cancel
Showing results for 
Search instead for 
Did you mean: 

Asus Maximus VI Gene + 4670K @ 4GHz random failures at boot

Faravid
Level 7
Hi,

I recently upgraded from MSI mobo and 2500K to Maximus VI Gene + 4670K.
I'm running the setup with CPU turbo @ 4GHz for all cores without touching any voltage controls.
Only other thing I've changed from the BIOS is turning on XMP profile for my 4+4GB of Corsair Vengeance LP.

I'm having a problem with this, computer tends to boot and immediately reboot and after forcing computer down and rebooting I've been greeted with overclock failed message (doesn't do this every time).
If I enter the BIOS and without any changes reboot the computer (from BIOS) my computer boots up fine and Windows 8.1 runs completely stable without any kind of problems, stress tests pass fine and games run like they should and temperatures are fine.

This behavior is odd as my 2500K ran 4GHz turbo without touching volts perfectly fine. And this 4670K got higher default turbo clock.

Any ideas?
Main setup: Ryzen 7 1700x @ 3.8GHz + NZXT Kraken x72, ASUS Crosshair VI HERO, DDR4 3600MHz CL15 16GB @ 3400MHZ CL14, Samsung 850 EVO 500GB SSD + 2TB HDD + 3TB external backup drive, ASUS Cerberus GTX 1070 Ti, Corsair Carbide Air 540 Silver, Corsair RM750 PSU, Windows 10 Pro 64-bit.
Laptop: ASUS ROG Strix Scar II GL504GS
6,437 Views
13 REPLIES 13

Tokens210
Level 10
Those chips are totally different so you can't really compare them as your trying

Haswells also like more voltage then most CPUs in a lot of cases

Try switching windows power plan to high performance mode, then run the stress tests



My guess is atleast one of your cores wants more power but doesn't always require it, in which case slightly raising voltage may fix your issue


Switching the power plan to high performance, will force the CPU to remain at the turbo clock so u can see if it crashes its most likely starving

I don't remember 100% but I believe the bsod code for voltage is something like 000000000124


Also as I was saying there different the haswell at 4.0 technically is much faster then the 2500 at 4.0

Also haswells are touchier on memory so running 4 dimms is more stressful, as well as running them at xmp mode so that could also have upped the voltage required
CoolerMaster HAF 932 Advanced/ Maximus VI Formula/ I7-4770K/Swiftech H320/ Corsair HX850/ G.Skill Trident X (2x8) 16gb 2400MHz/ 2x 840 EVO 120gb(Raid 0)/ WD 1TB HDD (Backup/Storage)/ EVGA GTX 1gb 560 TI/ Asus 12x bluray combo

Tokens210
Level 10
Those chips are totally different so you can't really compare them as your trying

Haswells also like more voltage then most CPUs in a lot of cases

Try switching windows power plan to high performance mode, then run the stress tests



My guess is atleast one of your cores wants more power but doesn't always require it, in which case slightly raising voltage may fix your issue


Switching the power plan to high performance, will force the CPU to remain at the turbo clock so u can see if it crashes its most likely starving



Also as I was saying there different the haswell at 4.0 technically is much faster then the 2500 at 4.0
CoolerMaster HAF 932 Advanced/ Maximus VI Formula/ I7-4770K/Swiftech H320/ Corsair HX850/ G.Skill Trident X (2x8) 16gb 2400MHz/ 2x 840 EVO 120gb(Raid 0)/ WD 1TB HDD (Backup/Storage)/ EVGA GTX 1gb 560 TI/ Asus 12x bluray combo

Faravid
Level 7
Thanks for the quick answer!

Switched the power plan to high performance, all cores running 4GHz and 5 tests (Blend) with Prime95 passed.
Max package temp 61C
Also passed with In-place large FFTs (5 tests)
Max temp 65C
Same goes with small FFTs
Main setup: Ryzen 7 1700x @ 3.8GHz + NZXT Kraken x72, ASUS Crosshair VI HERO, DDR4 3600MHz CL15 16GB @ 3400MHZ CL14, Samsung 850 EVO 500GB SSD + 2TB HDD + 3TB external backup drive, ASUS Cerberus GTX 1070 Ti, Corsair Carbide Air 540 Silver, Corsair RM750 PSU, Windows 10 Pro 64-bit.
Laptop: ASUS ROG Strix Scar II GL504GS

Tokens210
Level 10
Hmm, still probly a voltage related issue at least I'd think

What are you using to monitor temps, also if you have multiple programs able to monitor temps there probly reporting the wrong temps
Not that I think temps are your issue, so far they seem to be in a really good range

But core temps are what you want, program like aisuite doesn't show that temp



Again the most practical guess I'd have is at least 1 core wants a bit more, I have a 4770k which is also a has well, for me core 3 always needs more then the others which requires me to increase the voltage or use a per core setup
CoolerMaster HAF 932 Advanced/ Maximus VI Formula/ I7-4770K/Swiftech H320/ Corsair HX850/ G.Skill Trident X (2x8) 16gb 2400MHz/ 2x 840 EVO 120gb(Raid 0)/ WD 1TB HDD (Backup/Storage)/ EVGA GTX 1gb 560 TI/ Asus 12x bluray combo

Faravid
Level 7
I used CPUID HWMonitor, core max temps are 62C for core 0 and 1, 61C for core 2, 60C for core 4.

CPU-Z says my CPU is running 1.12V when running Prime95, if I up it to 1.13 or 1.14V from bios it should do the trick, it's currently set on auto?
Looks like HWMonitor reports VID 1.124V, max 1.29V, so may be a bit more accurate?
When I run Prime95 for longer period of time CPU-Z shows VID 1.2 to 1.136V.

How does BIOS voltage tune work, does CPU still scale down voltage when it runs on lower power saving frequency?
Main setup: Ryzen 7 1700x @ 3.8GHz + NZXT Kraken x72, ASUS Crosshair VI HERO, DDR4 3600MHz CL15 16GB @ 3400MHZ CL14, Samsung 850 EVO 500GB SSD + 2TB HDD + 3TB external backup drive, ASUS Cerberus GTX 1070 Ti, Corsair Carbide Air 540 Silver, Corsair RM750 PSU, Windows 10 Pro 64-bit.
Laptop: ASUS ROG Strix Scar II GL504GS

Tokens210
Level 10
That's fine for temps and monitoring, just wanted to be sure you weren't only seeing package temps as they tend to be a bit lower then cores temp but you seem fine there

By default the Asus board allows the CPU core voltage to go as high as 1.250-1.275v if my memory serves right but the board/CPU does allow wiggle room


As for your question about it stepping up and down, the answer to that is yes and no
A manual voltage will not step down, an adaptive will, but you'll want to find a stable voltage using manual first, then use the adaptive
Don't run prime if your using adaptive and your voltage is higher then 1.3 something unless you have a very good cooling as prime95 with has wells can automatically add voltage whenever it wants to due to instructions in both the CPU and prime


I see you have an antec cooler, I'm not familiar with them but it sounds to be a water cooling solution, every single has well is different, sometime just a little sometimes a lot, but with that being said I'd think you could easily do a little more voltage, just watch our temps, no 90's try to steer as far as u can from there, having temp spikes in the 70-80c range will be fine but you also wanna watch for it sitting at any higher 80's as after that temp most haswells hit a downhill running
CoolerMaster HAF 932 Advanced/ Maximus VI Formula/ I7-4770K/Swiftech H320/ Corsair HX850/ G.Skill Trident X (2x8) 16gb 2400MHz/ 2x 840 EVO 120gb(Raid 0)/ WD 1TB HDD (Backup/Storage)/ EVGA GTX 1gb 560 TI/ Asus 12x bluray combo

Faravid
Level 7
Ok I'll try that, thanks.

Yeah this is a closed circuit water cooler, that I've rigged so pump operates 100% speed all the time from case fan plug and fan is operated from CPU fan plug, also I've replaced stock fan with slightly better one.
Main setup: Ryzen 7 1700x @ 3.8GHz + NZXT Kraken x72, ASUS Crosshair VI HERO, DDR4 3600MHz CL15 16GB @ 3400MHZ CL14, Samsung 850 EVO 500GB SSD + 2TB HDD + 3TB external backup drive, ASUS Cerberus GTX 1070 Ti, Corsair Carbide Air 540 Silver, Corsair RM750 PSU, Windows 10 Pro 64-bit.
Laptop: ASUS ROG Strix Scar II GL504GS

InFAMAS-ONE
Level 8
Faravid wrote:
Hi,

I recently upgraded from MSI mobo and 2500K to Maximus VI Gene + 4670K.
I'm running the setup with CPU turbo @ 4GHz for all cores without touching any voltage controls.
Only other thing I've changed from the BIOS is turning on XMP profile for my 4+4GB of Corsair Vengeance LP.

I'm having a problem with this, computer tends to boot and immediately reboot and after forcing computer down and rebooting I've been greeted with overclock failed message (doesn't do this every time).
If I enter the BIOS and without any changes reboot the computer (from BIOS) my computer boots up fine and Windows 8.1 runs completely stable without any kind of problems, stress tests pass fine and games run like they should and temperatures are fine.

This behavior is odd as my 2500K ran 4GHz turbo without touching volts perfectly fine. And this 4670K got higher default turbo clock.

Any ideas?


As Tokens said, those chips are totally different. However, it's not that you can't compare just those chips, but you can't really compare the same chip. The maximum overclock you will achieve and the voltage it will require depends on the quality of your chip itself. That's not to say that you have a faulty processor if it won't overclock by much, or if it needs a higher than expected voltage to get there, it's just the luck of the draw. Some CPU's will go way beyond expectations and require a very modest voltage, others really need a kick up the butt and still might not get that high. You could buy two 4770K's for example, and try them both in the same computer, and they will probably overclock to a different degree, or need more or less voltage than the other will.

I currently have an i5 3570K on my Maximus V GENE. I read a guide that ROG did for overclocking this rig:
http://rog.asus.com/184142012/maximus-motherboards/guide-overclocking-the-core-i5-3570k-to-4-5ghz-on...

The guide says, and I quote: "1.25V is a good starting value that should be enough to get any i5-3570K to 4.5GHz, but this CPU was able to run stably at 1.21V."

However, I can only get mine up to 4.4GHz and I'm having to set a voltage of 1.34v. The CPU actually draws close to 1.38 under stress testing. I did manage to get it to 4.5GHz, but it wouldn't run at idle under 1.4v and wasn't stable enough to pass Real Bench, so I had to settle at 4.4GHz. This really is the limit for my CPU, some people have gotten them to 5GHz! This is extremely rare though.

I think ASUS have built an "Auto tune" function into AI Suite 3, I think it's called "4-way optimization". It might be worth using that, stressing it out to check it's stable, and then go into the BIOS and start dropping the voltage from there. AI Suite makes hardware level changes, so the overclock will be applied to the motherboard, so you should see the settings it's saved if you go into the BIOS.

Faravid
Level 7
I've upped turbo voltage by very small increase of 0.005 to see if it has an effect on the boot problems, so far no boot failures, I may try some higher clocks later on just for kicks of it with higher voltage boost.
I'm using adaptive mode with only the turbo voltage increased and full manual mode disabled.
Main setup: Ryzen 7 1700x @ 3.8GHz + NZXT Kraken x72, ASUS Crosshair VI HERO, DDR4 3600MHz CL15 16GB @ 3400MHZ CL14, Samsung 850 EVO 500GB SSD + 2TB HDD + 3TB external backup drive, ASUS Cerberus GTX 1070 Ti, Corsair Carbide Air 540 Silver, Corsair RM750 PSU, Windows 10 Pro 64-bit.
Laptop: ASUS ROG Strix Scar II GL504GS