Blue Screen of Death 4770k 75C

[andrewmont 91]

I have an i7 4770k that came pre clocked to 4.2Ghz, At idle it would run between 50-60C and under load anywhere from 60-75C.

 

Playing Day Z a couple of times I've had the blue screen of death and the temperatures were hitting 75C (maybe at times closer to 80 but I never saw this myself) So I set my bios back to defaults to remove the overclock and any other fiddling they may have done. The only thing I'd personally changed was disabling hyperthreading. 

 

Once I'd reset it to defaults I played it for a wee bit and temps never went above 63C and I didn't get the blue screen of death. However when playing games like shogun 2 I saw my PC hit similar temperatures of 75C and it never crashed. So I'm not entirely sure whether it's just coincidence that it happened with Day Z and it simply is my CPU getting too hot. Or maybe if it's a software issue? 

 

Any help would be appreciated thanks. 

 

PS. currently in process of complaining to customer service because the PSU they supplied me with was a cheap powercool 750w 80+ bronze x-viper and has started making high pitched whine noises. Particularly noticeable when the PC is off. 

 

EDIT: The motherboard is a Z87-K (sorry) the cooler is the thermaltake contac 29

Radeon HD 7770, 8GB Crucial Ballisitks RAM

Edited January 15, 2014 by AndrewM

[enforcer1975 1111]

What exaclty did you overclock?

[SaveMeJebus 164]

Wow .. even being haswell it should really only hit those temps if you are stress testing with prime95 or something I would think, that seems hot to me.  What are you using for a cpu cooler?

** Temps actually seem typical for haswell chips, can't believe they purposely run that much closer to thermal limit than they used to.   Your idle temps I guess is the only thing that seems high, most of the idle temps I saw were under 50C.  Maybe a reapply of thermal grease or a better cpu cooler will help you out, or maybe you have a stick of ram that is giving you problems once it starts getting hammered.

Edited January 14, 2014 by SaveMeJebus

[rustledmyjimmies 361]

That's not at all hot for a cpu, 75c should not be a problem. It's probably that you'd done something that had cocked up the overclock that caused the crashes.

[Stuntz 10]

Can you post the full bsod message, error code included? Or just upload the dmp file so I can get a look at it? Might be able to narrow down the source of your crashes.

 

Would also help knowing what mobo you're on.

Edited January 15, 2014 by Stuntz

[andrewmont 91]

Added other system specs to the post (sorry bout that) I checked the event log but I have a lot of errors and warnings and I'm not sure how I would make sure I've got the right one as I don't know exactly what time it happened at, I just know it was the 11th or the 12th. I'm trying to attach a picture of the changes made in the bios by the company I bought the PC from the only thing I would've changed from what they did was disabling hyperthreading. I think it's the PSU as it is low quality and I'm making it clear I'm not happy with it either way, the reason I suspect the PSU is that the system also had an overclock failed message at one point during startup and the computer normally attempts to boot up for a few seconds once before it shuts down then actually starts normally.

 

 

 

  

Edited January 15, 2014 by AndrewM

[Ansssi 184]

Your crappy PSU is probably feeding your system with fluctuating voltage, causing instability.

[Stuntz 10]

Aye odds are its a low qual psu, but if you wanna post any of the dump logs/files, might be able to narrow down further, especially on an asus board, they're very specific in their blue screens compared to most. Can narrow down to the voltage giving the issue on them.

[andrewmont 91]

Aye odds are its a low qual psu, but if you wanna post any of the dump logs/files, might be able to narrow down further, especially on an asus board, they're very specific in their blue screens compared to most. Can narrow down to the voltage giving the issue on them.

Would you be able to help me find them? I've looked and there's a lot of error messages and warnings, I have no idea which one would be the blue screen event. 

[Stuntz 10]

C:\Windows\Minidump

any of the recent ones in there will do.

[andrewmont 91]

C:\Windows\Minidump

any of the recent ones in there will do.

I found one file there however it was from the 14th and its 900kb so it's too large to upload here. (sorry if I'm being an idiot and missing obvious things, Its the first time I've had to actually access dump files before.

[Stuntz 10]

http://www.resplendence.com/download/whocrashedSetup.exe

download that, install, click analyze, and copy/paste what the crash dump analysis says :P

 

Works too ^^

 

Edit: specifically the bugcheck code line!

Edited January 15, 2014 by Stuntz

[andrewmont 91]

On Tue 14/01/2014 17:42:10 GMT your computer crashed
crash dump file: C:\Windows\memory.dmp
This was probably caused by the following module: atikmpag.sys (0xFFFFF88006E36D88)
Bugcheck code: 0x116 (0xFFFFFA800D1794E0, 0xFFFFF88006E36D88, 0x0, 0x2)
Error: VIDEO_TDR_ERROR
file path: C:\Windows\system32\drivers\atikmpag.sys
product: AMD driver
company: Advanced Micro Devices, Inc.
description: AMD multi-vendor Miniport Driver
Bug check description: This indicates that an attempt to reset the display driver and recover from a timeout failed.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: atikmpag.sys (AMD multi-vendor Miniport Driver, Advanced Micro Devices, Inc.).
Google query: Advanced Micro Devices, Inc. VIDEO_TDR_ERROR

 

I'm not sure if this is the right problem as this happened yesterday and I never got blue screen yesterday. However I do have a windows error"solve a problem with AMD display adapter-ati radeon graphics cards". As I say though the crash happened on either the 11th or the 12th and this error is just recent. 

Edited January 15, 2014 by AndrewM

[andrewmont 91]

 

 

Here are the changes made by the company to the BIOS, is there anything here they've done that's perhaps causing instability or is even just stupid? I've lost complete faith in this company. 

Edited January 15, 2014 by AndrewM

[Stuntz 10]

0x116 indicates either a low IOH voltage, or an overclocked GPU failure. Is that 7770 oc'd?

 

VIDEO TDR ERROR suggests a display driver crash that windows couldn't recover from.

 

Try uninstall your graphics driver, reinstall the newest ones. Make sure your directX runtime is as up to date as possible.

 

Should your current drivers be the most up to date ones, try rollback to catalyst 13.10, they were quite stable on the 7000 series cards.

Edited January 15, 2014 by Stuntz

[andrewmont 91]

0x116 indicates either a low IOH voltage, or an overclocked GPU failure. Is that 7770 oc'd?

 In the CCC the Over Drive is enabled but none of the values have been changed. So it's at 1000 and 1125mhz which are the default values for it. The crash hasn't happened again since I reset the BIOS to default which is what made me think it was the CPU overclock. 

Edited January 15, 2014 by AndrewM

[Stuntz 10]

I'm guessing they only upped the vCore voltage and nothing else. Left the IOH @ stock, causing the issue.

[andrewmont 91]

I'm guessing they only upped the vCore voltage and nothing else. Left the IOH @ stock, causing the issue.

 

Could you explain this too me in layman's terms :L so that I understand what they've actually done and can use it against the ignorant customer service. Thanks very much for all the help btw, I know a bit about computers but not enough that when something like this happens I'm not confused, annoyed and have to resort to forums to try and get the help I need :L 

[Stuntz 10]

The basics of it is:

 

IOH stands for Input/Output Hub, something akin to the older style of northbridge.

 

Though someone correct me if I'm wrong - i believe from sandybridge chipsets onwards its PCH (Platform Controller Hub) now, not IOH. Functionality remains the same regardless, in essence.

 

I'm guessing your issue was there if a bios reset fixed the issue. If you get further bsod's, consider the drivers to be at fault.

 

Either way, a pre-OC'd pc should have been burn/stress tested to ensure the voltages were correct. No 2 chips/boards are identical, voltages that work for one won't necessarily work for another board/chip of the same model (only need to look at the massive variance in results of people's haswell overclocking to see that). The issue would have been readily apparent.

 

The pch 1.05v should be (estimating) ~1.08-1.085v, and the PCH 1.5v should be no higher than 1.525. Given that your BSOD's were only occuring under heavy stress (dayz in this case), I'd guess a small tweak would have done it.

Edited January 15, 2014 by Stuntz

[andrewmont 91]

The basics of it is:

 

IOH stands for Input/Output Hub, something akin to the older style of northbridge.

 

Though someone correct me if I'm wrong - i believe from sandybridge chipsets onwards its PCH (Platform Controller Hub) now, not IOH. Functionality remains the same regardless, in essence.

 

I'm guessing your issue was there if a bios reset fixed the issue. If you get further bsod's, consider the drivers to be at fault.

 

Either way, a pre-OC'd pc should have been burn/stress tested to ensure the voltages were correct. No 2 chips/boards are identical, voltages that work for one won't necessarily work for another board/chip of the same model (only need to look at the massive variance in results of people's haswell overclocking to see that). The issue would have been readily apparent.

 

The pch 1.05v should be (estimating) ~1.08-1.085v, and the PCH 1.5v should be no higher than 1.525. Given that your BSOD's were only occuring under heavy stress (dayz in this case), I'd guess a small tweak would have done it.

Well if I remember correctly the CPU Core Voltage was 1.291v, does this seem about right? I'll maybe just try and get a new PSU from them and see if I can recreate the issue when I get it, I'm just hoping there's nothing that's going to be damaging my computer in the long run, I'd hoped I could trust a computer supplier to do an OC correctly :L

[Stuntz 10]

1.29 is fine, On the haswell's I'd be wary about going over 1.35v even on a decent closed loop watercooler.

 

Depending on your sample luck, you might be able to hit 4.2GHz @ 1.25v, some can even do it at stock if really lucky.

 

@ 1.3v some would even be able to get 4.4 or 4.5 from it, but that's pushing it. Regardless, that seems fine.

 

There's plenty of voltage headroom in modern boards/cpus, just make sure everything you do is incremental, no big jumps. A few blue screens are nothing to be concerned about when trying to get an OC right. I've got my 4930k running @ 5.1GHz watercooled, 1.45V. Took a lot of fiddling to get everything stable, I needed a 1.2v PCH for that, and a 1.88v QPI/PLL.

Edited January 15, 2014 by Stuntz

[andrewmont 91]

1.29 is fine, On the haswell's I'd be wary about going over 1.35v even on a decent closed loop watercooler.

 

Depending on your sample luck, you might be able to hit 4.2GHz @ 1.25v, some can even do it at stock if really lucky.

 

@ 1.3v some would even be able to get 4.4 or 4.5 from it, but that's pushing it. Regardless, that seems fine.

 

There's plenty of voltage headroom in modern boards/cpus, just make sure everything you do is incremental, no big jumps. A few blue screens are nothing to be concerned about when trying to get an OC right. I've got my 4930k running @ 5.1GHz watercooled, 1.45V. Took a lot of fiddling to get everything stable, I needed a 1.2v PCH for that, and a 1.88v QPI/PLL.

So see once if I was wanting to tweak the overclock myself once I get a new PSU would I just reset the BIOS to the settings they had it at, stress test it and see if it crashes then put the CPU core voltage up incrementally until its stable?

[Stuntz 10]

It very much depends on the nature of the bluescreen.

 

If it remains as a 0x116 error, like the one you posted, then it is still down to the vid card drivers *or* slightly low pch voltage. (also a chance that the lower quality psu is not feeding the correct voltage, even the top psu's tend to be off by a few thousandths of of a volt, lower quality ones I've seen be as far off as a tenth of a volt output the vCore, and in such a case that'd kill your overclocking potential.

[andrewmont 91]

It very much depends on the nature of the bluescreen.

 

If it remains as a 0x116 error, like the one you posted, then it is still down to the vid card drivers *or* slightly low pch voltage. (also a chance that the lower quality psu is not feeding the correct voltage, even the top psu's tend to be off by a few thousandths of of a volt, lower quality ones I've seen be as far off as a tenth of a volt output the vCore, and in such a case that'd kill your overclocking potential.

Okay I really do appreciate all the help and patience, I shall beans all your replies lacking any other way to show thanks :L