BSODs/Crases with new build

[jaronmartin]

I recently built a new desktop. Everything seems to work fine, except when it decides to crash (anywhere from every few days to multiple times per day). I was hoping someone here might have some suggestions.

Specs:
Monitor 1: 23 Inch HP S2331
Monitor 2: 23 Inch LG IPS231
Case: Azza Hurrican 2000 CSAZ-2000 R
Processor: Intel i7 2600k
MoBo: ASRock Z68 Extreme3 Gen3
GPU: Radeon HD 6990
SSD (OS): 120GB Mushkin MKNSSDCR120GB-DX R
HDD (Data): 1.5TB WD Caviar Black WD1502FAEX
RAM: 16GB Corsair CML8GX3M2A1600C9 R
Optical 1: Lite-On Blu-Ray Burner IHBS112-29
Optical 2: HP DVD Burner 1270I LS SATA R
PSU: Corsair AX1200
CPU Cooler: Cooler Master Hyper 212+ RR-B10-212P-G1
OS: Windows 7 Professional x64

Symptoms:
Generally the computer will become progressively more unresponsive, quickly leading to a BSOD.
Here is a video: http://www.youtube.com/watch?v=MDV5uSf1yZA

Firefox becomes unresponsive.
I can open 'My Computer', however I can't access any drives.
BSOD: STOP: 0x00000000F4 A process or thread crucial to system operation has unexpectedly exited or been terminated.
Notice that the BSOD never initializes the disk and is unable to perform a crash dump.
During BSOD the hard drive light is solid on.

I ran memtest86 v4.0a overnight. I woke up to a corrupted display:
http://www.youtube.com/watch?v=b5ACPw50a8k

I suspected a graphics card problem, so I removed the HD6990 and began using the integrated graphics.

I started memtest86 v4.0a again at about 2:30AM.
At 7:43AM the screen was fine:
Alt Link
At 12:31 PM the screen was corrupted again:
Alt Link
http://www.youtube.com/watch?v=qKEAIgto2E4

Another BSOD I am receiving is 0x7A - KERNEL_DATA_INPAGE_ERROR
Alt Link

My CPU Temperature stays around 34C when idle and around 55C at full load (Prime95).

I'm not sure what to try next...any help would be greatly appreciated!

 

[rolli59]

Try running with your ram just two sticks at a time.

 

[jaronmartin]

Try running with your ram just two sticks at a time.

MemTest did the same thing with each two stick set...
I also physically disconnected both hard drives and the DVD drive that wasn't being used to boot MemTest before running, so those can't be the culprit. Any other ideas?

 

[ko888]

Can you post your OC Tweaker screens, DRAM settings and Voltage Control settings displayed in the ASROCK UEFI SETUP UTILITY?

 

[jaronmartin]

Can you post your OC Tweaker screens, DRAM settings and Voltage Control settings displayed in the ASROCK UEFI SETUP UTILITY?

I decided to run Memtest86+ v4.20 to see if it runs into the same problem. It has been running for about 9 hours so far, so I'm going to let it continue running overnight. I can say that I haven't overclocked or changed any default settings. Tomorrow I will post the screens you asked for.

 

[jaronmartin]

Can you post your OC Tweaker screens, DRAM settings and Voltage Control settings displayed in the ASROCK UEFI SETUP UTILITY?

I ran memtest86+ for over 24 hours/10 passes and no problems occurred.

Here are the screens you requested:
OC Tweaker: Alt Link
DRAM Settings 1: Alt Link
DRAM Settings 2: Alt Link
Voltage Control: Alt Link

I re-downloaded and re-burned memtest86 and will start that now to see if there was a problem with my copy since memtest86+ didn't have any issues.

 

[ko888]

I ran memtest86+ for over 24 hours/10 passes and no problems occurred.

Here are the screens you requested (sorry, the image host i used previously is down at the moment):
OC Tweaker
DRAM Settings 1
DRAM Settings 2
Voltage Control

I re-downloaded and re-burned memtest86 and will start that now to see if there was a problem with my copy since memtest86+ didn't have any issues.

I use the exact same brand and model of memory modules that you are using but with a different motherboard (i.e. ASUS P8Z68-V PRO). I also had to set the Memory Frequency manually to DDR3-1600MHz because the [Auto] settings would only pick up the JEDEC #4 entry in the SPD timings table and not the XMP-1600 entry.

What I notice immediately from your OC Tweaker\DRAM Configuration is that your DRAM Frequency is operating at DDR3-1333.

First thing to try is to change the Load XMP Setting from the default value of [Auto] to [Profile 1]. If this setting change does not cause the DRAM Frequency setting to display the correct DDR3-1600 speed after a reboot then you'll have to set the DRAM Frequency setting manually.

With the Load XMP Setting set to [Auto] the DRAM Frequency setting of [Auto] is only picking up the JEDEC #4 memory timings from the memory modules' timings table in the SPD EEPROM chip. You will need to change this setting from [Auto] to [DDR3-1600] if you want to operate the memory modules at their rated speed, since that's what you paid for.

If your DIMMs are unable to operate at their rated speed then they are defective and should be RMA'd.

 

[jaronmartin]

I use the exact same brand and model of memory modules that you are using but with a different motherboard (i.e. ASUS P8Z68-V PRO). I also had to set the Memory Frequency manually to DDR3-1600MHz because the [Auto] settings would only pick up the JEDEC #4 entry in the SPD timings table and not the XMP-1600 entry.

What I notice immediately from your OC Tweaker\DRAM Configuration is that your DRAM Frequency is operating at DDR3-1333.

First thing to try is to change the Load XMP Setting from the default value of [Auto] to [Profile 1]. If this setting change does not cause the DRAM Frequency setting to display the correct DDR3-1600 speed after a reboot then you'll have to set the DRAM Frequency setting manually.

With the Load XMP Setting set to [Auto] the DRAM Frequency setting of [Auto] is only picking up the JEDEC #4 memory timings from the memory modules' timings table in the SPD EEPROM chip. You will need to change this setting from [Auto] to [DDR3-1600] if you want to operate the memory modules at their rated speed, since that's what you paid for.

If your DIMMs are unable to operate at their rated speed then they are defective and should be RMA'd.

Thanks, I hadn't even noticed that my ram was clocked low. I changed the Load XMP Setting to Profile 1 and after reboot it showed the RAM clocked at 1600.

Unfortunately I am still experiencing the same problems even with the new Memtest86 v4.0a image I downloaded.

 

[jaronmartin]

Here's an update from more Memtests I've run.

I decided to try Memtest86 v3.5b which comes with v4.0a. I was looking at the configuration and when I selected 'Default Tests' from the menu I received this error: (Unexpected Interrupt - Halting CPU 0).
Alt Link

At 11:37 AM I rebooted and let the test run without messing with anything. Sometime between 4:10 PM and 6:47 PM I received another Unexpected Interrupt - Halting CPU 0 error.

Alt Link

I then decided to try Memtest86+ v4.20 again. After running for 40 hours there were no problems. However, when I checked back at the 43:09:09 mark it said it found a bad memory location.
Alt Link

I'm a little confused, because if it's a bad stick of memory then the previous Memtest86 v4.0a test with two sticks at a time should have narrowed it down. Is there a problem with Memtest86 v4.0a/3.5b? Or could a CPU or MoBo problem appear as bad memory? Or are there multiple problems?

 

[ko888]

Memtest86 v3.5b was put out two years before the Sandy Bridge processors were launched. Your pictures even shown that it can't even identify the i7-2600K's cache levels and sizes.

Memtest86 v4.0a is a massive rewrite of the utility from Memtest86 v3.5b so it may/will have introduced some new bugs.

The integrated memory controller resides on the i7-2600K's processor die so any problems with an unstable CPU may/will manifest itself as defective memory.

 

[jaronmartin]

I continued running Memtest+ (mainly because I don't know what else to run). I tested sticks 1/3 in slots 1/3 for 67 hours with no errors. I then tested sticks 2/4 in slots 2/4 for 70 hours with no errors. I just started it with all 4 sticks again.

Is there any way to explicitly test the CPU's memory controller, or are there any other tests that would help narrow down the problem? I purchased all parts from NewEgg, but as far as I know I can't do an RMA unless I know exactly which part is malfunctioning. I'm pretty much at a loss and don't really know what to do from here...

 

[ko888]

One important fact to note is that not all errors reported by Memtest86+ is caused by the memory itself. Be aware that Memtest86+ tests the entire computer memory subsystem which includes CPU, cache, memory controller (either on CPU or motherboard), and the motherboard itself.

If errors are reported by Memtest86+, this does not necessarily mean a problem with your memory.

Other components mentioned above (i.e. cache, memory controller) as well as incorrect BIOS settings can also cause Memtest86+ to report errors.

Overclocking far beyond specifications can also cause Memtest86+ to report errors.

It is always a good idea to run all components including memory within the manufacturer's specifications to rule out other causes of your stability problems.

Are you using UEFI BIOS Version 1.10 or 1.00? If you're using Version 1.00 maybe it's time to update to Version 1.10.