Overclocked System - BSOD Google Chrome?

[syketh]

Hello! This is my second post today (Two entirely different topics though).

Earlier, just after finishing posting my first thread here about water cooling I hit a blue screen while opening my bookmarks folder which had a total of 16 tabs. Once I done this I was hit with a BSOD - 0x124 WHEA_UNCORRECTABLE HARDWARE ERROR. This isn't the first time this has happened, it has been happening ever since I built my new system, but two weeks ago I really thought I had fixed this... Obviously not.

To start, my specs are -
i7 4930K @ 4.6Ghz
Corsair H100i (Push/Pull Noctua Fans)
Asus Rampage IV Extreme
Corsair vengeance @ 2133Mhz
Corsair AX 1200i
Dual GTX 780 Ti SC
Windows 7 64-bit Home Edition

The motherboard, CPU and ram, I purchased as an overclocked bundle which was set by the retailer (Scan). I have contacted them multiple times and they have actually been very helpful, they had me try Memtest, run Aida64 and set my Vcore voltage to various settings, it was originally 1.35, now it is 1.37. I was able to run Aida64 for two hours with no hiccups and I ended it myself thinking everything was fixed at last.

It just seems weird that I am able to trigger a BSOD when opening multiple tabs in Google Chrome.. Has this happened to anyone before? I have already contacted Scan again - I am dreading pulling my system apart and sending bits and pieces back for them to tell me nothing is wrong (Since they stress tested the bundle before shipping and claim everything was fine and dandy).

Currently, I am running on stock settings and have no trouble opening the tabs now, so it must be my Overclock. Drivers are updated, Windows updates are all installed & temperatures are fine...

Any idea what could be triggering this? Thanks in advance.

 

[brarboy]

did you have look upon your temperatures? Also try lowering your multiplier and run your cpu at 4.2ghz then 4.5ghz.

 

[syketh]

did you have look upon your temperatures? Also try lowering your multiplier and run your cpu at 4.2ghz then 4.5ghz.

Yeah, temperatures are fine - 19-25oC on idle - 75-78oC stress testing.
If I lower the multiplier to say 4.2Ghz, what should I set my Vcore, 1.35?

Thanks for the response, pal.

 

[brarboy]

did you have look upon your temperatures? Also try lowering your multiplier and run your cpu at 4.2ghz then 4.5ghz.

Yeah, temperatures are fine - 19-25oC on idle - 75-78oC stress testing.
If I lower the multiplier to say 4.2Ghz, what should I set my Vcore, 1.35?

Thanks for the response, pal.

just go with 1.200 for starters. Go higher if still unstable. For more information on overclocking refer these articles:
http://www.tomshardware.com/faq/id-1845070/overclock-intel-3930k-asus-rampage-extreme.html
and;
http://rog.asus.com/51072012/rampage-motherboards/asus-rampage-iv-extreme-easy-overclocking-guide/

 

[johnbl]

most often the CPU is detecting a error in its memory controller and calling a bugcheck.
you can look at the error it detected by connecting the windows debugger to the crash dump file
and using the !errrec command on the address of the WHEA_UNCORRECTABLE HARDWARE ERROR stucture
This will tell you what the processor is complaining about.
with overclocked systems it is most often blamed on the overclock cutting the timings too close.
in a non overclocked system, there can be bugs in the CPU that the BIOS can fix in microcode or the OS can fix in a driver file. (microcode has to be loaded on each boot of the machine to fix these hardware errors)

So, make sure you have the current BIOS update and OS updates and if it still fails you need to reduce your overclock.

if you put your crash dump on the cloud with public access I will dump the error data for you.
Note: chrome is just a victim of the hardware error because it is such a pig and runs so many background processes. Odds are that it will be the current process running when your CPU throws its exception due to a internal cache memory error.

note #2: I guess you could expect this error if you overclock and allow the processor to do its turbo boost.
(be sure to turn off the turbo boost in BIOS if you are pushing your overclock)

 

[brarboy]

here are 2 videos which may help you further:
http://www.youtube.com/watch?v=nP8Q-7Mdztg
and;
http://www.youtube.com/watch?v=cdypzNu4He0

 

[syketh]

[/quotemsg=12523003,0,966493]most often filePU is detecting a error in its memory controller and calling a bugcheck.
you can look at the error it detected by connecting the windows debugger to the crash dump file
and using the !errrec command on the address of the WHEA_UNCORRECTABLE HARDWARE ERROR stucture
This will tell you what the processor is complaining about.
with overclocked systems it is most often blamed on the overclock cutting the timings too close.
in a non overclocked system, there can be bugs in the CPU that the BIOS can fix in microcode or the OS can fix in a driver file. (microcode has to be loaded on each boot of the machine to fix these hardware errors)

So, make sure you have the current BIOS update and OS updates and if it still fails you need to reduce your overclock.

if you put your crash dump on the cloud with public access I will dump the error data for you.
Note: chrome is just a victim of the hardware error because it is such a pig and runs so many background processes. Odds are that it will be the current process running when your CPU throws its exception due to a internal cache memory error.

note #2: I guess you could expect this error if you overclock and allow the processor to do its turbo boost.
(be sure to turn off the turbo boost in BIOS if you are pushing your overclock)[/quotemsg]

Thank you very, very much. I will check this out once I'm out of work. If I can't check it myself I'll Be sure to take your help dumping the files!

Thanks to everyone helping me out, I will post back with results in a few hours

 

[syketh]

most often the CPU is detecting a error in its memory controller and calling a bugcheck.
you can look at the error it detected by connecting the windows debugger to the crash dump file
and using the !errrec command on the address of the WHEA_UNCORRECTABLE HARDWARE ERROR stucture
This will tell you what the processor is complaining about.
with overclocked systems it is most often blamed on the overclock cutting the timings too close.
in a non overclocked system, there can be bugs in the CPU that the BIOS can fix in microcode or the OS can fix in a driver file. (microcode has to be loaded on each boot of the machine to fix these hardware errors)

So, make sure you have the current BIOS update and OS updates and if it still fails you need to reduce your overclock.

if you put your crash dump on the cloud with public access I will dump the error data for you.
Note: chrome is just a victim of the hardware error because it is such a pig and runs so many background processes. Odds are that it will be the current process running when your CPU throws its exception due to a internal cache memory error.

note #2: I guess you could expect this error if you overclock and allow the processor to do its turbo boost.
(be sure to turn off the turbo boost in BIOS if you are pushing your overclock)

Hey pal, out of work. I tried downloading the Windows Debugger Tool but the installation failed - not sure why. Could you give me the instructions as to how to upload my crash dump to the cloud by any chance? Haven't a clue what that means, whether to post it here or not.

 

[johnbl]

you have to create a google doc or microsoft skydrive or free account and just copy the file up to your account and put it in a directory with public access. skydrive seems to work pretty well. the memory dump file is pretty big and it is binary. too big to email

the microsoft debugging tools are here
http://msdn.microsoft.com/en-us/windows/hardware/hh852365.aspx
you should only need the standalone debugging tools for your windows version. You don't want to install the whole sdk.

it just installs them into a directory and you end up having to find the directory withe winddbg.exe in it and run it.

most often the CPU is detecting a error in its memory controller and calling a bugcheck.
you can look at the error it detected by connecting the windows debugger to the crash dump file
and using the !errrec command on the address of the WHEA_UNCORRECTABLE HARDWARE ERROR stucture
This will tell you what the processor is complaining about.
with overclocked systems it is most often blamed on the overclock cutting the timings too close.
in a non overclocked system, there can be bugs in the CPU that the BIOS can fix in microcode or the OS can fix in a driver file. (microcode has to be loaded on each boot of the machine to fix these hardware errors)

So, make sure you have the current BIOS update and OS updates and if it still fails you need to reduce your overclock.

if you put your crash dump on the cloud with public access I will dump the error data for you.
Note: chrome is just a victim of the hardware error because it is such a pig and runs so many background processes. Odds are that it will be the current process running when your CPU throws its exception due to a internal cache memory error.

note #2: I guess you could expect this error if you overclock and allow the processor to do its turbo boost.
(be sure to turn off the turbo boost in BIOS if you are pushing your overclock)

Hey pal, out of work. I tried downloading the Windows Debugger Tool but the installation failed - not sure why. Could you give me the instructions as to how to upload my crash dump to the cloud by any chance? Haven't a clue what that means, whether to post it here or not.

 

[syketh]

you have to create a google doc or microsoft skydrive or free account and just copy the file up to your account and put it in a directory with public access. skydrive seems to work pretty well. the memory dump file is pretty big and it is binary. too big to email

the microsoft debugging tools are here
http://msdn.microsoft.com/en-us/windows/hardware/hh852365.aspx
you should only need the standalone debugging tools for your windows version. You don't want to install the whole sdk.

it just installs them into a directory and you end up having to find the directory withe winddbg.exe in it and run it.

most often the CPU is detecting a error in its memory controller and calling a bugcheck.
you can look at the error it detected by connecting the windows debugger to the crash dump file
and using the !errrec command on the address of the WHEA_UNCORRECTABLE HARDWARE ERROR stucture
This will tell you what the processor is complaining about.
with overclocked systems it is most often blamed on the overclock cutting the timings too close.
in a non overclocked system, there can be bugs in the CPU that the BIOS can fix in microcode or the OS can fix in a driver file. (microcode has to be loaded on each boot of the machine to fix these hardware errors)

So, make sure you have the current BIOS update and OS updates and if it still fails you need to reduce your overclock.

if you put your crash dump on the cloud with public access I will dump the error data for you.
Note: chrome is just a victim of the hardware error because it is such a pig and runs so many background processes. Odds are that it will be the current process running when your CPU throws its exception due to a internal cache memory error.

note #2: I guess you could expect this error if you overclock and allow the processor to do its turbo boost.
(be sure to turn off the turbo boost in BIOS if you are pushing your overclock)

Hey pal, out of work. I tried downloading the Windows Debugger Tool but the installation failed - not sure why. Could you give me the instructions as to how to upload my crash dump to the cloud by any chance? Haven't a clue what that means, whether to post it here or not.

Yeah, I tried the link and tried only downloading the debug tools, still fails for some odd reason. Here is a link to my SkyDrive Dump Files

https://skydrive.live.com/redir?resid=7E2B9EA1D05ADDA8%21489

Again, I appreciate your help pal. You have been the most patient person helping me so far!

 

[syketh]

here are 2 videos which may help you further:
http://www.youtube.com/watch?v=nP8Q-7Mdztg
and;
http://www.youtube.com/watch?v=cdypzNu4He0

Thanks for the links, I checked them out, however, my CPU is the 4930K. Not sure if there's much of a difference, but I don't feel safe using the same settings for another CPU 😛

 

[johnbl]

machine info:
Manufacturer ASUSTeK COMPUTER INC.
Product RAMPAGE IV EXTREME
Version Rev 1.xx
BIOS Release Date 10/18/2013
Socket Designation LGA2011
Processor Version Intel(R) Core(TM) i7-4930K CPU @ 3.40GHz
Processor Voltage 8ah - 1.0V
External Clock 100MHz
Max Speed 4000MHz
Current Speed 3400MHz


memory
ChannelA_Dimm1 4096MB 2133MHz Corsair CMY8GX3M2B2133C9
ChannelB_Dimm1 4096MB 2133MHz Corsair CMY8GX3M2B2133C9
ChannelC_Dimm1 4096MB 2133MHz Corsair CMY8GX3M2B2133C9
ChannelD_Dimm1 4096MB 2133MHz Corsair CMY8GX3M2B2133C9



just looked at the first crash dump:

BugCheck 124, {0, fffffa8010945028, be200000, b110a}
WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 0000000000000000, Machine Check Exception
Arg2: fffffa8010945028, Address of the WHEA_ERROR_RECORD structure.
Arg3: 00000000be200000, High order 32-bits of the MCi_STATUS value.
Arg4: 00000000000b110a, Low order 32-bits of the MCi_STATUS value.

1: kd> !errrec fffffa8010945028

shows
Error : GCACHEL2_ERR_ERR (Proc 1 Bank 19)

processor 1 detected a error in its cache in bank 19 and called the bugcheck.

Note: the BIOS reported a processor voltage of 1.0 volts. (this may or may not be correctly reported)
if it is correct, i think you want to look up the proper processor voltage for you CPU. A low voltage can cause these type of errors.

all of your bugchecks show a internal cache error in the cpu

GCACHEL2_ERR_ERR (Proc 2 Bank 19)
GCACHEL2_ERR_ERR (Proc 11 Bank 19)
GCACHEL2_ERR_ERR (Proc 2 Bank 20)

-check your voltage to your processor
- confirm that you have the most up to date BIOS for your machine
(BIOS will fix known bugs in CPU with a microcode patch)
BIOS Release Date 10/18/2013
2: kd> !sysinfo cpumicrocode
Initial Microcode Version: 00000416:00000000
Cached Microcode Version: 00000416:00000000
Processor Family: 06
Processor Model: 3e
Processor Stepping: 04

 

[syketh]

My current bios is a few months old, but apparently (after checking Asus site) it's more than capable for the 4930k.

However, If I do update my bios, will I lose my overclock profiles?

 

[johnbl]

you will want to reset the profiles if you update the BIOS.
I would grab any update i could get in the chance that there is a known issue with the CPU stepping. the BIOS can fix it or a patch to the OS that windows loads can fix it. if a voltage tweak to your CPU does not fix the issue you are left with a CPU that has a cache memory error and those CPU cost some money to replace.

best bet is to check your voltage, make sure you have all CPU enhancements turned off (turn off intel speed step and intel turbo boost)

My current bios is a few months old, but apparently (after checking Asus site) it's more than capable for the 4930k.

However, If I do update my bios, will I lose my overclock profiles?

 

[syketh]

you will want to reset the profiles if you update the BIOS.
I would grab any update i could get in the chance that there is a known issue with the CPU stepping. the BIOS can fix it or a patch to the OS that windows loads can fix it. if a voltage tweak to your CPU does not fix the issue you are left with a CPU that has a cache memory error and those CPU cost some money to replace.

best bet is to check your voltage, make sure you have all CPU enhancements turned off (turn off intel speed step and intel turbo boost)

My current bios is a few months old, but apparently (after checking Asus site) it's more than capable for the 4930k.

However, If I do update my bios, will I lose my overclock profiles?

A while ago, I was given advice by my retailer to change my Vcore from 1.35 to 1.37 and I ran Aida64 stress test without any problems for a couple of hours - And I thought I sorted the problem out, but it seems it hasn't, so I think the Vcore isn't the issue here.

I would like to update my bios but I will wait to hear back from Scan's support team just incase they advise against it or want to call it back. I hope I don't have to replace my CPU, I've only purchased it about a month ago so I reckon if there are any problems with it, they'll replace it free of charge.

Thanks for looking through the dump files - I'll send Scan a copy of them too just so they know.
For now, I may try and play about with the multiplier and set it to maybe 4.4Ghz.

I'll be sure to keep this updated, just incase anyone else in the future has this problem!

 

[brarboy]

here are 2 videos which may help you further:
http://www.youtube.com/watch?v=nP8Q-7Mdztg
and;
http://www.youtube.com/watch?v=cdypzNu4He0

Thanks for the links, I checked them out, however, my CPU is the 4930K. Not sure if there's much of a difference, but I don't feel safe using the same settings for another CPU 😛

that is just to provide you an idea, how you can overclock. For your cpu, linked forum and same mobo. I hope that helps

 

[syketh]

you will want to reset the profiles if you update the BIOS.
I would grab any update i could get in the chance that there is a known issue with the CPU stepping. the BIOS can fix it or a patch to the OS that windows loads can fix it. if a voltage tweak to your CPU does not fix the issue you are left with a CPU that has a cache memory error and those CPU cost some money to replace.

best bet is to check your voltage, make sure you have all CPU enhancements turned off (turn off intel speed step and intel turbo boost)

My current bios is a few months old, but apparently (after checking Asus site) it's more than capable for the 4930k.

However, If I do update my bios, will I lose my overclock profiles?

Good news and bad news.

Good news is that I updated my bios to the latest version and tested my overclock out, I tried opening multiple tabs on Google Chrome (Which caused BSOD's before) and it no longer happens.

Bad news is that after trying to stress test the overclock (Which was stable on Aida64 previously) I end up with a BSOD almost instantly with 1.37 Vcore and about 1 minute in with a 1.35 Vcore.

https://skydrive.live.com/?mkt=en-GB#cid=7E2B9EA1D05ADDA8&id=7E2B9EA1D05ADDA8%21489

I have uploaded two more dump files - I think they are caused by the same thing it seems. Also, I added a few screen shots of my Bios when I am using the overclock setting - If anyone could take a quick look and check if everything's in order. I am using the retailers settings.

Thanks.

 

[brarboy]

reset everything to default. Go to voltage made and set to manual. Then try stress testing and chrome tab thing. See if everything goes on smooth.

 

[johnbl]

same as before:

ection 2 : x86/x64 MCA
-------------------------------------------------------------------------------
Descriptor @ fffffa80108bf138
Section @ fffffa80108bf2c0
Offset : 664
Length : 264
Flags : 0x00000000
Severity : Fatal

Error : GCACHEL2_ERR_ERR (Proc 2 Bank 20)
Status : 0xbe200000000b110a
Address : 0x0000000224805400
Misc. : 0xe0fe289088800086
2: kd> !sysinfo cpumicrocode
Initial Microcode Version: 00000416:00000000
Cached Microcode Version: 00000416:00000000
Processor Family: 06
Processor Model: 3e
Processor Stepping: 04


both crashes were on bank 20

 

[johnbl]

You have CPU PLL voltage is 1.796 V (auto)

VccPPL on the CPU is 1.7V according to intel CPU spec.(min 1.624V max 1.77V)
try and reduce your CPU PLL voltage to 1.65 to 1.70V and see if your cache errors go away

 

[syketh]

You have CPU PLL voltage is 1.796 V (auto)

VccPPL on the CPU is 1.7V according to intel CPU spec.(min 1.624V max 1.77V)
try and reduce your CPU PLL voltage to 1.65 to 1.70V and see if your cache errors go away

Just lowered my multiplier to 44 and set the CPU PLL voltage to 1.65 like you recommended. Managed to boot alright, tried opening tabs on Google, perfect. Tried stress testing the overclock, fine so far. Only tested it for 10 minutes, but it's 10x times longer than it lasted previously. I think that may just have fixed it... I'm just curious as to why my PLL voltage was set to auto by the retailers, surely they had it working perfect when they tried it.

I will continue using the stock profile for now and wait until Scan get back to me with a go ahead or another way around using the original Overclock profile they set up for me.

By the way, if I was to just continue with the current overclock - 4.4Ghz @ 1.37 Vcore, should I try lowering the Vcore - or is it alright? I don't want to add too much incase I damage something.

Again, I will keep posting if anything comes up and I will let you know what Scan recommends. Thank you so much guys for all the help so far!

 

[brarboy]

you can go below 1.37 volts, once you surely know that you are stable at that current oc settings. Might check at 1.300 volts, still stable, go below. Good luck

 

[syketh]

you can go below 1.37 volts, once you surely know that you are stable at that current oc settings. Might check at 1.300 volts, still stable, go below. Good luck

Thanks, i'll give it a shot

Do you think leaving the CPU PLL at 1.65 should be alright too, rather than Auto like Scan had set it to?

Cheers buddy.

 

[johnbl]

I think the auto value was too high for your other settings. It was set with the assumption that you would be turning off some other Intel fucntions that you left on. (c states, hyperthreading, ...)

you can go below 1.37 volts, once you surely know that you are stable at that current oc settings. Might check at 1.300 volts, still stable, go below. Good luck

Thanks, i'll give it a shot

Do you think leaving the CPU PLL at 1.65 should be alright too, rather than Auto like Scan had set it to?

Cheers buddy.

 

[syketh]

Scan got back to me today while at work and suggested calling the bundle back as they suspect a problem with the overclock or one of the components.

So I'll revert the bios back and ship it off. I'll post back in a week (or however long they take to test it and send it back!) To tell you how it goes.