nVidia Grahpics Cards Problems

[daytripper]

Hi guys. Let me start out by saying that I usually research the problem at hand, before asking for help of any kind. But I feel that my "problem" seems to be a little out of the scope of my knowledge, and I am hoping there is someone else out there that has had this same issue, and can help me resolve it.

I built myself a gaming computer earlier this year, but I also consider myself a "power-user". I like to be able to "fiddle" with things so to speak. I like to do stuff in Photoshop, and well as edit and render videos, etc.

Anyways, I have been having a very annoying problem since July or so.

The issue I seem to encounter is with my video card. It is a Galaxy GTX 680 (4gb) GC.

I was having issues when Windows was booting up, albeit the problem has been intermittent, and has been very hard to diagnose.

Usually I would just freeze at the Windows logo. But occasionally I would make it past there, but later come crawling to a halt at the desktop or ANY OTHER activity, such as web browsing, or searching my music in folders.

There have been stretches where I would be able to use the computer normally as well, but when I would reboot, the computer would display the same symptoms again.

Usually when this happens, I have to go through a process of rebooting the computer 20+ times just to make it to the desktop, in hopes of being able to actually use the computer.

Below I will list the parts in my build (as well as parts I have tested), and my methods behind the tests I have performed so far.

----------------------------------------------------------------------------------------------

CPU: Intel Core i7 3770k (4.5ghz, I have also tested it at stock speeds.) - I also have an i5 3570k at my disposal, but I have not tested it yet.

RAM: Patriot Viper IEM 1600mhz 4x8gb (32gb total) - I also used a 8gb set of Corsair Vengeance RAM in my testing.

MOBO: Gigabyte Z77X UD5H (I am currently running an MSI Z77 MPOWER, which was purchased as part of the "troubleshooting" phase.)

PSU: XFX Pro 1250w Black Edition (My original plan was to go multi-GPU on this PSU) - I also have a Corsair HX750 on tap as well.

GFX: Galaxy GTX 680 4gb (I also have a Gigabyte GTX 770 4gb which displayed most of the same issues.) I ALSO have an HIS Radeon HD 7850 which hasn't had ANY problems, whatsoever.

STORAGE: Samsung 840 Pro 256gb, WD Black 2tb, WD Green 2tb.

OS: Currently on Win 8.1 but I have also tested Win 8 and Win 7.

-----------------------------------------------------------------------------------------------

I have made sure every time I have installed a new driver, I have removed the old one, and swept over the previous with Driver Fusion Pro to remove any traces.

I have also performed a clean install of three different OS's multiple times.

Windows copies are legit, and have been fully updated through Windows Update.

Currently the Radeon displays none of the problems that it's nVidia counterparts are displaying.

-----------------------------------------------------------------------------------------------

I have RMA'd the Galaxy card, in which the symptoms disappeared for about a day or so, and then randomly resumed again. I have been trying to troubleshoot with them, but haven't made it very far.

They have suggested that I update the BIOS on the card, but I really don't think that I should have to do that to get the card working again.

Please note that I have recently tested my card in another rig, in which it worked (Same rig I have the i5, HX750, and GTX770 from).

The only other things I can think of is the PSU or CPU is faulty, or it is a software issue. - Which is what I am leaning towards. Particularly the nvlddmkm file.

Any input on this matter is greatly appreciated.

 

[EzioAs]

Just to pitch in a little bit but have you tried updating the motherboard's BIOS?

 

[daytripper]

Thanks for the reply, EzioAs.

I just realized that I have forgotten to mention a few things. I also have a Gigabyte Z77x-D3H and 8gb of Corsair Vengeance Pro memory on hand as well. AND I have an entire AMD based system (mid-entry level) that I have laying around.

I have also tested the newest driver versions from nVidia I believe R331, which was out right after BF4. All the way back to drivers from 2012, and the problem still persisted. I have also performed "selective installs" where I opt out of GeForce Experience and 3D surround, etc. This didn't work either.

I have been closely monitoring my temps and voltages, and everything seems within spec. My system as a whole idles around 20-30 degrees Celsius.

On the Gigabyte (UD5H) I have tried the BIOS the board shipped with, I believe it was BIOS version F14... although I can be mistaken. I have downloaded BIOS versions F15o and F15q which are listed as being in the Beta stage. To my dismay, it didn't resolve anything.

On the MSI Board I believe I am running on the current BIOS:

American Megatrends Inc. Version V17.6 (12/19/2012)

I have tried to download the 17.11 version off of MSI's website but I remember it was doing some weird stuff, cant remember what exactly, but IIRC it was locking up at the BIOS and rendering the system nearly unusable. Luckily I was able to boot into the backup BIOS and recover the original one.

Should I incrementally update the BIOS? For example 17.6>17.7> and each one following until I get to the newest BIOS?

I never thought that it really mattered, because I thought that once you "flash a BIOS" that there was nothing remaining on the actual chip, except for the BIOS code itself (all the code you would need would be on this new BIOS, so skipping a couple of BIOS updates shouldn't negatively impact you. So it wouldn't be like installing some kind of driver that you have to worry about the remnants hanging around, right?

I may be incorrect in assuming so, since I am not expert in BIOS's, but that was how I thought they worked.

______________________________

TL - DR

I have tried multiple BIOS versions of my original board (UD5H). I have tried to update the bios of my current board (Z77 MPOWER) but after some technical difficulties I was forced to flash to the original BIOS again.

 

[EzioAs]

Might be the PSU, the XFX unit might be a little bit faulty since the Corsair HX unit works fine. Have you tried replacing only the PSU while keeping the rest of specs the same?

 

[daytripper]

That was also a thought that had run through my mind. I was kind of hoping that wasn't the case though since I bought a pretty high end unit that was manufactured by Seasonic, which is probably one of the best companies on the market.

Also, another thing that had bugged me was that I can run BF4 for hours on my 7850 without an issue (and as we all know, when you are gaming on a card, it is going to draw more power) - but, while I am using the desktop (I would assume that the card would be in a near idle mode, and would have substantially lower power consumption at idle than a Radeon would while gaming.) My 680 or the 770 will freeze, stutter, lag, etc.

But I can hang around in the BIOS all day long and everything is going to be fine.

Also, I would like to make a note as well that when I go to boot up into Windows, any time it freezes, it always does so on "nvlddmkm.sys". I also get the bit where the screen flashes, goes black, and it tells me that the graphics driver has recovered.

And if this helps at all, but when Windows does freeze at the starting flag logo, sometimes my monitor will immediately go to sleep. Like a fussy kid that can't even be bothered to try...

I originally went with a 1250w because I don't believe in "skimping by" and my original idea was to put 2 GTX 680's in SLI. I didn't want to push a PSU within an inch of it's life, potentially causing problems down the road. However, after buying one and seeing the performance, I was content for the time being and was going to hold off on getting an new card until the new wave came around. (Keep in mind, this was before the Titan, 7 series, and the new Radeon R series.) - That is why I went overkill on a PSU.

I am going to tear down the i5 build I have laying around, and hook the Corsair PSU up to my build, and see what happens. I am glad I finally got my hands on another quality PSU to test this with, since I didn't have any spares laying around at the time I did most of this testing.

Sorry for rambling, but I am trying to provide as much detail as I can while I am typing this, so that way people know exactly what we are working with.

I will try what you suggested, but in the mean time I am open to any/all suggestions.

Here is the part breakdown from PCPartPicker, for easier reference:

http://pcpartpicker.com/p/1WLGU

 

[daytripper]

Ok. Just tried running it with the HX750, and the same exact symptoms are still there.

Any other suggestions for me?

 

[lancerzero9]

This smells of a dying HDD.

Have you tried installing windows and therefore the Nvidia drivers on a different drive?

If you're installing windows on an SSD then forget it as that is not your problem. If your windows install is on a HDD it could be "going bad." This would explain why it can't load/read the drivers everytime.

The only other explination is that you somehow got 2 bad nvidia cards back to back. Its rare, but it can happen. Do you know anyone that can loan you a nvidia card to see if this is the case? Or loan one of your cards to someone for a week and see if they have the same symptoms.

 

[daytripper]

Thanks for your suggestion Lancer. If it was a dying HDD wouldn't the AMD drivers have trouble loading as well?

Anyways I have Windows 8.1 installed on a Samsung 840 Pro (256gb). The drive is fairly new, as I purchased it over the summer. From what I read on the reviews, it seems to be a very reliable drive. Also Samsung Magician states that the drive is still in good/excellent health.

My first card I had sent back and RMA'd it. I got another one back. (Galaxy GTX680 4gb GC)

My second card which is a HIS Radeon 7850 1gb seems to be working flawlessly throughout the intermittent problems, even when there are remnants of the nVidia cards on the system.

I have borrowed a Gigabyte GTX 770 4gb OC from my friend for trouble shooting purposes, and the problems have still been displayed, despite the fact.

I have also tested my 680 on the HX750 power supply, and after the drivers and everything have been cleaned, the card is still displaying the same issues on the different PSU. (Also to eliminate the possibility of interference, I have only hooked up the CPU and MOBO headers, the PCI headers for the card, and one SATA power for the SSD.

No fans, LEDS, AUX drives, or anything else were powered during this time.

Is there any known issues that cause nVidia drivers to do this, while AMD drivers remain unaffected?

My system, in terms of software, is pretty clean, since I just did a fresh install of Windows 8.1 around 2 weeks ago (whenever the release was). So I am not sure if there is something on my system causing the issues.