Posted by Ket

I'd not pay too much attention to the UEFI CPU temp reading, I suspect its uncalibrated or it measures actual core temp, and not IHS surface temp.

---

Indeed CPU temp in BIOS are usually off but to my complete surprise Extreme4 is the first Intel MB that shows me CPU temp the same as hottest core shown by ThrottleStop/RealTemp. Always. :shocked:
http://img502.imageshack.us/img502/2...esaving.th.png

Quote:

---

Originally Posted by Strike@ps3

I realized that the realtemp 3.60 was like outdated.. last year oct.

---

Try latest beta. ;)

is PLL overvolt ok to use for 24/7 setup? Also, for the sake of minimizing degradation what core voltage should I be trying to stay under?

Quote:

---

Originally Posted by Sailindawg

Ket, can you confirm this with Asrock? Is this a BIOS issue or a limitation of Sandy Bridge because of the on die PCIE controller?

---

I get the impression that this is a firmware issue are some controllers work fine while others even though there from the same manufacture don't.

Whether it's an issue with the controller or MB firmware I don't know.

EDIT:
There's a couple of threads floating around for the adaptec cards. Below are the ones I can find again.
http://hardforum.com/showthread.php?t=1600521
http://forums.tweaktown.com/asrock/4...tml#post389351
http://forums.tweaktown.com/asrock/4...tml#post388014

Quote:

---

Originally Posted by burebista

Try latest beta. ;)

---

thanks... it seems like coretemp , hwmontior and this realtemp are reporting the same temp of the 4 cores... since it already been mentioned, my temp is fine.. then i shall not bother too much abt it...

Quote:

---

Originally Posted by dworks

I have a P67 Extreme 6 and a 2500k OC'd @ 4.6Ghz. Everything works great except one thing that keeps bothering me. When I turn on the computer it will turn on for 2 seconds then power off, remain off for about 5 sec then power itself back on and then run through the POST process and boot to windows 7 like normal. This happens every time I shut my computer down and I have to turn it on again. Is this normal behavior for this board? I'm a little concerned with the wear on the parts.

---

You need to update the UEFI, that bug was fixed with UEFI 1.6A onward.

Quote:

---

Originally Posted by burebista

Indeed CPU temp in BIOS are usually off but to my complete surprise Extreme4 is the first Intel MB that shows me CPU temp the same as hottest core shown by ThrottleStop/RealTemp. Always. :shocked:
http://img502.imageshack.us/img502/2...esaving.th.png


Try latest beta. ;)

---

You have a better IHS application than I do. Theres a maximum delta of 10c between hottest and coolest cores on my 2500k and even the average is 6-7c hotter between coolest and hottest cores and thats only idle :down: at load it can increase to as much as 13c. Thats why I think its just TIM used under the IHS and why I'm so very tempted to take a razor blade and set it on my 2500k.

Quote:

---

Originally Posted by vafan13

is PLL overvolt ok to use for 24/7 setup? Also, for the sake of minimizing degradation what core voltage should I be trying to stay under?

---

For 24/7 usage, you should be looking at around 1.38v max.

Quote:

---

Originally Posted by Ket

You need to update the UEFI, that bug was fixed with UEFI 1.6A onward.

---

He has got Extreme6 so he need to flash UEFI 1.50A and higher.

http://www.pctreiber.net/asrock-bios...69&sortby=date

Quote:

---

Originally Posted by Ket

So in summary for those that didn't follow all that here it is bullet style;

- Spread Spectrum helps reduce EMI
- In PCs with audio going across the PCI-E lanes theoretically it may help with a cleaner audio signal
- Spread Spectrum is all benefits if a PC is left at stock
- Disable Spread Spectrum when OCing as it can lead to clock frequency / data misalignment; aka instability with OCs and data corruption.

---

Yeah I was having sound issues and you suggested the heatsink idea (never tried it though).

I played with stock & overclock setting and on/off spectrum spread and it made no noticeable difference. I may just have picky speakers but it also came over the headphones too.

In the end I grabbed a Xonar DX and problem solved. Crystal clear audio.

BTW the issue was still present on my P45 Asus Deluxe but to a lesser a degree - but so what the clock speeds!

Testing its easy - run Maxon Cinebench rendering test - the 1st one with the cars and you can actaully hear a sound (not unlike car engines!) low in the background. Totally silent now with a offboard audio solution.

I've never tried the onboard audio with my Extreme4, but instinct tells me I wouldn't have any issues, and if I did, I would stick a aluminium or copper HS on the audio CODEC to block most of the EMI. Its just something to be expected with onboard audio, though the quality and age of speakers also plays quite a big role. I've used the same Logitech X540 5.1 setup for a few years now, never had any issues even before I got my Xonar and had to make do with onboard audio.

Update:
Prime (blend) stable @ 45x.
I only mentioned briefly that I was running my RAM @ XMP.
F3-17000CL9D-4GBXL (2x2gb sticks)
Well, I had thought the RAM was stable but it wasn't. Switching it back to 1600mhz (auto), and the system is perfect @ 45x with 1.36v (I can't believe I was trying 1.4v earlier for a mere 45x). The vcore might even work lower...
Very good info on spread spectrum btw. I'd read about it elsewhere but I don't think it ever made much sense to me until now. :up:

So now I have two problems:
- I am fine staying @ 45x, but I want to be able to run my RAM at XMP settings. G.Skill "guarantees" those settings AFAIK, but XMP causes instability when the system is OCed @45x. I've tried various combinations of voltages without luck so far. Ket, your post on voltage ranges is invaluable for me (you did post that earlier). I was also thinking that high VTT was better for RAM than PLL, very good info. Any other recommendations? :(

- I do indeed have a 45x wall; attempting to boot @ 46x causes black screen after POST. Falkentyne, I was dismissive when it came to your idea about PLL Overvolt, but now I'm definitely open to the possibility. For the record, PLL overvolt on or off makes no difference up to 45x (can boot into Windows just fine, system stable without it). Just can't boot to 46x. Seeing that other people have no problem with 48x at low voltages, I'm left scratching my head here... It's weird because 45x is Prime & IBT stable...why would such a small jump in multi be impossible? :confused:

I'm mighty tempted to test out the Fatal1ty's onboard audio though. What's the THX studio pro for? But I'm still lovin my Xifi xtreme gamer even if creative drivers suck.

Quote:

---

Originally Posted by Ket

You need to update the UEFI, that bug was fixed with UEFI 1.6A onward.



You have a better IHS application than I do. Theres a maximum delta of 10c between hottest and coolest cores on my 2500k and even the average is 6-7c hotter between coolest and hottest cores and thats only idle :down: at load it can increase to as much as 13c. Thats why I think its just TIM used under the IHS and why I'm so very tempted to take a razor blade and set it on my 2500k.



For 24/7 usage, you should be looking at around 1.38v max.

---

Thanks. It works.

Quote:

---

Originally Posted by dworks

I have a P67 Extreme 6 and a 2500k OC'd @ 4.6Ghz. Everything works great except one thing that keeps bothering me. When I turn on the computer it will turn on for 2 seconds then power off, remain off for about 5 sec then power itself back on and then run through the POST process and boot to windows 7 like normal. This happens every time I shut my computer down and I have to turn it on again. Is this normal behavior for this board? I'm a little concerned with the wear on the parts.

---

I too get this problem when overvoltage is enabled so i use 4.6 cause i dont like the double boot up.


Ket can you check into the extreme6 board honest all the bugs the extreme 4 had the 6 has to but seems the extreme 4 is getting more bios update is there gonna be a fix for the extreme 6 or unsure
??

Question should i mess wih the bclk lvls at all im 100% stable at 100.5 but i dont kno if this is hurting or making things better

- I never use XMP settings, always set speed timings and voltage manually and never had a problem. If you are using the WHITE DIMM slots on your board, try switching to the BLUE slots, they are better for stability with high RAM clocks.

- That sounds like a physical processor wall, 45x may be all she wrote. I can boot to Windows @ 48x but its not stable for long, I'm blaming poor IHS mounting. Still trying to confirm if 2500k CPUs just use TIM under the IHS. My intuition tells me it is just TIM as at idle I have a maximum of 10c delta between cores and at load up to 13c. If a IHS was soldered on, theres no possible way a temp delta would be that high between coolest and hottest cores.

Quote:

---

Originally Posted by blacksun1942

I was also thinking that high VTT was better for RAM than PLL, very good info. Any other recommendations? :(

---

Not sure where he got the idea that PLL is what's related to Vdimm, as there are plenty of online documents showing you were right in that the relation is between VTT and Vdimm since they are both tied to the IMC. Here's one for example: http://www.madshrimps.be/articles/ar...#axzz1KGAGQv1d

Thanks for the info once again guys.

- The RAM is indeed in the blue slots.
- Been trying all day, and so far I'm unable to get it stable @ DDR3-2133....
- Would help if I knew the exact timings it was rated for @ that speed, but different software gives me different SPD info. Can I trust HWiNFO32?
- At this point, I've considering making do with a lower speed.

...in other news, I've successfully booted into 4.8ghz (testing now)...

Oddly, this Japanese fellow's exact settings got me there:
http://ts2peca.no-ip.info/?p=604
I'm using LLC lvl2 not 1.
I did not disable Hardware Prefetcher or Adjacent Cache Line Prefetcher; I'm not sure why he claims that it will "slow you down" (according to Google translation)...

On the one hand, I *do* want to run this thing with "decent" voltage. If I don't, then my Tuniq Tower 120 Extreme was nothing but a waste of money. On the other hand, I do *not* want to kill this CPU.

I ran my old Core 2 Duo E6400 @ 3.2ghz + over-volt with EIST and C1E disabled non-stop for 4 years (bought it in summer of 2006 when I first registered for this forum) without problems. That's kind of what I'm hoping for from this build.

So far, my temps are WAAAAAAAAY lower than anything I ever got on my Core 2 Duo; that thing would idle @ ~40C, max @ 70C+ (cheap cooling). I would Prime 95 or IBT it every few months or so, and it stayed rock-solid error-free for all those years. With Tuniq Tower 120 Extreme, I'm so far <30C idle, maxing @ ~62C with 1.47v...

- Ket you said

Quote:
Originally Posted by dworks
I have a P67 Extreme 6 and a 2500k OC'd @ 4.6Ghz. Everything works great except one thing that keeps bothering me. When I turn on the computer it will turn on for 2 seconds then power off, remain off for about 5 sec then power itself back on and then run through the POST process and boot to windows 7 like normal. This happens every time I shut my computer down and I have to turn it on again. Is this normal behavior for this board? I'm a little concerned with the wear on the parts.

You need to update the UEFI, that bug was fixed with UEFI 1.6A onward.

Can you please link us to a 1.6a cause there is only 1.60 and 1.52 and they both still have the problem this is for the extreme 6 i dont know if you didnt saw that in dworks or if theres a bios i dont know about

Quote:

---

Originally Posted by blacksun1942

Thanks for the info once again guys.

- The RAM is indeed in the blue slots.
- Been trying all day, and so far I'm unable to get it stable @ DDR3-2133....
- Would help if I knew the exact timings it was rated for @ that speed, but different software gives me different SPD info. Can I trust HWiNFO32?
- At this point, I've considering making do with a lower speed.

...in other news, I've successfully booted into 4.8ghz (testing now)...

Oddly, this Japanese fellow's exact settings got me there:
http://ts2peca.no-ip.info/?p=604
I'm using LLC lvl2 not 1.
I did not disable Hardware Prefetcher or Adjacent Cache Line Prefetcher; I'm not sure why he claims that it will "slow you down" (according to Google translation)...

On the one hand, I *do* want to run this thing with "decent" voltage. If I don't, then my Tuniq Tower 120 Extreme was nothing but a waste of money. On the other hand, I do *not* want to kill this CPU.

I ran my old Core 2 Duo E6400 @ 3.2ghz + over-volt with EIST and C1E disabled non-stop for ~4 years without problems. That's kind of what I'm hoping for from this build.

---

Those settings are for a 2600k and be careful with llc 1 its very strong

Is LLC lvl2 safe for 24/7?

Honestly im not sure i just know that 1 and maybe 2 to can be just a little to powerful but then again i use offset mode so it may be a little differnt but basicly the same all LLC does is lessin the vdroop so that the voltage is closer to what you set it.

Hello everyone my account was finanly activated. Im new not to the site or anything computer related just was able to register :) Iv wanted to post here so bad because I know ket has contacts oo did I ever want to tell you that the extreme 6 had all the same problems but we arnt getting many bios updates i wonder why. Last but not least Ket you are the man thank you so much for your insight I hope i can help sharpin up the best looking 2 1155 boards extreme 4-6

Ket-

The problem dwork stated is the only thing keeping me from running 4.8 24/7 because on a cold boot up the computer will double boot.... and sometime it wont pass post at all and restart 2 -3 times before it goes into windows Bios 1.60

Other than that i cant complain these boards are amazing.

Quote:

---

Originally Posted by Nw0rb

Those settings are for a 2600k and be careful with llc 1 its very strong

---

Like I said, I'm using LLC 2. :up:

My only concern is... Will 1.47v constant (NO EIST, NO SPEEDSTEP, NO C1E, NO POWER SAVING) be safe if I want this build to last >3 years?
Max temp is now @ 67C.... (Prime95 blend)
Remember, this is NOT my 24/7 build, it is only for gaming/folding/etc, and it will likely only be in use a few days a week. :sofa:

[QUOTE=blacksun1942;4825055]Like I said, I'm using LLC 2. :up:

My only concern is... Will 1.47v constant (NO EIST, NO SPEEDSTEP, NO C1E, NO POWER SAVING) be safe if I want this build to last >3 years?
Max temp is now @ 67C.... (Prime95 blend)
Remember, this is NOT my 24/7 build, it is only for gaming/folding/etc, and it will likely only be in use a few days a week. :sofa:[/QUOT


Try using the auto oc to 4.8 and use overvoltage and see how that runs because 1.57 even for day to day use is a little high

Quote:

---

Originally Posted by Nw0rb

Try using the auto oc to 4.8 and use overvoltage and see how that runs because 1.57 even for day to day use is a little high

---

It's 1.47v, and remember this is not a day-to-day machine; it is only for performance. I will not be checking e-mails or "surfing the internets." ;) That is why speedstep, etc will remain *off* (I want every single window to load as fast as possible).

To be clear I am *NOT* going to be using Turbo OC, offset vcore, etc. My goal is for non-stop OC when the machine is powered on.

Nevertheless, I would like this system to last at least ~3 years if possible, like my old Core 2 Duo build did....

Quote:

---

Originally Posted by Poohead

Not sure where he got the idea that PLL is what's related to Vdimm, as there are plenty of online documents showing you were right in that the relation is between VTT and Vdimm since they are both tied to the IMC. Here's one for example: http://www.madshrimps.be/articles/ar...#axzz1KGAGQv1d

---

I tested myself, VTT doesn't make any difference on the Extreme4, but PLL voltage does. Can't say I have a explanation for it, just the way it is. With low PLL I'll BSOD when running 2133, up the PLL to 1.85, no problems.

Alright, which of these is likely to be easier on the cpu...

1.395v and lvl 3 LLC

or

1.375v and lvl 2 LLC

I ask because I can run 5Ghz stable at both of those (the voltage under full load is approx the same).

The one with offset....

hi there ...

not sure whether I should asking about this

my extreme4 bios is P1.60 .. date is 18/02/2011... according to cpu-z
it had not given me any prob yet..
how come i went to the asrock website .. it stated bios version 1.60 .. date is 18/32011

okay... I d/led the 8/3/2011 1.60 version from the asrock website and flash it ... when i opened the cpu-z .. it is still stated there P1.60 18/02/2011... how come ?

should I go d/l the 1.62 beta and update it ???

hi guys :)
is there anyone who tried this board for extreme oc? ss or ln2? I'm really interested in this one...or giga p67x ud3r...

Quote:

---

Originally Posted by dworks

I have a P67 Extreme 6 and a 2500k OC'd @ 4.6Ghz. Everything works great except one thing that keeps bothering me. When I turn on the computer it will turn on for 2 seconds then power off, remain off for about 5 sec then power itself back on and then run through the POST process and boot to windows 7 like normal. This happens every time I shut my computer down and I have to turn it on again. Is this normal behavior for this board? I'm a little concerned with the wear on the parts.

---

Quote:

---

Originally Posted by Nw0rb

The problem dwork stated is the only thing keeping me from running 4.8 24/7 because on a cold boot up the computer will double boot.... and sometime it wont pass post at all and restart 2 -3 times before it goes into windows Bios 1.60

---

Add me to the list of people who get this problem too. :(
EXACTLY the same here....
No 1.6A available for Extreme6; I'm using the latest- 1.60

Some say 'Asrock is crappy' but I said before buying that Asrock will be a great one for OC.
I wasn't wrong!

My Fatal1ty one gave the +1 working multiplier to the processor which was capable of 55 multi.
Maximus IV extreme + my 2600K = 5580max, 55 multi (SS)
Deluxe, Pro + my 2600K = 5560 max, 55 multi (SS)
MSI GD65 + my 2600K = ~~5300MHz max, 52 multi (SS)
Asrock Fatality + my 2600K= 5600MHz, 56 multi :O (Air cooled, Prolimatech Megahalems)

http://img824.imageshack.us/img824/2...litymax.th.jpg

I can't wait for - temperature benches today ;)

Please don't look at the SPI run time etc. (I was just testing the 56 multi :))

Quote:

---

Originally Posted by blacksun1942

Add me to the list of people who get this problem too. :(
EXACTLY the same here....
No 1.6A available for Extreme6; I'm using the latest- 1.60

---

Have you spread spectrum enable in bios guys ?

I got an Extreme 6 too and no double boot with the P1.60 bios at start. :shrug:

Quote:

---

Originally Posted by Olivon

Have you spread spectrum enable in bios guys ?

I got an Extreme 6 too and no double boot with the P1.60 bios at start. :shrug:

---

A. you dont have overvoltage enabled

B. its more of a double fan/post then a double boot

C. Its only when overvoltage is enabled and yes spectrum is disabled but that wouldnt have anything to do with it seeing it boots fine at 4.6 spectrum disabled but overvoltage disabled.

It is a problem with the board bios there hasnt been a bios fix so everyone still has this problem execept the extreme 4 which is gettin many more bios updates then us Ket must hate me i tell him hes my hero and doesnt talk to me ima cry in a corner lol

I get the double post on my Ex4. I'm running BIOS 1.6, the shipping BIOS. The double post reminds me of my X48 board. That board did it as it adjusted voltages before fully handing off to the operating system. I just assumed it was the same type of thing with P67 once you started changing voltages and manually setting voltages.

my extreme4 wasn't getting any double post .. once i switch on the com.. it will come to the logo page then start loading of the window 7 screen.

but ... the question.. I asked above is i saw the latest 1.60 bios from the asrock website date is 08/03/2011... when I opened up the cpu-z 1.57.1 .. it say bios date is 18/02/2011... so I went to d/led the bios from the website... flash it.. successfully.. but I use the cpu-z 1.57.1 .. the bios date still stated there 18/02/2011. how come ?

so far my board wasnt even giving me any prob... should I d/l the 1.62 beta and flash it.. how come it is still in beta form..

Quote:

---

Originally Posted by Nw0rb

A. you dont have overvoltage enabled

B. its more of a double fan/post then a double boot

C. Its only when overvoltage is enabled and yes spectrum is disabled but that wouldnt have anything to do with it seeing it boots fine at 4.6 spectrum disabled but overvoltage disabled.

It is a problem with the board bios there hasnt been a bios fix so everyone still has this problem execept the extreme 4 which is gettin many more bios updates then us Ket must hate me i tell him hes my hero and doesnt talk to me ima cry in a corner lol

---

If the system is stable with spread spectrum enabled, try turning it back on.

Quote:

---

Originally Posted by radier

The one with offset....

---

Thanks, I was using fixed volts

I'm able to run 5Ghz stable with +.05 offset and lvl4 LLC. Idles around 1.075v on average :up:

Quote:

---

Originally Posted by KUBA_TM

Some say 'Asrock is crappy' but I said before buying that Asrock will be a great one for OC.
I wasn't wrong!

My Fatal1ty one gave the +1 working multiplier to the processor which was capable of 55 multi.
Maximus IV extreme + my 2600K = 5580max, 55 multi (SS)
Deluxe, Pro + my 2600K = 5560 max, 55 multi (SS)
MSI GD65 + my 2600K = ~~5300MHz max, 52 multi (SS)
Asrock Fatality + my 2600K= 5600MHz, 56 multi :O (Air cooled, Prolimatech Megahalems)

http://img824.imageshack.us/img824/2...litymax.th.jpg

I can't wait for - temperature benches today ;)

Please don't look at the SPI run time etc. (I was just testing the 56 multi :))

---

interesting. I knew m4e was the worst buy for price/performance...but the fatality board has something more than the xtreme 6 or it's just about the graphic layout?

Quote:

---

Originally Posted by blacksun1942

Add me to the list of people who get this problem too. :(
EXACTLY the same here....
No 1.6A available for Extreme6; I'm using the latest- 1.60

---

Keep in mind that Pro3, Extreme4, Extreme6 or Fatal1ty have different UEFI names for the same changelog.

For e.g. 1.60A for Extreme4 is the same as 1.50A for Extreme6.

You can grab it here:
http://www.pctreiber.net/asrock-bios...69&sortby=date

Quote:

---

Originally Posted by simone18

interesting. I knew m4e was the worst buy for price/performance...but the fatality board has something more than the xtreme 6 or it's just about the graphic layout?

---

If you think about overclocking capabilities I don't know because I didn't have xtreme 6 in hands :)
Sry :) Forgot to log out from XA's account ;)

Quote:

---

Originally Posted by radier

Keep in mind that Pro3, Extreme4, Extreme6 or Fatal1ty have different UEFI names for the same changelog.

For e.g. 1.60A for Extreme4 is the same as 1.50A for Extreme6.

You can grab it here:
http://www.pctreiber.net/asrock-bios...69&sortby=date

---

That fixes the spread spectrum double boot not the ppl over voltage enabled double boot.

Quote:

---

Originally Posted by Nw0rb

A. you dont have overvoltage enabled

B. its more of a double fan/post then a double boot

C. Its only when overvoltage is enabled and yes spectrum is disabled but that wouldnt have anything to do with it seeing it boots fine at 4.6 spectrum disabled but overvoltage disabled.

It is a problem with the board bios there hasnt been a bios fix so everyone still has this problem execept the extreme 4 which is gettin many more bios updates then us Ket must hate me i tell him hes my hero and doesnt talk to me ima cry in a corner lol

---

Answer A Sir :D

Right, on my H24 profile, no PLL Overvoltage but 4.8GHz with 1.37v max load with Offset mode.
This mode works really well with nice lil' overclocking, Fixed mode seems to work better for massive overcloking though. Way less BSOD from my experience but maybe I made bad tweaks during my tests with Offset mode and high clocking :shrug:

Quote:

---

Originally Posted by Olivon

Answer A Sir :D

Right, on my H24 profile, no PLL Overvoltage but 4.8GHz with 1.37v max load with Offset mode.
This mode works really well with nice lil' overclocking, Fixed mode seems to work better for massive overcloking though. Way less BSOD from my experience but maybe I made bad tweaks during my tests with Offset mode and high clocking :shrug:

---

Nice but mine is at Multi wall at 46 so it has to be enabled to get to 4.8 i get the black screen with blinking curser without it. I have benched at 5 at 1.41 basicly i dont know whats causing it i think it could possibly be my ram preventing me or its a multi wall and i have to have it enabled

Currently im at offset +.05 which is 1.32 4.6 60 full load prime and ibt

Anyone else having or had troubles adjusting memory settings, like timings and what not? Whenever I try to tune my timings at all, my computer will boot fail til boot guard resets the settings and I'm back in UEFI. Any solutions? I have 8 gigs G. Skill sniper series 1600mhz ram sticks.

Quote:

---

Nice but mine is at Multi wall at 46 so it has to be enabled to get to 4.8 i get the black screen with blinking curser without it. I have benched at 5 at 1.41 basicly i dont know whats causing it i think it could possibly be my ram preventing me or its a multi wall and i have to have it enabled

Currently im at offset +.05 which is 1.32 4.6 60 full load prime and ibt

---

If changing your timings causes the "double boot" (basically a boot failure, then boot guard rolls back ur settings on the third one), then I think we have the same problem. I can't touch my ram settings without having my new rig fail boot. And it can't be a power issue, i've got a very good 650W PSU and I'm only running one HD6870 atm.

Quote:

---

Originally Posted by trees415

Anyone else having or had troubles adjusting memory settings, like timings and what not? Whenever I try to tune my timings at all, my computer will boot fail til boot guard resets the settings and I'm back in UEFI. Any solutions? I have 8 gigs G. Skill sniper series 1600mhz ram sticks.


If changing your timings causes the "double boot" (basically a boot failure, then boot guard rolls back ur settings on the third one), then I think we have the same problem. I can't touch my ram settings without having my new rig fail boot. And it can't be a power issue, i've got a very good 650W PSU and I'm only running one HD6870 atm.

---

weird cause i only get that if my memory is unstable with pll over voltage enabled even then tho it usually starts up on second boot only if timings are way off it will reset. Honestly its a weird boot it sounds like the video card fan starts up twice with pll overvoltage enabled with it off the video card fan starts normal

Quote:

---

Originally Posted by Nw0rb

That fixes the spread spectrum double boot not the ppl over voltage enabled double boot.

---

Maybe for you. I have Spread Spectrum disabled and CPU PLL Overvoltage enabled with Multi x50 and since UEFI 1.60A for Extreme4 dualboot is not a problem anymore.

Quote:

---

Originally Posted by radier

Maybe for you. I have Spread Spectrum disabled and CPU PLL Overvoltage enabled with Multi x50 and since UEFI 1.60A for Extreme4 dualboot is not a problem anymore.

---

Yes and as i said before its not so much a dual boot next time you boot up listen to your video card fan spin up twice.... and exactly extreme 4 we dont have the same bios and Ket has a lot to do with why you guys have a more solid board then us you still recieve bios updates you have had 2 sense our last......

Did you try 1.50A for Extreme6 ?

Link for Extreme6 UEFIs; http://www.pctreiber.net/asrock-bios...earch=Extreme6

At the very least load, save & exit UEFI defaults after the flash prior to any OC attempt.

Preferably though follow these steps, as it may help;

- Load UEFI defaults
- Download earliest DOS flashable UEFI from Asrock site (MUST be DOS)
- Flash
- Load UEFI defaults, save & exit
- Flash latest UEFI
- Clear CMOS (remove battery, power cord, set jumper to "clear", press power button a few times, wait 1 min)
- See if the UEFI bugs are still there

Quote:

---

Originally Posted by Olivon

Fixed mode seems to work better for massive overcloking though. Way less BSOD from my experience but maybe I made bad tweaks during my tests with Offset mode and high clocking :shrug:

---

My experience has been the same. I was LinX stable at 1.36v fixed (probably could go one notch lower) at 50x and even with offset hitting 1.376 I still error out of LinX, although I can run intelburntest, which seems to be a bit less demanding, just fine. Offset voltages vary more and when you OC that high and then run something like LinX that reduced voltage stability is enough to cause errors. At least that's my take on it.

in your opinion - how much wattage does the 2600K need for 4.8GHz and above

My PSU is 650W. and is powering the AsRock E4, 2600K, gtx470 (750/1800), 3HDDs, DVD RW, 5-6 fans and nothing more special....

Quote:

---

Originally Posted by Ket

Link for Extreme6 UEFIs; http://www.pctreiber.net/asrock-bios...earch=Extreme6

At the very least load, save & exit UEFI defaults after the flash prior to any OC attempt.

Preferably though follow these steps, as it may help;

- Load UEFI defaults
- Download earliest DOS flashable UEFI from Asrock site (MUST be DOS)
- Flash
- Load UEFI defaults, save & exit
- Flash latest UEFI
- Clear CMOS (remove battery, power cord, set jumper to "clear", press power button a few times, wait 1 min)
- See if the UEFI bugs are still there

---

Did this and it still gives me the double fan spin up at start its like the video card fan spins up fast lowers fast then spins up fast again and this only happends when over voltage is enabled and my best friend has a extreme6 as well with the same problems except he can do 4.8 without over voltage and i cant

@radier

Have you got any solution for your Arctic Freezer 13 cooler having lowest speed as 1700 RPM?

I have a friend with Asrock P67 Extreme4 and he has exactly the same problem with a Zalman 9700NT nForce cooler. Lowest rotational speed is 1700 RPM.

I was sent beta UEFI 1.62A from ASRock support.
http://www.megaupload.com/?d=2OP2XBMX

With this version my problem is partially fixed. On level1 fan is spinning at full speed which is a bug, but on Level2 it now slows down to ~1500RPM which is a step forward.
Also it is a little bit more stable with 2 core stress and offset voltage. At least this is what I have observed.

Please send info to ASRock Support about fan issue that your friend have because I had an impression that they don't belive this problem really exist.

Quote:

---

Originally Posted by radier

Please send info to ASRock Support about fan issue that your friend have because I had an impression that they don't belive this problem really exist.

---

Thanks.:up: I will ask him to report the problem to the Asrock support.

Lates ASRock Extreme Tuning Utility ver:0.1.79.1 is available at Extreme4 Beta Zone:

http://www.asrock.com/mb/beta.asp?Model=P67%20Extreme4

Quote:

---

Originally Posted by Nw0rb

Did this and it still gives me the double fan spin up at start its like the video card fan spins up fast lowers fast then spins up fast again and this only happends when over voltage is enabled and my best friend has a extreme6 as well with the same problems except he can do 4.8 without over voltage and i cant

---

Strange. As far as I'm aware any bugs found on the Extreme4 that applied to the Extreme6 as well got fixed with a UEFI update. As I don't have a Extreme6 theres not much more I can say on the matter. Should a Asrock rep read this post though, by all means send me a Extreme6 and I'll bug test it.

Quote:

---

Originally Posted by Ket

Should a Asrock rep read this post though, by all means send me a Extreme6 and I'll bug test it.

---

Is this thread followed by Asrock?

Ket-

Ya it seems that the bios for extreme 6 is still kinda buggywish i had the contacts you have. Im not the only one with this problem iv post other fourms with kids who have the same problem honest listen to your computer next tme you boot it up if overvoltage is enabled its more of a video card fan double boot the fan boots twice. spins like it normally does then it spins down and spins back up like its double posting and its only when overvoltage is enabled

Quote:

---

Originally Posted by bat4

Is this thread followed by Asrock?

---

Lets just say I can neither deny or confirm that but Asrock do know I have made this thread.

Quote:

---

Originally Posted by Nw0rb

Ket-

Ya it seems that the bios for extreme 6 is still kinda buggywish i had the contacts you have. Im not the only one with this problem iv post other fourms with kids who have the same problem honest listen to your computer next tme you boot it up if overvoltage is enabled its more of a video card fan double boot the fan boots twice. spins like it normally does then it spins down and spins back up like its double posting and its only when overvoltage is enabled

---

I might try and see if Asrock want to give me a Extreme6 to bug test, but I'm going to need good reasons. I think its clear to see my bug testing is pretty unrivalled by all the bugs I uncovered with the Extreme4 but the fact remains I'm just a guy whos done R&D for some high profile name manufacturers before, been asked to design certain things for other manufacturers in the past, but I am not a Asrock rep and the decision is entirely up to them if they would find my services useful.

Ket,
I sincerely hope Asrock takes you up on your offer. It seems to me that the Extreme6 has more bugs/issues than the Extreme4.

I'm still at a 45x multi wall. Still get double-boot problem with PLL Overvoltage enabled. I'm very intrigued by Radier's ~5ghz settings, but they just cause a black screen with blinking cursor on my end... :(

after using the extreme4 for 2 weeks.. I would said the board is cool and able to perform well. so far it didn't gave me any problems at all... im glad that i had chosen this board. so far this got to be one of the brand choice i would consider and picked next time if I happen upgrade to another chipset.

my concern for now is the Asrock Z68 extreme 4 and 6 will be coming out .. will asrock be forgotten those users like us who are on the P67 chipset motherboard whether on extreme4 or extreme6 ... any continued support ?

Quote:

---

Originally Posted by blacksun1942

Ket
I'm very intrigued by Radier's ~5ghz settings, but they just cause a black screen with blinking cursor on my end... :(

---

Don't worry. My settings are universal exept multi and vCore. I was very lucky with my CPU batch because as it can run 4,5GHz effortlessly with default vCore and 5GHz with very low vCore (1,336V).



Sent from my GT540 using Tapatalk

Quote:

---

Originally Posted by Strike@ps3

after using the extreme4 for 2 weeks.. I would said the board is cool and able to perform well. so far it didn't gave me any problems at all... im glad that i had chosen this board. so far this got to be one of the brand choice i would consider and picked next time if I happen upgrade to another chipset.

my concern for now is the Asrock Z68 extreme 4 and 6 will be coming out .. will asrock be forgotten those users like us who are on the P67 chipset motherboard whether on extreme4 or extreme6 ... any continued support ?

---

Of course they will continue to offer support, ASRock p67 UD3 mobo's were only released just a couple months ago. While the release of the new chipset is not expected until mid-may at the earliest, the p67 boards will continue to receive support as long as it's needed.

Just been doing a little more testing. My max multi seems to be 55, I can only achieve this with HT turned off, otherwise chip gets too hot.

Now is it better to run slower with HT turned on or faster with HT turned off?

I'm guessing the later, do many games/applications actually take advantage of HT?

Quote:

---

Originally Posted by Sync0r

Now is it better to run slower with HT turned on or faster with HT turned off?

I'm guessing the later, do many games/applications actually take advantage of HT?

---

Games are not using HT. Therefor for gaming it is enough a I5 2500K processor which has no HT.

HT is needed for example in case of video editing or industrial process simulator applications.

Its better to leave HT on. While games won't really get any advantages theres a lot of other software that will. It depends what you mainly use your system for though.

For those having trouble with PCIE controllers the below beta bios might do the trick.

http://www.asrock.com/mb/beta.asp?Model=P67%20Extreme4
http://forums.tweaktown.com/asrock/4...rollers-2.html


EDIT:
My testing seems to show that the adaptec cards now work correctly with the BIOS L1.64a, (at least when they're placed in the 4X slot).

Just an update:
I'm finally stable @ 4500ghz and DDR3-2133
My settings:
- All c-state stuff disabled
- EIST/Speedstep disabled
- PLL Overvoltage disabled
- Spread spectrum disabled
- DRAM timings: 9-11-9-28-2N (the rest auto)
- DRAM voltage: 1.635v
- Wattage limits: 250/250/250
- Power saving mod: disabled
- Vcore: 1.375 (fixed)
- LLC: Lvl 3
- PCH Voltage: 1.059v
- PLL Voltage: 1.832v

The key was upping the VTT and VCCSA:
VTT: 1.155v
VCCSA: 1.107v

Increasing those last two was *the* solution to make my 45x+2133 clocks stable. I'll try lowering each voltage a bit at a time to see how low I can go, but 100% stable for now.
...What I'd *really* like is 46 or 47x... How much of a difference would that make in gaming anyway?

Quote:

---

Originally Posted by Redbill

For those having trouble with PCIE controllers the below beta bios might do the trick.

http://www.asrock.com/mb/beta.asp?Model=P67%20Extreme4
http://forums.tweaktown.com/asrock/4...rollers-2.html

EDIT:
My testing seems to show that the adaptec cards now work correctly with the BIOS L1.64a, (at least when they're placed in the 4X slot).

---

Very good to hear, that Asrock thinks about storage systems users.

Now give me ATA password support in UEFI, Asrock!
Just be the first! Be the first who fully supports intel G3 ssds with hardware encryption or SandForce SF-1200 and upcoming SF-2xxx ssd drives which require ATA Security Extension implemented.

Quote:

---

Originally Posted by blacksun1942

Just an update:
I'm finally stable @ 4500ghz and DDR3-2133
My settings:
- All c-state stuff disabled
- EIST/Speedstep disabled
- PLL Overvoltage disabled
- Spread spectrum disabled
- DRAM timings: 9-11-9-28-2N (the rest auto)
- DRAM voltage: 1.635v
- Wattage limits: 250/250/250
- Power saving mod: disabled
- Vcore: 1.375 (fixed)
- LLC: Lvl 3
- PCH Voltage: 1.059v
- PLL Voltage: 1.832v

The key was upping the VTT and VCCSA:
VTT: 1.155v
VCCSA: 1.107v

Increasing those last two was *the* solution to make my 45x+2133 clocks stable. I'll try lowering each voltage a bit at a time to see how low I can go, but 100% stable for now.
...What I'd *really* like is 46 or 47x... How much of a difference would that make in gaming anyway?

---

Absolutely zip. 100MHz is bugger all. You might want to leave speedstep enabled. I disabled it along with all other power saving options and my CPU really didn't like me doing that. Turned speedstep back on and.. everything is fine again. Very odd.

Quote:

---

Originally Posted by Ket

Absolutely zip. 100MHz is bugger all. You might want to leave speedstep enabled. I disabled it along with all other power saving options and my CPU really didn't like me doing that. Turned speedstep back on and.. everything is fine again. Very odd.

---

Excellent advice.
Odd thing is that my "100% stable" settings actually failed Prime95 last night on a second test after initially passing ~12 hours.

Now I'm running everything with the C states/Speedstep enabled since this morning and so far so good.... :up:

You can leave a lot of the CPU power saving stuff off, just not speedstep. Dunno why thats such a big deal, maybe the CPU tries to apply "turbo" speeds on top of the already OCd CPU and it just makes it unstable.

Didn't go through this whole thread but I had to return my Pro3. Would BSOD with any thing above stock settings. No amount of voltage or settings changes helped. Latest bios didn't help. Putting my meter on it found that even though the voltage would remain constant, the wattage would slowly drop until it crashed. It would drop at stock also but never enough to crash. Swapped out everything 1 part at a time including PSU. Nothing helped. Too bad. I've had other ASRock boards and really liked this one.

I think its probably just best to avoid the B3 Pro3 boards, nobody seems to have any success with them.

ok i just bought this board..i2500k batch number 3052A185 and 2 4gb sticks of gskill 1600 F3-12800CL8D-8GBXM 1.5 VOLTS....any recommendations on what bios to use and if my batch cpu is any good? any replies with any info is much appreciated..thanks in advance to all of u

Give a try to the latest beta 1.64a from ASRock website.

Sent from my GT540 using Tapatalk

At the very least use UEFI 1.6A. IMO 1.6A is probably the best, has a few bugs but nothing serious and seems to be the most stable in my testing.

ok ty!

Quote:

---

Originally Posted by Ket

At the very least use UEFI 1.6A. IMO 1.6A is probably the best, has a few bugs but nothing serious and seems to be the most stable in my testing.

---

Ket, does 1.6A solve the problem of removing saved BIOS profiles when clearing CMOS or is it just 1.61?

I'm currently running 1.6, the shipping BIOS, aside from the double post bug :shrug: , it seems pretty stable.

I'm currently running 100.9x46 @ 1.375Vcore, 2133 XMP profile on the RAM and HT / SpeedStep / CiE / C states all enabled. Pretty cool chip & board so far. Seems to run more smoothly than my QX9650! :up: The board is on my tech bench for the time being. I'm pretty much at the max of my air cooling though. I'm dusted off my Big Typhoon and it does ok. With 1.375Vcore, 4c/4T Linx, I see temps just touching 80C. I will eventually put use water cooling for the system once I move it to my Stacker.

If you want the fix for saved profiles being wiped you need to use UEFI 1.61 and newer. If the saved profiles being wiped aren't a big issue for you though, UEFI 1.6A in all my testing by some distance has been the most stable and has the double POST bug fixed. I do wish Asrock would look at that, AFAIK the only thing fixed with UEFI 1.61 was the saved profiles bug but stability somehow decreased when the system gets OC'd. Bad omen.

ok so i got all hardware installed but i cant get to desktop
says it cant detect harddrive
sata3 operation mode:ide mode
marvell sata3 bootable:yes
sata mode:ide
sata controller 0:compatible
sata controller 1:enabled
hard disk smart:enabled

sata3_0 st3250310as
sata3_1 not detected
sata2_2 my dvd rom

After fresh install of OS ?

i cant even get windows to install..says i need to make sure my boot drive is enabled in bios