Please help with my OC - ASRock Z68 Extreme4 + 2600K

[SabreWulf69]

I'm looking at you specifically Mr. Ket for help on this one as you have this board yourself (AFAIK) and am a great admirer of your abilities and skills.

Currently OC'ing a 2600K on the ASRock Z68 Extreme4 and having some issues. I am going for the standard sort of fair clock with this setup of a 4.6GHz clock using decent enough air cooling. As far as other relevant parts of this system are concerned there is - a Gigabyte 560Ti Superoverclock GPU in it, a Enermax NAXN 850W PSU, Vertex 2 240GB SSD, and 8GB of G.Skill RAM (G.Skill Ripjaws X F3-17000CL9D-8GBXLD (2x4GB) DDR3), a Lian-Li V2120B chassis, a Prolimatech Super-Mega CPU cooler with 2x 12cm Gelid Wing Blue LED fans on it using IC Diamond Carat 7 TIM with the rear fan being an Enermax Apollish Vegas fan.

As you will see from the following UEFI shots the idle temp in there suggests a temp of 51 degrees Celsius, but RealTemp in Windows suggests an idle temp of around 44-46 degrees Celcius. Using LinX AVX the max load temps of one of the cores hits around 86 degrees (Australian summer and no aircon for the loss :-() but it will run a full 50 times run without BSOD'ing.

I provide these UEFI screenshots as I am sure you will be able to spot something that I may have done really wrong. My understanding is to use SpeedStep although I have run fine without it before besides when trying to clock even higher to something stupid on air like 5.4GHz just to do a 'see if I could' sort of thing, of which would refuse to play right if I did disable SpeedStep. Getting down to the actual issue now and I hope I have provided enough background information - it randomly BSOD's when clocked but NOT being pushed. Any ideas or anything you can spot from these shots that I have done wrong?

I have read the Sandy Bridge guides, but information remains elusive to my specific problem. Everything else has been ruled out as when running on stock settings I do not have the issue. I must be doing something wrong with the OC but am not sure where and would really love assistance from Ket or anyone else that would care to chime in with some tips. In advance I thank anyone for their help.


UEFI shots as attached -

[SabreWulf69]

I don't usually make a habit of bumping things, but any help on this one guys?

UEFI is now the latest but the random BSOD issue still persists. I don't know where I'm going wrong.




I'm going to try putting the vcore on Auto and see what it does for now. So far I have gathered some information from elsewhere, and hope it may serve me and any others out there just as well -

For Asrock Mobo's
Turbo Boost Power - Manual
Short Duration Power Limit - 300
Long Duration Power Limit - 300
Core current Limit - 300


For the ASRock 1155 boards, 5 is the lowest LLC setting. You will probably want it to be at 2 or 3 if you're overclocking. Don't set it to 1 because that will cause spikes in your voltage which is dangerous.

Definitely take that voltage off auto! You can do much better with fixed/offset.

The "Long Duration Maintained (Sec)" setting is how long in seconds before the TDP limit kicks in and the CPU is throttled down.
So if I have a higher Turbo Power Limit, it will be unlikely to get to a situation where the above setting needs to kick in and take effect.

I had to turn on Intel SpeedStep, and increase the Long Duration Power Limit from 95W to 110W. With this, I hit 4.6GHz and 4.7GHz (hitting 73 ºC, and reducing to 4.6GHz so I increased the power to 120W)

Fixed Mode: Your CPU will run full voltage and speed all the time, even when idling.
Enhanced Halt State (C1E): Enabled
CPU C3 State Support: Enabled
CPU C6 State Support: Enabled
Package C State Support: Auto

- Turn on Intel SpeedStep
- If using 'turbo mode' then -
Turbo Boost Power = Manual and set all 'limits' to 300, and then dial back incrementally until load testing goes unstable (may not do anything above a certain point)

- Turbo mode may be disabled for better results to prevent multiplier auto-switching inside OS if it is an issue
- Try 1.365 Vcore + for 4.6GHz
- PLL Overvoltage enabled and set at 1.7V
- RAM at X.M.P.
- Try LLC of 2 or 3 (lower numbers equal stonger LLC effect and vice versa)
- Run at 'Fixed Mode' and disable all 'C-States'
- VCCSA should be set to Auto
- 85 degrees max when load testing
- Select CPU PLL Overvoltage to enabled, this allows most people to go beyond 4.8 or 5.2 GHz whatever your barrier is. If you have a D1 stepping processor, enabling this option will cause you to fail to boot. (Refer to CPU-Z)
- All fans as high as possible


BSOD Codes
0x124 = add/remove vcore or QPI/VTT voltage (usually Vcore, once it was QPI/VTT)
0x101 = add more vcore
0x50 = RAM timings/Frequency add DDR3 voltage or add QPI/VTT
0x1E = add more vcore
0x3B = add more vcore
0xD1 = add QPI/VTT voltage
“0x9C = QPI/VTT most likely, but increasing vcore has helped in some instances”
0X109 = add DDR3 voltage
0x0A = add QPI/VTT voltage


VCC/VCCIO/CPUIO
It's CPU Input/Output (IO voltage terminals. This voltage work just like QPI/VVT/IMC and the default voltage is
1.05v but it needs to keep within 1.10 - 1.20v when your overclock beyond 4GHz, increasing this voltage also help
to improve the stability when 4 memory DIMMs occupied and its max speed (2133MHz). This high voltage
doesn't seem to affect overclocking.

PCH Voltage - Platform Control Hub
The default value is 1.05v and no need to change this voltage because it doesn't contribute to any overclock capability
therefore keeps it at default setting.

[Ket]

I have a Beta UEFI for the board you can try for starters; Clicky. This UEFI is date stamped newer than official 1.7. Increase your power and current limits to 250, set LLC to level 1 (level 1 is less aggressive than it was on the P67), set BCLK to 100.1, leave IGPU voltage at auto, increase CPU PLL to 1.857v, increase VTT to 1.155v, increasing VCCSA to 1v might help as well. You should now be good to go. Do remember only roughly 35-40% of CPUs will be able to handle 4.6GHz+ completely stable so worst case scenario go for 4.5GHz. Don't worry about the temps the UEFI reports too much, AFAIK the UEFI reports actual core temp, where other apps report the IHS surface temp. Install the AXTU software and you will see that even reports the same temps as things like realtemp or coretemp will

You should also be sure you are running with the latest drivers, I haven't got a link for them yet I'm still rebuilding all my archives and directory trees after the FBI being complete retards and just killing the entire megaupload site.

[SabreWulf69]

Thanks for that. Seems to be stable on +0.020V in Offset mode atm without a bluescreen to report, although every now and then it will 'hickup' with a microfreeze of sorts. Will try adjusting some of those values to what you have suggested, and once again thankyou very much for your assistance. I also look forward to your repacked newest lot of drivers :-)

[Ket]

NP Heres the latest set of drivers; Clicky.

Drivers in package:

Broadcom BCM57781 LAN Driver v14.8.0.2
Intel HD Graphics Driver v8.15.10.2509 (modified by me, should be more stable and give better performance)
Intel Management Engine Driver v7.1.20.1119
Intel PCH Driver v9.3.0.1019
Marvell SE9120 SATA3 Driver v1.2.0.1016
Realtek HD Audio Driver R2.67
EtronTech USB 3.0 WHQL Driver v1.109
Intel RST v10.8.0.1003
Lucidlogix VIRTU Driver v1.2.112

If you have any odd problems with the Etron driver let me know. Although the USB issues are reasonably under control now it can sometimes be a bit fickle. As for the microhiccups, that might be related to RAM or CPU voltages occasionally dipping a touch lower than normal, just up vcore by 0.05v-0.1v and dial in RAM timings manually, also set RAM voltage a touch over whatever stock voltage is just to allow for any drop under heavy load.

[TammyWulf]

Image of this PC in it's current form --v