ga-z68x-ud3h-b3 rev 1.0 problem with PCIe x16 slot for my videocards

[flippiej]

MOBO: Gigabyte GA-Z68X-UD3H-B3 rev. 1.0 (updated from f10 to f12 during the story below)
CPU: Intel core i5-2500k @ 3.30GHz (no OC)
RAM: Corsair Vengeance 8GB(4x2GB) DDR3 1600MHz
GPU1: MSI GeForce GTX 560 Ti, 1GB, Twin FrozrII (no OC)
GPU2: MSI GeForce GTX 560 Ti, 1GB, Hawk (no OC)
PSU: CoolerMaster GX 750W
SSD: OCZ vertex 3 Max IOPS 120GB
OS: Windows 8.1 64 bit

Well I'll tell first what the issues were that I encountered and how I got to the conclusion of it being a problem with the PCIe x16 slot on my motherboard. My main problem was freezing of my pc without any reason and then also having difficulty to restart the computer.
The long story is mainly there to give as much information as possible. In advance I apologize for my writing style and I hope I made the story understandable

A short summary from the story below to keep things clear:

- pc freezing after a couple of hours of being turned on -> programs won't open anymore and shutting down takes a long time.
- reinstalled windows 8
- problem occurred on the same day
- wiped my SSD clean + updated firmware of SSD
- reinstalled windows 8
- problem reoccurred after about 5 weeks of having no problems
- bought a new SSD and copied windows + files from old drive to it
- after using it for a few hours -> same issue
- started testing videocards in their slots (PCIe x16 and x8 )
- both videocards failed in the PCIe x16 slot apart from each other with booting windows
- reapplied motherboard drivers and updated BIOS from f10 to f12
- pc couldn't boot past windows logo anymore at all -> BSOS loop
- repaired windows 8 through windows 8 install drive
- pc is stable for a couple of days with only 1 videocard attached in the PCIe x8 slot


The last few months I'm encountering a problem with my pc from time to time in which my pc seems to freeze while working on it.
This occurs after a couple of hours of having the pc on. The pc seems to be working normally but suddenly I cannot open new programs or tabs in my browser, but I can still talk through Hangouts for a short while (and get messages back). Once this happens everything on the pc that might open will do so very slowly(takes minutes instead of half a second), while sometimes my RAM would rocket to 100% and other times not. Taskbar cannot be opened when this occurs as well (I keep getting an error when opening it during this freeze) so I wasn't able to check if any process in particular was the issue.
The only option to get my pc working again is by restarting it. Shutting down the pc takes about 5-10 minutes before anything happens and the real shutdown is after 5-10 more minutes. After this most of the time the pc also has problems restarting and will freeze at the moment that the loginscreen of windows would show. My mouse will load into the screen, but the screen stays black.... sometimes loading after a few minutes(would be as slow as before the restart) and other times not even after 1 hour. Restarting the pc for a second time would most of the time do the trick to get the pc running again smootly.

I've tried reinstalling windows 8 to get rid of this problem and also to see if any of my software was to be blamed, but it reoccurred the same day already.
I then thought that my SSD might be the one to blame and wiped it clean with the OCZ toolbox and also updated its firmware and then reinstalled windows 8.1.

My freezing problem didn't occur for about 5 weeks after this, so I thought I was right about my SSD being the problem, but when it occurred again and also the next day for 2 times, I decided to buy a new SSD. (samsung 840 evo 250gb)
This time I was lazy and let Samsung's Magician software copy Windows and all it's files from the old SSD to the new one and let it boot from there. On the next day however my pc froze again with the exact same problem. So it wasn't due to my old SSD at all.

On advise of a friend I then started to take components out of my pc and test if they were causing the problem. I took out both of my videocards (PCIe x8 and x16 slots) and my computer started up normally. Retried it multiple times with a single videocard put in and my computer actually wouldn't start anymore with only 1 videocard in the PCIe x16 slot. In the PCIe x8 slot both of my videocards are working normally on their own but neither is working in the PCIe x16 slot. I've also tried changing the power cables for the videocard from the PCIe x8 slot for the x16, but changing this power cable didn't change anything. Combining the 2 videocards together (with and without the SLI cable attached) in the PCIe x8 and 16 slot made the computer start up normally most of the time.

I've been using my 2 videocards together in SLI for about 1.5 year without any problems, so I don't know what caused the problem with the PCIe x16 slot all of a sudden. Speaking for the nvidia drivers, while having the freezing problem I've been using multiple versions of it, with at latest now the beta 337.50.

Anyway I thought maybe a BIOS update and/or motherboard drivers might fix this issue for me so I installed the motherboard drivers for windows 8.1 again (with the Intel Management Engine Interface driver for the chipset not able to install, because the ''minimal requirements were not met''). I then made the mistake to update my bios from f10 to f12 through @bios.
The update itself seemed to go fine, but afterwards I couldn't start my pc anymore My pc would go past the bios screens and I could see the windows 8 logo. It then took a little while before the wheel underneath came, but after a few rotations it crashed and the screen would show the BSOS (blue smiley of sadness ) screen and restart instantly after that resulting in a loop. This was also the case when trying to boot from my old SSD.

I then let the windows 8 dvd repair windows and it did start up normally (1 videocard in the PCIe x8 slot). I've now been running windows 8.1 on this computer normally and so far my pc didn't do the strange freeze up or anything.


However... now finally coming to my question after this story. Do you guys know anything that I might be able to do to get the PCIe x16 slot to work normally again? I'm quite sure that it worked by itself for 1 videocard at some point, because before I got my 2nd gpu I had my pc running with just one in the PCIe x16 slot for a year.

I still have about 8 months of warranty left on the motherboard, but I would prefer to keep my pc running and fix it myself instead of being without one for a week or 2 So does anyone have any idea what I might do apart from that?

Thanks for in advanced and also for reading through this wall of text.

[dmdilks]

Boy that was a mouth full lol. One thing you didn't put your story was did you ever try different memory.

[flippiej]

No I didn't, but I also don't have any other RAM to test with.
While trying out my videocard through the PCIe x16/8 ports I did switch their places, but I didn't try it with only 4GB in total attached to the board.

Do you think it might be useful to test my videocards in the PCIe x16 slot again but then with only 4GB attached as RAM to see if that makes any difference? I could try some combinations with that.

[dmdilks]

Yes that would be something to try. I had a gigabyte board one time run for months with Kingston 2gb 1066.

Then one day the computer wouldn't boot. I change it to the same memory but 1gb sticks 1066 it boot and it is still running to this day.

[flippiej]

Well just to come back on this I've found solutions (I think) for both my problems.

The PCIe x16 slot probably didn't work because the powerconnection to the motherboard was a bit loose for the last few pins xD
I connected them back properly and made sure they cannot get out of it again and now the PCIe x16 slot works normally.

I have no idea why this power issue would then only affect this slot in particular (and why the motherboard would actually start at all if it didn't have enough power) but that seems to work. Fixed that 3 weeks ago or something and is still working.


Now for the freezing issue it still occurred to me yesterday so I'm not 100% sure I fixed it now, but will report so if it's not the case.

It seemed that some hidden process was taking up a lot of RAM so I've installed Poolmon and RamMap to look at hidden processes and if anything was out of the ordinary.
During the freezes my Nonpaged Pool is very large according to RamMap (3-4GB+) which would mean that something is using my ram but not giving the used space back. With poolmon I found the tag Tag3 to be the one with a very large memory amount in Nonpooled during freezes as well as Vepp. (vepp only comes together with Tag3, but seems to be Verifier.exe which was probably recording the heavy ram usage)

Through a search Tag3 seems to be associated with 2 of my programs. iaStorV.sys which is a process for Intel Rapid Storage and cmdguard.sys which is a process for Comodo Firewall.

Disabling Intel Rapid Storage didn't change anything, since the freeze came back. I saw through verifier.exe that the cmdguard.sys process indeed used almost the same amount of Nonpaged Pool RAM as tag3 did, so this confirmed for me that my firewall is the problem.

I've deinstalled it yesterday and I'm just sticking with windows firewall + my routers firewall as that probably should be enough . I'm hopeful that my pc won't freeze anymore, but since my pc is only a few hours on since deleting the firewall I can't be sure yet.

Anyway tnx for your help and hopefully my pc will run smoothly from now on again

[dmdilks]

When I'm having a problem like that I use Task Manager. You watch the CPU usage.

Than you go to the Processes and see what program is sucking up the CPU and Memory. Plus make sure you do with all users.

Unless you are on a server you don't really need all the extra programs to protect yourself.

I use Microsoft Security Essentials,  Malware-bytes and windows fire wall. That is all you really need. Plus Good Luck.

[flippiej]

If it only was that simple
Task manager never showed any real abuser when I could get it to open. With my 8 GB of RAM at 100% during a freeze, I combined all the memory used according to task manager and the number would only get to 50% of it max and not the 100% it displayed. Apart from that the memory use sometimes did not go over 50% (or would climb up steadily to 100% then slowly drop down after 30 minutes of doing nothing with the pc still being frozen and not able to use any of the ''free'' memory.)

Processor wasn't active during those freezes at all. At most 4% or something. When I now think back on it comodo firewall was never really using any resources according to task manager so I'm quite happy that I found these other programs in the end to pinpoint my problem. Task manager was useless :p

And I'm the only user account on this pc, so I know no one else is breaking things apart from me
Pc still running smoothly so far with no random nonpaged pool climber and I found another person on the Comodo forums complaining about tag3 and it's nonpaged memory leak, so I guess I'm not the only one.

But again thanks for your help

[MagnusB]

I have the exact same issue with a z97x-ud3h. Driving me crazy.

Video just starts behaving weirdly, freezing apps and I can't shut the PC down.
Have to reboot via the power button.

It's not my video card, I have the same issue with just the internal Intel video GPU.

I have now uninstalled all the Gigabyte apps to see if that helps. The CFOS Speed Internet Accelerator was such a pile of garbage anyway ...

[flippiej]

Well for me it was Comodo Firewall as you can read above. Pc still running without any freeze so far, so I'm positive that Comodo was causing my issue at least.

If you don't have that running, I think your freezes are unrelated though I would suggest installing RamMap and have it running the whole time to check if your Nonpaged Pool is growing extremely large during a freeze.

[MagnusB]

It's running better now after uninstalling all the Gigabyte stuff but has frozen once in two days.
I suspect the remaining issue is due to some power save feature.