IXTU benchmarking overclock problems PC shuts down

[TeamMinions]

Win 7

5930k 45x 4.5ghz 1.269 stable

45x cache 1.269 stable

Xmp 2666 1.2 ballistic elite

Asrock mini itx x99

Ax860i power supply

Gtx 970

1.9 input

72 degrees max in stress testing

 

 

Those settings verify and for the life of me I cannot go higher then 45x. Whatever I do when I start to benchmark at 46x+ it immediately shuts down. Even if I put the core voltage at 1.3 1.4 1.5 no matter what it shuts down within 1 second. I don't see any thermal throttling. I can't tell if it's the VRM shutting down or what's going on. I have tried no xmp , lower cache settings, 1.8-2.1 input. Speed step on and off, blck at 125 and lower multiplier, nothing works.

 

Any ideas as to why it is shutting down when going beyond 45x?

 

Note when doing 46x at 1.269 volts it BSOD and I am thinking the voltage was too low but when I raise it to even 1.299 it shuts down. I can't tell if it's a bios issue when the voltage goes to high or what. I don't see any temps over 75 so far.

[Massman]

What value are you booting up from?

[TeamMinions]

Those values. Boot fine, benchmark fine, stress test fine. Anything over 4.5 hence 4.6 and pc turns off when testing. I read somewhere that it could be my power supply with ocp enabled and its drawing too many Amps for the 12v rails but after reading a lot online my ax860i should be able to handle any overclock and single gpu. I don't have corsair link enabled to disable and check but that doesn't sound right

[Massman]

Hm, the AX860i did have some issues related to random shutdowns but since you're running fine at 4.5GHz that shouldn't be the problem.

 

Can you try booting up with 46x ratio and 1.35V? Even if it's not XTU stable, it should be able to pass POST (boot) and even go into Windows.

 

It almost sounds like there's an issue with the µITX board. OCP is possible; might also be BIOS/board related. I haven't seen many builds with the ITX, let alone overclocks, so it's hard to say if asrock build in some protection. I would assume they did since the PCB real estate is very small on ITX.

 

Maybe someone closer to Asrock can check?

[TeamMinions]

Yeah I have no idea. It won't even pass boot at 46x and 1.35v. I can only boot and load windows with my 4.5 settings then try to tune to see anything. If I do higher then 45x in bios and restart it just hangs. doesn't even display anything. It just shuts off after around 15-20 seconds. Very frustrating because this is a brand new build, days old.

[Massman]

I have a vague memory of something similar happening when testing Haswell a long time ago.

 

Did you disable turbo/eist/c-states?

[TeamMinions]

Yeah tried everything. It has to be some kind of protection being enabled because its highly unlikely the 1x increase can reach tjmax In 1 second when 4.5 is stable at 70 degrees 100% load. When previously overclocking Asus rampage 5 extreme and Formula z97 I always got BSOD when trying for higher clocks or saw crazy temps at higher voltages, never a full power off in 1 second. This is my first build with asrock and I am not very happy.

[Guest TheMadDutchDude]

It sounds like a CPU limit. It happens, fella. I had a 2500K which used to do 4.7GHz all day long, but it wouldn't even complete POST at crazy high voltages at 48x. They just hit walls, and there's nothing you can do about it unless you buy another chip that may be able to go further.

[BoTSkill]

Not on Hasswell-E. They don't act like that. To exclude the PSU run xtu with 4.5ghz settings that work but with voltage set to 1.35 or 1.4. If it passes than it's not the PSU. If it shuts down than most probably is the motherboard, subdimensioned vrm or some king of OCP protection to protect the motherboard. The PSU should be more than enough for your cpu at 1.35 -1.5v.

[Massman]

It sounds like a CPU limit. It happens, fella. I had a 2500K which used to do 4.7GHz all day long, but it wouldn't even complete POST at crazy high voltages at 48x. They just hit walls, and there's nothing you can do about it unless you buy another chip that may be able to go further.

 

Sandy Bridge architecture was just like that - it's different from Ivy Bridge and Haswell.

[BoTSkill]

According to this review - http://www.tweaktown.com/reviews/7112/asrock-x99e-itx-ac-mini-intel-x99-motherboard-review/index5.html - the mobo should have no problem booting with 4.7-4.8ghz on the cpu, it says that:

 

"The VRM wasn't the limitation, so with better cooling, I am sure you can hit the same high overclocks as a normal ATX board."

 

What cache do you try to run? Tried higher Input voltage? Like 2.1?

 

Settings used in tweaktown review to find max bootable frequency:

"Max CPU Overclock is found by setting the VCore to 1.5v, Input voltage to 2.1v, cache voltage to 1.15v, booting with a CPU multiplier of 45x and disabling any features that would result in CPU frequency fluctuation."

 

So try these:

1. Set the VCore to 1.5v, Input voltage to 2.1v, cache voltage to 1.15v and see how high frequency u can boot at.

 

2. Connect the usb plug, install the Corsair Link Software and disable OCP for the rails.

 

3. Try with another PSU.

 

Al of these with cpu cache ratio set at low 30 or under.

[Massman]

Good ideas

 

- try 4.6G with higher input voltage

- try running 3GHz cache

 

Based on your XTU submissions I guess BIOS isn't the issue since you're running the same like in the Tweaktown review.

[TeamMinions]

4.5 at 1.3 1.35 1.4 passes fine just higher temps.

 

4.6 and 2.1 input still doesn't work.

 

4.6 and even 2.3 cache doesn't work

 

Just tried 125 strap stock speeds 3.5 and 1333 memory. Auto vcore was at 1.14 and passed fine.

 

Now just tried 4ghz same setting and vcore wanted 1.15 and bam instant shut down with benchmarking, wtf??? Ugh

Edited July 3, 2015 by TeamMinions

[Guest Ximi]

Try 4.4 ghz cache with 1.42v on cache system agent.

[intender]

I hate to bring this back from the dead, but did you ever figure this out. I just built a 6600k system that is doing the same thing. I can bench and stress test for hours with no trouble at 4.7 and still tweaking to get 4.8 stable. But xtu is killing me. I can stress test in xtu with no problems at 4.7. It ran for 3 hours yesterday, but as soon as I click benchmark windows blue screens. Even at 4.6 and 4.5 I get random blue screens in xtu sometimes.

 

Sent from my SM-N910V using Tapatalk

[nhaxinhcenter]

I also have questions like this, who can answer the details?