Memory Fail?

[gruxer]

Hi. I have custom built computer and it runned just fine since first couple months. Then problems started. I started to receive random bluescreen errors. Almost everyday. I have my comp on warranty so i returned it and they detected ram failure and replaced it for brand new. After couple weeks the same thing happened i returned it and they detected memory failure again and that story repeated good 4 times and problem came back after they replaced all 4 modules of my memory. I have bluescreen showing up randomly does not matter whether im playing game or my comp is just running. I looked up 100 's different forums and nothing seems to help. I just found general idea of component failure nothing detail. I m trying to firgure out what can cause those kind of failures and how can i possibly fix it. Today i installed Windbg application on my computer to display dmp files. I got the info out of it but i dont understand it. I wish you can help me. My comp : windows vista 64x, 4 gb dual channel ddr2 ram 1080, nvidia geforce 9800 gtx, intel quad core processor 2,4, sound blaster fatal1ty championship sound card, asus IE5 pro motherboard. (memory cooling fans installed, average temperature of video card: 62 C, cpu 45 C, cpu cores 30 C, motherboard 35 C, and info i got from Windbg:

*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

IRQL_NOT_LESS_OR_EQUAL (a)
An attempt was made to access a pageable (or completely invalid) address at an
interrupt request level (IRQL) that is too high. This is usually
caused by drivers using improper addresses.
If a kernel debugger is available get the stack backtrace.
Arguments:
Arg1: 0000000000000028, memory referenced
Arg2: 0000000000000002, IRQL
Arg3: 0000000000000000, bitfield :
bit 0 : value 0 = read operation, 1 = write operation
bit 3 : value 0 = not an execute operation, 1 = execute operation (only on chips which support this level of status)
Arg4: fffff8000208e740, address which referenced memory

Debugging Details:
------------------


READ_ADDRESS: GetPointerFromAddress: unable to read from fffff800021fa0b0
0000000000000028

CURRENT_IRQL: 2

FAULTING_IP:
nt!MiFindNodeOrParent+0
fffff800`0208e740 48f7412800ffffff test qword ptr [rcx+28h],0FFFFFFFFFFFFFF00h

CUSTOMER_CRASH_COUNT: 6

DEFAULT_BUCKET_ID: COMMON_SYSTEM_FAULT

BUGCHECK_STR: 0xA

PROCESS_NAME: svchost.exe

TRAP_FRAME: fffff980199265e0 -- (.trap 0xfffff980199265e0)
NOTE: The trap frame does not contain all registers.
Some register values may be zeroed or incorrect.
rax=0000000000000000 rbx=0000000000000000 rcx=0000000000000000
rdx=0000000000000000 rsi=0000000000000000 rdi=0000000000000000
rip=fffff8000208e740 rsp=fffff98019926778 rbp=0000000000000000
r8=fffff980199267c0 r9=fffff880036a8ae8 r10=fffffa80040240a8
r11=fffffa8004024010 r12=0000000000000000 r13=0000000000000000
r14=0000000000000000 r15=0000000000000000
iopl=0 nv up ei ng nz na pe nc
nt!MiFindNodeOrParent:
fffff800`0208e740 48f7412800ffffff test qword ptr [rcx+28h],0FFFFFFFFFFFFFF00h ds:00000000`00000028=????????????????
Resetting default scope

LAST_CONTROL_TRANSFER: from fffff8000204db33 to fffff8000204dd90

STACK_TEXT:
fffff980`19926498 fffff800`0204db33 : 00000000`0000000a 00000000`00000028 00000000`00000002 00000000`00000000 : nt!KeBugCheckEx
fffff980`199264a0 fffff800`0204ca0b : 00000000`00000000 fffffa80`045d2b98 00000000`00000000 fffff800`02057080 : nt!KiBugCheckDispatch+0x73
fffff980`199265e0 fffff800`0208e740 : fffff800`0207d0e7 00000000`00000000 fffff800`020d5c7a 42506650`02000000 : nt!KiPageFault+0x20b
fffff980`19926778 fffff800`0207d0e7 : 00000000`00000000 fffff800`020d5c7a 42506650`02000000 fffffa80`045d2000 : nt!MiFindNodeOrParent
fffff980`19926780 fffff800`020d585e : 00000000`00000000 00000000`000007ff 00000000`000018c0 fffff800`0213d2a4 : nt!MiLocateAddressInTree+0x17
fffff980`199267b0 fffff800`020e45ee : 00000000`000292d8 fffffa80`045d2b98 fffffa80`08b66bb0 00000000`00000000 : nt!MiIdentifyPfn+0x77b
fffff980`19926850 fffff800`023b3705 : fffffa80`045d2000 fffff980`19926ca0 fffff980`19926918 00000000`00000000 : nt!MmQueryPfnList+0x13e
fffff980`19926890 fffff800`02238cbc : 00000000`00000000 00000000`00000000 fffffa80`045d2000 00000000`00000001 : nt!PfpPfnPrioRequest+0x115
fffff980`199268e0 fffff800`022c0a9c : 00000000`00000000 00000000`00000000 00000000`021feb20 fffff980`19926901 : nt!PfQuerySuperfetchInformation+0x1db
fffff980`19926950 fffff800`0204d833 : fffffa80`08b66bb0 00000000`05dd2ff0 00000000`00000008 00000000`00000000 : nt!NtQuerySystemInformation+0x11aa
fffff980`19926c20 00000000`770105da : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : nt!KiSystemServiceCopyEnd+0x13
00000000`021fea78 00000000`00000000 : 00000000`00000000 00000000`00000000 00000000`00000000 00000000`00000000 : 0x770105da


STACK_COMMAND: kb

FOLLOWUP_IP:
nt!MiFindNodeOrParent+0
fffff800`0208e740 48f7412800ffffff test qword ptr [rcx+28h],0FFFFFFFFFFFFFF00h

SYMBOL_STACK_INDEX: 3

SYMBOL_NAME: nt!MiFindNodeOrParent+0

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: nt

DEBUG_FLR_IMAGE_TIMESTAMP: 49ac932a

IMAGE_NAME: memory_corruption

FAILURE_BUCKET_ID: X64_0xA_nt!MiFindNodeOrParent+0

BUCKET_ID: X64_0xA_nt!MiFindNodeOrParent+0

Followup: MachineOwner
---------


Thank You for all your help.

[ronallandottk]

From the information you've posted it appears that you really have a memory problem.

IRQL_NOT_LESS_OR_EQUAL (a)
An attempt was made to access a pageable (or completely invalid) address at an
interrupt request level (IRQL) that is too high. This is usually
caused by drivers using improper addresses.

In simple terms this means that the CPU tried to access a memory location (address) that is invalid, meaning that during startup the OS has identified this memory as being available, but when it tried to access it, it wasn't there anymore, causing your PC to experience a BSOD.

It's not that easy pinpointing the culprit responsible for your memory failures, but here are some things you can explore:

1. Maybe you need a BIOS update. The present BIOS may be bugged causing it to experience memory failures with your specific memory configuration;

2. You've mentioned using 4 modules of RAM for a total of 4 GB. Some motherboards tend to be less stable when all of the slots are populated. Try using 2 sticks of 2 GB RAM instead of 4 1 GB sticks. I've personally seen this happen on a number of motherboards, some motherboards just don't like being full. Is it a defect? Not necessarily. All the slots work, it just tends to be unstable when all slots are full.

You can also try removing 2 of the sticks temporarily and see if stability improves. Try using one stick in single-channel mode just to isolate the problem.

3. Are you using a reputable brand of memory? Sometimes this also has an effect. I've had an MSI motherboard that just didn't work with 2 1 GB sticks of generic RAM (which worked without issue in another PC), but the MSI board worked flawlessly when I installed 2 branded 1 GB sticks (Kingston). Try changing to a different brand. It works sometimes;

4. Are you running your memory at it's rated speed? You've mentioned using "ddr2 ram 1080". If I'm not mistaken most common DDR2 modules run at 533 MHz effective (1066 Dual Channel), 666 MHz (1333), and 800 MHz (1600). If you're running at a non-standard 540 MHz (1080), you may experience memory failure if your RAM isn't up to snuff. That being said, most branded (and even non-branded) RAM can handle a mere 7 MHz (actually 3.5 MHz) overclock;

5. Are your RAM voltages standard? Some motherboards, especially those intended for overclockers have BIOS options for tweaking the RAM voltage. It's possible it's set wrong causing your memory to fail;

6. Are your BIOS memory settings (CAS, RAS, etc.) set correctly? I just had to ask, because if they're set too aggressively, you'll experience system instability. Having BIOS memory settings set wrong are not really know to ruin RAM though, it just makes the system unstable;

7. Are your modules matched? If they're rated at different speeds, and their speeds are set manually to run at the faster module's rating, the slower rated module may be stressed leading to failure;

8. Are the modules exposed to excessive heat? Some less than ideal designs have RAM sticks directly in the path of hot air coming from the CPU heatsink/fan, GPU fan, or power supply. If they're continuously exposed to heat they can die prematurely;

9. Lastly, if you considered all of the above and your board keeps killing RAM modules, the motherboard itself could be defective. It isn't unheard of for some motherboards to have a flaky RAM voltage regulator causing it to fry any RAM installed on it. I have to admit it's been a while (several years?) since I last encountered this phenomenon, and it usually happens to no-name brand boards which were popular in the late 90s. Your board is an Asus, so I doubt it if it has this problem. Then again, you never know;

10. This is a longshot, but you never know. Some assemblers don't really take static electricity into account and carelessly handle electronic components such as RAM, video cards or motherboards. Static electricity can kill such components, and it doesn't necessarily kill them immediately, only failing after some time.

http://www.pcworld.com/article/82184/avoid_static_damage_to_your_pc.html

11. Do you use a UPS or a surge suppressor? Voltage fluctuations from plugging directly to a wall outlet have been known to ruin sensitive electronic components such a computer parts. Nowadays most modern power supplies can take the stress, but it's also a possibility.

To test your RAM for defects try downloading Memtest86+ and burn it to a CD.

http://www.memtest.org/

Boot from the CD and run the tests. If there are any flaws this should detect it.

I hope you find this long-winded reply helpful. I've diagnosed RAM-related problems on other PCs before, mine and others, and it can be really quite annoying dealing with BSODs.

[gruxer]

Hello. I appreciate your immediate response. Your hint about changing ram for different one makes sense. It is worth a try and I’m going to return the ram and exchange it for another module 2x 2gb. If the new ram fails as well it means that my motherboard is responsible for that. First I thought it might be my electricity in my house. The computer was plugged into the outlet without ground wire. But after all I don’t think that is what causes problem. I purchased power surge protector with battery backup and all other protectors and the problem still exists. Then I thought that my computer might simply overheat. I installed everest on my computer and is started to monitor all components. I even opened the window so it was really cool in my room I took the computers case off and I installed additional cooling system on memory module. After all that I checked temperature of components and my video card reached 63 C which seems to be normal temperature because geforce 9800 operate always hot. Then I checked my processor and its temperature is 43 which should be normal as well. All other were cool including motherboard (35C) I also checked my ram using my fingers because I don’t have temp sensor in my ram and it seemed to be very cool. After all I don’t think it is overheating. I mentioned that I use 1080 mhz of memory I apologize for mistake, it is regular 1066 Mhz I thought the standard is 1080. So I never overclocked anything and bios settings are set on default because I never changed anything. Bios update seems to be reasonable and I’m definitely going to try it. I also check my ram using memory test and it told me that some serious problems with ram were detected and I should contact with technician to solve issue. Of course I know what they are going to do. They will prove my point that memory failed and they will replace memory but the problem still exists and they never try to detect what causes the memory fail. I will be talking with them today. One last possible causing issue matter is that since I purchased my computer it has been good 2 years. It was perfectly stable at first but after some time I remember I purchased new cooling for my cpu. I had to take all motherboard out of the tower, take all components of, and unscrew mobo because cpu had screws that I had to install on the back of my motherboard. I did everything that was necessary plus I installed isolating cover on the back so screws don’t touch mobo. Then I putted it back again and everyting seemed to be normal but I still have doubts about if I did everything correctly and if I accidentally did not damage my motherboard. Those are all possible things that may cause memory problem. Is there anything critical from the list above that may cause the problem? Thank you For your all Help Gruxer.

[ronallandottk]

I guess we can rule out heat, power fluctuations and overclocking issues in your case.

When replacing/reinstalling a motherboard (or any other internal component), it's always advisable to be wary of static electricity and you should discharge yourself (touch any grounded unpainted metal surface first) before physically handling them. Assuming nothing as extreme as a spark was generated when you were handling your motherboard, I guess we can rule that out as well.

It may be a good idea to inspect your motherboard's mounting points, make sure none of your screws are overtightened. While there may not be any metal to metal contact with your case, it's always a good idea to leave some clearance. Things expand as they heat up, what doesn't touch when cool may touch when warm. Look around if you have accidentally left any screws around, these could cause shorts.

Since we've basically gone through everything, if you want to tackle the RAM problem conservatively, here's what I would do if it were my PC:

1. I'd update the BIOS if one were available. Instabilities can often be cured by a BIOS update;
2. I'd set the RAM related BIOS settings to their defaults. You can always set them back for performance later on;
3. I'd spray the RAM slots and the RAM connecting edges with a good contact cleaner (the fast drying kind, NOT petroleum based ones like WD-40) like this:

http://www.polywater.com/typefd.html

You really don't have to do this of course, but I'd recommend it, just to make sure there isn't any invisible debris or gunk shorting out your memory.

4. I'd install only ONE module of RAM. Yes, you'll have less memory, and run in single channel mode only, but we're doing a process of elimination here.

5. I'd run the PC for a few days and see what happens. If nothing happens, I'd put another stick in. And wait another few days.

6. Keep populating the the slots every few days until all the slots are full, or you reach the amount of memory you intend to install, or until you experience a BSOD. If you do experience a BSOD, make note of the unstable RAM slot, and use another.

If you're going for 4 GB, I'd recommend 2x2GB instead of 4x1GB. Don't go replacing all your RAM first, wait until we're sure that your board won't ruin the new modules.

Hopefully you'll end up with a more stable system.

[gruxer]

Hi once again Thank You for your response. I talked to technician today (different one then before) and he offered me to replace ram for different brand ram. Based on your response that 2x2 is better than 4x1 i suggested him to replace my ram using 2 modules 2gb each and he agreed with me. I gave computer to the service and in few days i should have it back with new memory installed. Hopefully this will solve issue, if not, well i will be forced to take care of my motherboard. Since all my components are on warranty I don't bother if any of new installed components will fail. I just hope it will work this time and i will inform you about that as soon as I will get my computer back from service. Thank You for all your time and great help! Gruxer

[ronallandottk]

Good luck, hope your memory problems get sorted out.