Graphical glitches with a stable graphics card, where does the issue come from?

Hi, my current system is equipped with a Gigabyte GV-N670OC-2GD (with two monitors) and an Intel i5-3750K, running Windows 10 Pro and Linux Mint 18.2 Cinnamon in dual boot.

I've been sighting graphical glitches for the past few days, sometimes one of my screens flashes black for a few frames, or the bottom half of one them flashes a random color. And it keeps getting worse, I now get several black screen flashes in a row. I also had the Cinnamon desktop environment crash a few times under Linux Mint.

I first thought that my graphics card was dying, so I swapped it with a Gigabyte GV-N770OC-2GD, but the glitches still occurred. I also ran Uningine Heaven under Linux Mint for 9 hours on my Gigabyte GV-N670OC-2GD, the system didn't crash, and the graphics card didn't overheat (75° maximum).

As I mostly use Linux Mint 18.2 and recently upgraded from 17.3 to 18.2, I started thinking that the Linux Nvidia drivers could be the issue. So I tried a few things, I first tried upgrading the Linux kernel to the latest available version, I tried rolling back to Linux Mint 17.3, and then I tried using only the recommended driver under Linux Mint 18.2 (nvidia-375), but none of those did anything to fix the graphical glitches.

As I don't use Windows 10 Pro that much, I didn't had the chance to sigh any major glitches when running it. I only got the screen flash black once, but I was installing software when it occurred, so it might be unrelated.

Here is the complete list of my computer's components:

• CPU: Intel i5-3750K (No overclocking applied);
• Graphics card: Gigabyte GV-N670OC-2GD (No overclocking applied);
• RAM: 2x 4Gb DIMMs DDR3 @ 1600MHz;
• Motherboard: Asus P8Z77-V PRO/THUNDERBOLT;
• Storage: 1x Intel SSDSC2CT120A3 120Gb SSD, 1x Western Digital WD10EZEX-00RKKA0 1Tb 7200RPM HD;
• Power supply: Seasonic G-750 750W Semi-Modular.

Notes:

• Neither the graphics card, CPU or RAM are or were overclocked;
• The graphics card requires two separate PCI-E power connectors (1x 6 pin + 1x 6+2 pin), each connector uses a separate power cable;
• The system was recently moved (carefully) by car, but I didn't notice any damage upon arrival and the system booted fine.

So my graphics card isn't to blame here, neither are the drivers. What else could I try to pinpoint the issue?

Thank you in advance and have a nice day.

is always the same screen you're seeing it on? If so, you might try moving that screen to another system and see if the problem goes away.

ryan_975 said:

is always the same screen you're seeing it on? If so, you might try moving that screen to another system and see if the problem goes away.

Click to expand...

The glitches happens on both screens, moreover, I had Cinnamon crash under Linux Mint.

Could something be overheating? Maybe try testing the memory?

RazorWind said:

Could something be overheating? Maybe try testing the memory?

Click to expand...

I'm constantly monitoring the temperature of the CPU and GPU under Linux Mint, my i5-3750K runs between 30° and 40°, my GTX 670's GPU runs at approximately the same temperature, sometimes it gets around 50° or a few degrees higher, which isn't even close to overheating.

I didn't test my RAM, I'll run tests overnight, but why would bad RAM cause these kind of issue?

As my power supply is semi-modular, I checked every single cable, I unplugged and plugged back each connector.

I also checked my motherboard and I didn't notice anything (all capacitors are solid and none seemed to be leaking or in bad condition). Could it be the motherboard? Could a poor electrical contact or a bad capacitor cause this?

Thanks for your answer.

P.S: I didn't use my computer a lot today but I didn't notice any glitches, maybe a cable's connector got loose? I don't know, but I still don't consider the issue fixed as I didn't have much time to observe my computer's behavior today, so maybe the glitches will come back soon, I'll keep the thread updated on that.

Flaky memory can cause all kinds of strange problems. Everything the system does goes through the memory, so if 1 + 1 suddenly equals 4, strange things can happen. A flaky power supply or monitor cable is a possibility as well.

You might also try running from either a liveCD type USB installation or running windows for a couple of days, and see if your problem goes away. This could be entirely software related, and just an artifact of the combination of your linux distro and graphics drivers.

Could be as simple as a flaky monitor cable/connector. Double check all those I'd say.

My first thought was graphics ram, but you say you swapped the card out and get exactly the same issue? And is it a new card or another old one?

spine said:

Could be as simple as a flaky monitor cable/connector. Double check all those I'd say.

My first thought was graphics ram, but you say you swapped the card out and get exactly the same issue? And is it a new card or another old one?

Click to expand...

Both graphics cards I have are used (GTX 670 and GTX 770). They both seemed to be stable when I stressed them with Uningine Heaven (no overheating, no graphical artifacts and no crashes). And yes, I got the exact same issue with both, black flashes and the bottom half of one of the screens flashes a random color.

As for the monitor cables, I don't think the issue comes from there, as I had the Cinnamon desktop environment crash multiple times under Linux Mint, and I don't think a bad VGA/DVI cable could cause this in any way. Or maybe these crashes were unrelated? Moreover, I have two monitors, one uses DVI, and the second uses VGA, it is extremely unlikely that both the cables got bad at the same time.

Plug the monitors into the iGPU and remove the discrete GPU.

If the conditions continue then you can rule out GPUs and that particular PCI-e slot. If the motherboard has more than one 16x slot try moving the discrete card to it.

+1 on checking the RAM. Run tests, if it passes then try running just one stick at a time. See if the problem exists or goes away.

I've had odd graphical issues with Mint before on different machines. After switching to a different Linux distro, never had any graphical issues. Have you tried a different distro to see if the problem is still there?

This sounds like a cable or monitor issue to me, not a GPU or PC issue. Check the cables aren't pinched and are properly plugged in at the monitors.

Yep, 'looks' like a dodgy cable or connection.
Image blanking is common for this type of problem.
Strange colour sections for a moment an extension of this.

It happening on more than one screen could point to a bad earth problem that is disturbing the monitors operation by causing strange current flows through the earth connection of the HDMI cable.
If you see a spark when connecting an HDMI cable I would suspect this. If not its still worth trying the following.
Try different mains cables.
Failing that try different mains sockets in the house, direct to the sockets. No extensions.
See if that changes anything.

Today I've had weird Firefox graphical glitches and a crash under Windows 10 Pro, meaning the issue does NOT come from Linux Mint. I'm also pretty sure that under NO CIRCUMSTANCES a bad cable could cause issues at the operating system level (and I do not use any HDMI cable, I use VGA and DVI).

The glitches consisted of a black flash with a withe bar on the top of the screen when I was playing Youtube videos in full screen.
The crash happened with a pretty similar glitch, when switching to an already opened tab, it displayed only black with the top window bar being solid white, the weird thing is that EVERY other tabs worked fine (the faulty tab had a Youtube video opened), Firefox crashed a minute later. The crash report CLEARLY stated a graphics issue, stating an error while rendering bitmap and mentioning "gfx" and "displayAdapter" everywhere in the crash log.

I've got new graphical glitches under Windows 10 Pro. Sometimes, 1 out of 2 frames when playing a video (Adobe Flash Player) is replaced with a solid random color (mostly green or navy blue), this goes away if I pause the video and then play it again.

I think neither the drivers nor cables are the culprits.

Then it's time to rule out the GPUs. Get a cheap ass GT1030 or something and see if that doesn't experience the same issues.

doubletake said:

Then it's time to rule out the GPUs. Get a cheap ass GT1030 or something and see if that doesn't experience the same issues.

Click to expand...

I have a lot of spare graphics cards, I'm not going to buy a new one, I can't afford it right now. Moreover, I already tried a GTX 770 that I own, but I still had the same issues.

I also ran Memtest86 from the Ultimate Boot CD to check if my RAM wasn't the culprit, but it didn't report any errors after several passes (Memtest86 ran for 10 hours).

I'm starting to think that my motherboard might be the culprit, I'll try to use another PCI-E x16 slot for my graphics card and see if it fixes anything.