Solved BSOD and crashes playing BF3, happens with 2 different GPUs

CaptainDju · May 23, 2012

Hi all, and first of all thank you in advance, even if you don't manage to solve my issue

First things first, what happens:

I have played BF3 since it was out, and a couple of weeks ago it started crashing more and more often. Now I can barely play for 5 minutes. It also occurs with Mass Effect 3, but I haven't tested other games.

Most of the crashes no longer lead to BSODs, I "only" need to Ctrl-Alt-Del out of the game, but sometimes I'll have a cold BSOD.

Note that I had for a while a second GPU from the same model (except 1Gb of RAM instead of 2) and switching between them did not solve the issue.

My Windows is a Home Premium x64 version, which I bought with the computer and installed myself. I built the PC (and installed the OS) in early November 2011.

I hope you guys can acces the specs I entered in my profile, otherwise I'll copy/paste them.

I have run the BSOD Dump & System File Collection app, and attached the zipped result, let me know if anything is missing. Note that some dumps might be old, as I used to have a malfunctioning GPU which was replaced by the store where I bought my computer.

I also attached the perfmon html file as advised.

I'm running AVG Antivirus Free, which is always turned on, and Windows Update are made on a regular basis.

Here's a quick summary of what I tried so far:

- Uninstalled everything Logitech-related
- Uninstall/Driver Sweeper/Re-install NVIDIA drivers several times, with WHQL and BETA drivers
- Installed the last NVIDIA drivers from MSI's website
- Unplugged every USB device except mouse & keyboard
- Uninstall and re-install both Origin and BF3
- Uninstalled several non-critical applications

I am at a loss, not being able to play my games is really no fun, so I hope you'll be able to help

Once again, thanks in advance!

writhziden · May 23, 2012

Code:

Start Menu\Programs\MSI Afterburner	DJU\Dju:Start Menu\Programs\MSI Afterburner	DJU\Dju

The above has known problems with Windows 7. Recommend removing MSI Afterburner.

AVG does not seem to be starting on startup. Was that intentional?

Crashes are pointing to memory problems or driver conflicts.

If you are overclocking any hardware, please stop.
Run the boot version of Memtest86+ paying close attention to Parts 2 and 3 of the tutorial. Also, in case Memtest86+ misses anything and comes up with no errors, run the extended version of the Windows Memory Diagnostics Tool for at least five passes. These you may want to run overnight since they take a long time to complete (run them an hour before bed each of the next two nights and check before going to sleep that they are still running).

Warning

Before you proceed with the following, answer these two questions: Are you still under warranty? Does your warranty allow you to open up the machine to check hardware? If you are unsure of the answers to these questions, contact your system manufacturer. :warn:WARNING: The steps that follow can void your warranty!!!

For Part 3: If You Have Errors: If you swap any memory components, follow these steps for ESD safety:
1. Shut down and turn off your computer.
2. Unplug all power supplies to the computer (AC Power then battery for laptops, AC power for desktops)
3. Hold down the power button for 30 seconds to close the circuit and ensure all power drains from components.
4. Make sure you are grounded by using proper grounding techniques, i.e. work on an anti-static workbench, anti-static desk, or an anti-static pad. Hold something metallic while touching it to the anti-static surface, or use an anti-static wristband to attach to the anti-static material while working. If you do not have an anti-static workbench, desk, or pad, you can use your computer tower/case by finding a metal hold in it, such as a drive bay.
Once these steps have been followed, it is safe to remove and replace components within your computer.
An underlying driver may be incompatible\conflicting with your system. Run Driver Verifier to find any issues. To run Driver Verifier, do the following:
a. Backup your system and user files
b. Create a system restore point
c. If you do not have a Windows 7 DVD, Create a system repair disc
d. In Windows 7:
- Click the Start Menu
- Type verifier in Search programs and files (do not hit enter)
- Right click verifier and click Run as administrator
- Put a tick in Create custom settings (for code developers) and click next
- Put a tick in Select individual settings from a full list and click next
- Set up the individual settings as in the image and click next
  View attachment 213245
- Put a tick in Select driver names from a list
- Put a tick next to all non-Microsoft drivers.
- Click Finish.
- Restart your computer.
If Windows cannot start in normal mode with driver verifier running, start in safe mode. If it cannot start in safe mode or normal mode, restore the system restore point using http://www.sevenforums.com/tutorials/700-system-restore.html OPTION TWO.

Thanks to zigzag3143 for contributing to the Verifier steps.
If you are unable to start Windows with all drivers being verified or if the blue screen crashes fail to create .dmp files, run them in groups of 5 or 10 until you find a group that causes blue screen crashes and stores the blue screen .dmp files.
The idea with Verifier is to cause the system to crash, so do the things you normally do that cause crashes. After you have a few crashes, upload the crash reports for us to take a look and try to find patterns.

When you are ready to disable Verifier: Start Menu -> All Programs -> Accessories -> Right click Command Prompt -> Run as administrator -> Type the following command and then Enter:
verifier /reset
-> Restart your computer.

zigzag3143 · May 23, 2012

Related to asmtxhci.sys XHCI Host Controller Driver from ASMedia Technology Inc. Yours is from 2010. I would re-install the newset available.

ASMedia Technology Inc. ????

CaptainDju · May 23, 2012

writhziden said:
Code:

Start Menu\Programs\MSI Afterburner DJU\Dju:Start Menu\Programs\MSI Afterburner DJU\Dju

The above has known problems with Windows 7. Recommend removing MSI Afterburner.

Okay, I can certainly do that, even though I'm currently using it to underclock my GPU which is factory overclocked.

writhziden said:
AVG does not seem to be starting on startup. Was that intentional?

It certainly is not intentional, but the weird thing is it DOES start with Windows :huh: I'm puzzled on this one.

writhziden said:
Crashes are pointing to memory problems or driver conflicts.

I'll run Windows Memory Diagnostic Tool this night, will post the results

writhziden said:
If you are overclocking any hardware, please stop.

See above, my GPU is overclocked by default, how can I circumvent that?

writhziden said:
An underlying driver may be incompatible\conflicting with your system. Run Driver Verifier to find any issues. To run Driver Verifier, do the following: [snip]

I'll see tomorrow what the result of the above is, and if I still get crashes I'll do the Driver Verifier bit

zigzag3143 said:
Related to asmtxhci.sys XHCI Host Controller Driver from ASMedia Technology Inc. Yours is from 2010. I would re-install the newset available.

ASMedia Technology Inc. ????

As I don't seem to find a more recent one, I have currenty uninstalled (via the Windows Programs and Features application) the USB 3.0 driver. Do you think there might be leftover files and/or registry entries I should delete?

Anyway, going to sleep now, thanks a lot for the help so far, I'm impressed!

writhziden · May 23, 2012

You'd better make sure AVG is running properly. It was not listed in your running tasks for your crash report... That was why I asked. That type of activity where an antivirus seems to be running but actually is not usually indicates a virus or malware is present.

For the factory overclocked GPU: You can use MSI Afterburner, but make sure you have the most recent version. Older versions caused more issues than the newer versions.

Yours should be okay:

Code:

RTCore64	fffff880`062f1000	fffff880`062f7000	Tue Sep 06 06:24:50 2011 (4e661112)	00008bee		RTCore64.sys

You could try re-installing the USB 3.0 drivers. It could just be that yours were corrupted at the time and caused the crash.

For the Windows Memory Diagnostic tool: make sure you also run Memtest86+. Memtest86+ stresses the memory is more ways than the Windows tool and is more thorough as a result.

EDIT: Ah, nevermind, AVG is running as it should. It is usually listed in Program Files, but yours is in Utils. That threw me for a loop.

CaptainDju · May 24, 2012

Okay, here's a recap of what I've done and what's to come, will edit this post with the results as they come in:

- I'm keeping MSI Afterburner 2.2.1 which is the latest version so that I can keep my GPU de-overclocked. My current settings are the stock ones
- AVG is confirmed as fine, keeping it active, and confirmed as up to date
- I cannot find any reliable source for the USB 3.0 drivers, so I will keep them uninstalled for now as I only have one USB 3.0 device which works fine in USB 2.0 mode
- Windows Memory Diagnostic Tool returned no issues, see attached results

EDIT: Okay, now for the possible cause:
I ran Memtest86+ for 5 passes at first, and it reported 40 errors during test 6 around the address 00074B08810 (1867.0Mb)

I had no idea if that could be a random error that can be disregarded as a false positive, so I ran it again, for 10 passes.

This time it reported 3421 errors, and the last ones that were on screen were also at 1867.0Mb.

With this in mind, I have the following questions:

- Can this be considered as final proof that the error is coming from the RAM?
- As it seems it only affects on of the two RAM slots, I'd like to remove it to check if the problem disappears, can I assume it is the one in the lowest numbered slot?
Or can this 1867,0Mb thingy be on any one of the two?

Thanks in advance!

CaptainDju · May 24, 2012

One new thing that I started thinking about: Is it possible that the RAM can be fine but the mobo is bad? Or is MemTest a surefire way of pinpointing RAM issues?

zigzag3143 · May 24, 2012

CaptainDju said:
One new thing that I started thinking about: Is it possible that the RAM can be fine but the mobo is bad? Or is MemTest a surefire way of pinpointing RAM issues?

Memtest only tests ram BUT by taking a known good stick and running it in each mobo slot you can tell if an individual slot is bad,

CaptainDju · May 24, 2012

Okay, so as the MemTest errors all seemed to point to the same memory area that was below the first 4Gb, I tried removing the first DIMM and played for a while.

After 1.5 hours with not a single glitch, I'm starting to think I got my culprit.

Will run a MemTest with the remaining DIMM overnight.

zigzag3143 · May 24, 2012

CaptainDju said:
Okay, so as the MemTest errors all seemed to point to the same memory area that was below the first 4Gb, I tried removing the first DIMM and played for a while.

After 1.5 hours with not a single glitch, I'm starting to think I got my culprit.

Will run a MemTest with the remaining DIMM overnight.

good job and good luck

CaptainDju · May 25, 2012

Well, the remaining DIMM showed 0 errors after 12 passes, so I guess I removed the correct one

What I'll do now is gradually re-install whatever I uninstalled as noted in my above posts and test the heck out of this computer this week-end (read: finally play BF3 after weeks

)

I guess I'm going to have to ask a refund on that RAM and buy another set of DIMMs, hopefully I'll have it fast.

In the end, if all goes well and I have 0 crashes, I'll come back and mark this thread as solved.

In any case, your support has been incredibly fast, friendly and focused, so thank you very much!

EDIT: Well, the store at which I bought the RAM told me to do one more test before returning it: As I have an i5 2500K processor they would like me to test by changing the BIOS as follow:

RAM frequency to 1333MHz
Timings to 9-9-9-24
Voltage to 1.25V
Command rate to 2N or 2T

What do you guys think of that approach?

writhziden · May 25, 2012

It is not a bad approach, and I have seen it resolve problems maybe 5-10% of the time when errors show in Memtest86+. Those are not great odds, but if the company suggests you do the changes, it should be done at least to test.

CaptainDju · May 29, 2012

Okay, I finally found some time to re-test both RAM sticks as requested by the vendor.

I attached the two BIOS screens to display the RAM settings I used, as well as the MemTest results: 0 errors after 14 passes.

I'll keep an eye on this as I haven't had a chance to give it a "real" test, but after one long and crash-free gaming session I think i'll be able to mark this one as solved anyway

writhziden · May 29, 2012

Great news! Glad to see the RAM settings provide more stability. Look forward to hearing the final results.

CaptainDju · May 31, 2012

Well, bad news (for me), after using my computer for a while with a few RAM-intensive games the issues re-appeared almost instantly.

Once again the same symptoms, either the game, Windows Gadgets or Firefox would crash.

Let's remove the second DIMM and all goes well.

Here's what's on my ToDo list:
- Wonder why MemTest went fine for 14 passes but computer crashed right after normal usage was resumed
- Currently the fan on my cooling system prevents me from using the A1 and B1 slots so I'd like to move the fan on the other side of the cooling and test the same 2 DIMMs in different slots
- I'm getting a new kit of exactly the same DIMMS so I'd like to test those in place of the current ones

For the first item, I have no clue, I can only note the results.

The second item is a test to check the memory slots, which would then point to a MB failure instead of a RAM failure, but again, see item one for my question

The third should give a final answer to the "Is it the RAM or the MB" question. Indeed if the system still crashes with new DIMMs then the mobo is the culprit, otherwise the first RAM kit is back to the store.

Is the process above worth anything or am I overlooking something?

writhziden · May 31, 2012

Memtest86+ is programmed to test the RAM in certain ways. There are two reasons it may not find errors when errors do exist:

It was not run long enough. It may take up to 21 passes to find an error in RAM.
It does not stress the RAM as thoroughly or in the same ways as Windows itself stresses it. This means it may not stress the electronics within the RAM that are faulty and may miss those areas of the RAM having errors. Sometimes, running Memtest86+ on a cold boot, multiple times, or for more passes will cause it to catch areas it missed the first time.

Your troubleshooting steps you have outlined are sound. Proceed as described.

CaptainDju · Jun 6, 2012

Okay, finally got the new identical RAM kit delivered, and had time to run some tests, I think the conclusion is indeed a faulty RAM kit:

First I ran a positive control test where I reset the BIOS settings to Auto for all items I had changed in a previous post, and re-installed the faulty RAM in the A2-B2 slots
--> Crashes as frequent as before, as expected

Then I placed the faulty RAM in the A1-B1 to see if it could be a faulty slot (mobo) issue instead
--> Crashes as frequent as before, consistent with the faulty RAM theory, but does not rule out the faulty mobo

Third step was to remove the faulty RAM and install the new RAM kit in slots A1-B1
--> Not a single crash in a total of 10 hours of gaming, where I used to have crashes every few minutes. Still consistent with the faulty RAM theory, but does not rule out the faulty mobo

Last step was to move the new RAM to the A2-B2 slots to test them:
--> Not a single crash in over 18 hours of gaming/intensive use.

So in conclusion I think it is safe to assume the mobo is fine and the old RAM kit is faulty, so I'll send it out for replacement.

Here are my lessons learned on this one:

- Even though everyone hopes for a quick fix, ALWAYS perform test steps one by one (e.g. remove ONE supposedly faulty component and test right away). If you change multiple things and the issue is resolved, you won't know for sure what actually solved the issue.

- MemTest results are definitely not 100% accurate unless you leave it to run long enough. I think it's pretty safe to say that running it for less than 10 passes is worthless.
Note that even with enough successful passes (20+ have been advised) it does not replace real-life situations a.k.a. whatever you usually do on your computer that makes it crash.

- If the component you're suspecting is plugged in the motherboard, don't rule out the possibility that the motherboard itself is faulty, ideally have a negative control test with another similar component that is known to work fine

- RAM issues are a pain to diagnose as you can easily have false negatives as I have had. Also there's nothing you can do with your computer without RAM

I'll mark this one as resolved, hopefully it can help other people.

Anyway, thanks a ton for your super fast follow up and very knowledgeable answers.

TL;DR: The RAM was borked in the end, tadaaa

writhziden · Jun 6, 2012

Thank you for posting back with an excellent set of steps and tips for users. I have added it to my links to refer other users to when Memtest86+ tests show no errors, but the RAM may still be faulty. I appreciate you posting that!!!

Glad to see this resolved. :party:

Solved BSOD and crashes playing BF3, happens with 2 different GPUs

New member

My Computer

Closed by request.

My Computer

Account closed by request

My Computer

New member

My Computer

Closed by request.

My Computer

New member

My Computer

New member

My Computer

Account closed by request

My Computer

New member

My Computer

Account closed by request

My Computer

New member

My Computer

Closed by request.

My Computer

New member

My Computer

Closed by request.

My Computer

New member

My Computer

Closed by request.

My Computer

New member

My Computer

Closed by request.

My Computer