PC Freezes When Gaming

[Grabino]

Whenever I play anything that requires high graphics (Witcher 3, WoW on Mid level), my PC has one of the following reactions:

• 
  Screens freeze and become unresponsive, with random artifacts on the screen. Then the screen goes black and all my external ports become unresponsive.
  My PC will have a BSOD with this error: "Stop code: Video_Scheduler_Internal_Error"

Mind you, once I explain my components, you will realize that this is not really high for my GPU, but its been occurring for some reason. My setup is:

ASRock Z270 Killer SLI/AC
Intel Core i7-7700
EVGA GeForce GTX 1080 Ti FTW3 ELITE
EVGA SuperNOVA 750 G3, 80 Plus Gold 750W

I ran the Furry Donut and got the same issue as well. Also, recently I got a new chassis for my PC and moved components, so it is possible that I may have caused this issue, but the question is where is the issue coming from so I can resolve it. I was thinking it was the GPU at first, but I wonder if its the Power Supply or the MB. Whenever I run games, I don't have an issue at max graphics, but then it just freezes. Any thoughts community?

 

[Grabino]

Another note is that when I turn the graphics down in those games, I never freeze. But when it isn't turned down, I have constant BSOD/freezes and then sometimes my PC will not even complete the boot process and my externals will not even start.

 

[panathas]

It could be a hardware issue (GPU, PSU) or a software (drivers). Also make sure that the GPU isn't overheating. You have to do a clean re-installation of the GPU drivers. If that doesn't work you have to test your system with another PSU and/or test your GPU in another system. That way you'll find what's causing this. Keep in mind that there is chance that something else may be responsible, such as another system driver or the motherboard (PCIe slot) but you should start from the most likely suspects. Good luck.

 

[panathas]

Another note is that when I turn the graphics down in those games, I never freeze. But when it isn't turned down, I have constant BSOD/freezes and then sometimes my PC will not even complete the boot process and my externals will not even start.

This points to a failing PSU, You should start from that and test your system with another PSU. Try to borrow one from a friend if you don't have a spare PSU available.

 

[Grabino]

I have a backup one. I will give it a try! Thanks! I'm out of town atm but I should be back in a couple of days to test it.

 

[Grabino]

So I ran some more test on this. I still haven't tried a different PSU (I've been gone for a while so I just had a chance to check on my GPU), but I noticed that in games where I downgrade the graphics, no issue.

Also, when I ran another semi-high graphic game, I received this error: 887a0006 "The GPU will not respond to more commands, most likely because of an invalid command passed by the calling application."

The game had the same graphic error occur, but this game actually presented an error code with it. Thoughts? Also, I just updated my drivers to the most recent build.

 

[panathas]

Are you sure the GPU isn't overheating? All those issues may be heat related. As I previously said it could be the PSU, GPU or a strange software issue. You have to check the PSU and GPU and if the issues aren't gone then check the software side because eventually you may need to reinstall windows.

First test your system with another PSU. It's a necessary step and it'll give you a lot of information. Testing the GPU in another system is the next step, but it's more difficult and not as conclusive as the PSU test. If you manage to complete both tests you'll probably find out who's responsible.

 

[Grabino]

Okay. I will check it out and see.

 

[Sithtiger]

I too believe you are clearly having an overheating issue! I'm assuming everything was working before you moved to a new chassis, but it sounds like your GPU, CPU or both are overheating. I'd make sure you have the newest drivers, specifically your GPU. I'd also check your fans, make sure their spinning. Check your temps. Use HWMonitor, something like that that checks your CPU and GPU temps. Run a benchmark or stress test. I'd start with the GPU. Make sure the fans are spinning. Check the temps and make sure they aren't shooting up. You said when you lower the game settings games run fine, which seems to suggest a software issue or a heating issue. A 1080, even though it's coming up on 2 years old should handle ANY game out there at 1080P....even if you were running at 1440p! Also, make sure everything is seated properly.

It's probably a driver issue or perhaps your HSF or liquid cooler (if you're using that) might have come unseated and not making good contact with your CPU. You might need to remove the thermal pad or thermal paste and reapply. It could also be your GPU fans have died. I know Pascal cards...at least the 1080 and 1070 fans don't spin up until it hits 60C I think it is, or if you manually set the fan speed using Afterburner or some other overclocking software.

I don't believe it's a PSU issue because if it was, it should happen when spontaneously or when you start any game at any setting, but that's not to say it couldn't be, it would just be far down on the list from what you're describing. Oh, one more thing. Make sure you have both PCI-E cables connected to the 1080. I believe they take two 8-pin connectors, or at least one 8 pin and one 6 pin. One last thing. If you're overclocking anything at all, make a note of your settings and then restore everything to a default setting. I think you'll find, this is most likely heat related. That would be #1 on my list! Please report back. I'd like to know the outcome.

 

[Grabino]

So I just installed a new PSU. So far I haven't been able to recreate the issue that is occurring. However, I realize I have been doing this with my chassis open. My last test is to do the setup in its original location and figure out if overheating is the true cause.

 

[panathas]

Temp plays a huge role especially when you are using high end components. You should always be running temp monitoring software in the background. Also GPUs are more sensitive to overheating issues than CPUs. CPUs will just throttle back but GPU overheating will also cause display artirfacts. That could also explain why you couldn't complete the boot process afterwards. Your system (CPU, GPU or both) had to cool down. But then again this is a symptom that a faulty PSU can certainly cause. That PSU could also be overheating or having a failing fan.

Make sure that every fan inside your system is spinning (including the PSU fan). That said a lot of newer system components have fans that spin only under load or under various predefined temperature thresholds, so you have to make them spin via software (GPU, Case fans), or hardware switches (PSU).

Please monitor your PC's various temps, it's very important and it's always the first thing you check in case of unexpected system behaviour. Good luck.

EDIT: Since you changed the PSU, take everything that I wrote about PSUs, as general advice.

 

[Grabino]

Nevermind...I was gaming today and the same thing happened again.

I looked at all the fans, I am still using the new PSU unit, I'm looking at my temps and my CPU is at a steady 40C and my GPU is at 37C.

If it isn't the PSU or Temps, would it be safe to assume the MB or GPU?

 

[Raiddinn]

Did you by chance happen to change RAM recently? Did you change CPU? I find freezing to most commonly be related to CPU or RAM, but not always.

Everything points to a video card that is failing here to me. Helps that all the errors seem to have "GPU" in the error message. That isn't definitive by any means, but it's a strong signal in any event.

If you can use the computer fine with the side off and it fails again when the side goes back on, that tells me that something somewhere is getting too high in temperature. Maybe that's happening to the video card. That's the best suspect I have atm.

You might try getting a different case with better airflow if your current one sucks, or maybe just add more fans if you have open slots. That's cheap. You can do it for like $10 and it can make a big difference.

Also, I just want to throw it out there that you have to take temps from various apps with a grain of salt. Not every MB displays temps the same way and it will confuse some temp measuring programs. They are most often right, but they absolutely can be wrong too.

 

[Grabino]

So I took the side off and the same failure occurs while gaming on higher demand games.

My final test, I think, would be to try another GPU at this point. The PC runs without issue without the GPU slotted. I'm thinking, at this stage, it could be the PCIE slot or the GPU itself. On that same note, I tried both slots without issue.

 

[Raiddinn]

I would agree that trying with a different (ideally the same model) video card would be a sensible next troubleshooting step.

 

[Karadjgne]

You are missing something. There's 2 areas to any graphics card. The gpu (processor) which has a temp sensor, and the vram/VRM's which does not have a temp sensor. Just because the gpu is reading low temps doesn't mean that the vram/VRM's aren't seeing high temps. Check the gpu fans, all should be spinning at about the same rate. Also, when was the last good cleaning?

 

[Grabino]

I'll have to look into the fans and see if they are spinning. I cleaned it the when I made got my new chassis approximately a month ago and again when I attempted another PSU. How would I check the VRAM temps?

 

[Karadjgne]

Only way is with a IR camera or other laser temp reader

 

[panathas]

A lot of EVGA cards have VRM overheating issues. One thing you could do is to manually set the fan curves much higher but you'll be forced to leave with increased noise levels. If higher airflow is able to fix those issues then your card is probably defective and you should RMA it. Good luck.

 

[Grabino]

So I just went a purchased a PNY 1060 GPU. No problems. Sigh, I have to RMA this thing...Thanks for the help everyone.

Also, once I get it back, I'll see if I can ask Customer Support what happened with the GPU.