Question Application freezes / crashes (games & discord) + weird artifacting?

[BananaMonkey]

Hey all, seemingly randomly my computer has started to crash when attempting to run games or discord (The application freezes (no error) then I usually close it myself). I first experienced the crashes just last week while playing a Minecraft Modpack, my game would freeze after a few minutes, however I managed to avoid further crashes after dedicating more RAM to it. Since then I have attempted to play BattleBit Remastered (if you haven't seen it before, Google it and you will see it should be effortless to run) but after a few minutes in that game it froze, sometimes it's in the menu before I can even load into the game. However, while troubleshooting this issue (I was under the assumption that it was just BattleBit with this problem) my Discord also crashed. I then realised there was a much bigger problem at hand. I have tried all the things listed in the dot points below to try and resolve this issue, but nothing has worked. Also, after resetting my PC I wanted to see if an even easier-to-run game "SpiderHeck" would work fine, but even that crashes. Sometimes it's just the game that crashes, but other times Discord crashes right after the game too.

Since I typed the above paragraph I tested an older graphics driver, which resulted in the artifacting shown in the attached images (Artifcating_1.png, Artifcating_2.png), hoping that someone will be able to recognise the issue based on that (btw the artifacted pixels flicker). A restart got rid of them for some minutes before they came back, as did upgrading my graphics driver to a newer version. Still have them now, seems like things are just getting worse!

Specs:
CPU: AMD Ryzen 7 3700X 8-Core Processor
GPU: RTX 2080 MSI Gaming X Trio
Motherboard: B45M2 AMD B350M M-ATX
RAM: 16GB (2x 8GB) T-Force DD4 [3200 Mhz]
PSU: Thermaltake SMART BX1 750W 80+ Bronze
Storage: Kingston M.2 [NVMe] 240GB SSD, 2 TB HDD (disconnected HDD during troubleshooting)

Graphics Driver: Started with the latest 536.67 but have since tried 536.23 & 536.40
OS: Windows 11 (Version 22H2 [OS Build 22621.2070]

• Have updated graphics driver to the latest version and tried the aforementioned versions
• Have tried with one 8GB RAM stick in at a time
• Have tried setting the game to compatibility mode (Windows 8)
• Have run Windows Memory Diagnostic tool with both sticks in ("mdsched.exe" into Run) returned "No problems detected"
• Have run "sfc /scannow" ...on first attempt it detected corrupt files and 'repaired' them, seemed to fix nothing. Subsequent attempts have all returned "Windows Resource Protection did not find any integrity violation."
• Have reset Windows with "Keep my files" option, when that didn't work I then completed a full reset
• Have tried running a game without Discord installed

Artifacting examples:

Task Manager:

HWMonitor:

Thanks to anyone that replies

 

[Aeacus]

Have reset Windows with "Keep my files" option, when that didn't work I then completed a full reset

Full reset is not equal to clean Win installation. So, best bet would be formatting your OS drive and making a new, clean Win installation.

Though, your "artifacts" look like malware infection. With this, you'd need to format all and any drives you have in your system, to get rid of it.

I tested an older graphics driver

Downloaded from where? Since it could be most likely source of malware. Either that, or your minecraft modpack.

PSU: Thermaltake SMART BX1 750W 80+ Bronze

Why did you cheap out on PSU?

And here i don't want to hear: "i didn't have enough money", since when you have money to buy RTX 2080, you'll also have money to buy proper, good quality PSU.

 

[BananaMonkey]

Full reset is not equal to clean Win installation. So, best bet would be formatting your OS drive and making a new, clean Win installation.

Is this any different to using a USB flash drive to install Windows and formatting the drives when up to the part "Where do you want to install Windows?" I've seen that described as a "clean install." I guess "full reset" can be confused with something else so really I meant to say I performed a clean install. (this was before any of the "artifacts" started)

Though, your "artifacts" look like malware infection. With this, you'd need to format all and any drives you have in your system, to get rid of it.

Interesting, do you think there is potential for the GPU to be the cause instead? I've had someone say the "artifacts" are due to a 'dead' GPU

Downloaded from where? Since it could be most likely source of malware. Either that, or your minecraft modpack.

I can see why you would suggest the modpack but I would rule that one out. As for the drivers, I acquired them from the "Manual Driver Search" directly from the NVIDIA website (https://www.nvidia.com/en-us/geforce/drivers/).

Why did you cheap out on PSU?

Erm, that's just what was part of the deal. Are you suggesting that the PSU could potentially be at fault? Though I don't think it's too strange for the average consumer (compared to an enthusiast who assists with tech problems) to not pay much attention to the PSU. Actually, I remember there was an option to upgrade to 80+ Gold, but in the related discussion some people argued for it while others argued against it, I just went with the majority which was upgrades towards the CPU cooler and RAM speed instead.

Thanks for the reply

 

[Aeacus]

Is this any different to using a USB flash drive to install Windows and formatting the drives when up to the part "Where do you want to install Windows?" I've seen that described as a "clean install." I guess "full reset" can be confused with something else so really I meant to say I performed a clean install. (this was before any of the "artifacts" started)

Formatting the drive is key part in clean Win installation. Without format, there are other options to "repair/reset" Windows, while still keeping the installed kernel intact. Hence why i asked of clean Win install, not just reset.

Interesting, do you think there is potential for the GPU to be the cause instead? I've had someone say the "artifacts" are due to a 'dead' GPU

I don't think the issue is due to GPU, namely thanks to this symptom:

A restart got rid of them for some minutes before they came back, as did upgrading my graphics driver to a newer version. Still have them now, seems like things are just getting worse!

If hardware would be at fault, you'd have the artifacts at all time. GPU is like on/off switch. Either it displays things properly, or it displays artifacts. These artifacts you're seeing, that appear on their own, at some random convenient time, some time after you've rebooted, is not hardware failure. Malware, in the other hand, works exactly like so. Won't "start" at an instant, but does "start" after some time after the reboot.

Also, your task manager, hwinfo screenshots are artifact free. If it would be GPU issue, artifacts would be there too.

Since you have Ryzen chip, you don't have the redundancy in form of iGPU. With this, you just can't pop out the GPU and hook monitor to MoBo, to see if the GPU would be the issue.

At this point, i'd do 2nd, clean Win install. It doesn't cost you anything and if issue is software (most likely), you can fix it with clean Win installation. Since clean Win installation will remove all software issues, bloatware and malware (except rootkits).
If the issue comes back as soon as you've made clean Win installation, then we can look towards hardware issue(s).

Erm, that's just what was part of the deal. Are you suggesting that the PSU could potentially be at fault? Though I don't think it's too strange for the average consumer (compared to an enthusiast who assists with tech problems) to not pay much attention to the PSU. Actually, I remember there was an option to upgrade to 80+ Gold, but in the related discussion some people argued for it while others argued against it, I just went with the majority which was upgrades towards the CPU cooler and RAM speed instead.

Part of the deal? Like GPU + PSU combo?
If so, ever wondered why such combos are sold in the first place?

Average consumer excuse doesn't cut it really. Since if you have know-how to upgrade CPU and/or GPU to better versions, which are compatible with existing hardware, you also have knowledge about PSUs. Average consumer, in it's true meaning, doesn't know 0 about what is in their PC (hardware wise), let alone the know-how to upgrade individual components.
Though, you must realize that PSU powers everything, right? And if so, this makes the PSU most important component inside the PC. <- Overlooking PSU is not because of not knowing the real importance of PSU, but instead ignoring the cost of proper, good quality PSU, in favor of more expensive GPU, to get more FPS. Since after all high FPS matters the most, right?

Your PSU is low quality unit, borderline crap, and it has high chance to produce all sorts of issues. At worst, PSU goes "pop", releases magic smoke and fries everything it is connected to, aka your whole PC.

So, the to-do list is:
#1 clean Win install
#2 new, good quality PSU

Regarding #2, good PSUs to go for are Seasonic Focus/PRIME or Corsair RMx/RMi/HX/HXi/AX/AXi.
(All 3 of my PCs, are also powered by Seasonic, while i have two PRIME TX PSUs and one Focus PX PSU. Full specs with pics in my sig.)

For other options besides Seasonic or Corsair, any PSU in Tier A will do fine,
PSU Tier list: https://forums.tomshardware.com/thr...er-list-rev-14-8-final-update-jul-21.3624094/
While your Tt "Smart" PSU is Tier D.

 

[LinuxDevice]

I will add that if the artifacting is not from malware, then it is probably failing hardware, or hardware with too much heat on the memory.

 

[BananaMonkey]

Formatting the drive is key part in clean Win installation...

Update:

I have used my laptop to create a USB with the Windows 10 installation media on it (note that I was previously running 11 but I thought that this was a good opportunity to see if using Windows 10 changes anything, plus the fact that I prefer it) I booted using that USB, deleted partitions and formatted the drives through the custom install.

I then did the following:
- Checked for Windows Updates (there were none)
- Downloaded Chipset Drivers from my motherboard's official website (https://www.biostar.com.tw/app/en/mb/introduction.php?S_ID=934#download) then restarted PC
- Updated the BIOS to its latest version (assisted by the aforementioned website)
- Downloaded the latest Nvidia Graphics Driver manually (without GeForce Experience) through the official website (https://www.nvidia.com/en-us/geforce/drivers/) then restarted PC
- Downloaded Steam and BattleBit (to now test if things are working, nothing strange had happened up to this point)
- But unfortunately, I was met with a black window after about a minute in-game (see image)

Also curious, was wondering why the shortcuts of "SpiderHeck" and "Discord" were still showing after the wipe. I guess it has something to do with the cloud due to that symbol in the bottom left of the SpiderHeck shortcut. (see image)

I will add that if the artifacting is not from malware, then it is probably failing hardware, or hardware with too much heat on the memory.

Thanks for the reply. Yet to see any artifacting since I did the above steps, but I will mention it if it starts again.

#2 new, good quality PSU

Okay, thanks for the information! I'll keep a lookout for the ones you mentioned. I assume you still believe that this could be the issue after seeing what I said above?

In the meantime, I'm still open to suggestions from anyone. Thanks all

 

[Aeacus]

But unfortunately, I was met with a black window after about a minute in-game (see image)

Hmmm....

Something to try:
* different game
* full screen mode
* borderless full screen mode (if game supports it)

Also, your temps look fine, even hotspot on GPU. So, i don't think your GPU is overheating. Which, if it would, could be one of the culprits.


As far as we can see, you got rid of your "artifacts" issue, right?

I assume you still believe that this could be the issue after seeing what I said above?

While poor PSU doesn't cause artifacts per se, it can cause plethora of different issues, including the black screen you saw after a minute or two of gaming. Other issues include PC restarting/shutting down randomly, especially during gaming. And a chance that PC can recover from that reboot/shut down, initiated by poor PSU, are small. Since when PSU acts up and kills the power, it also has a chance to die + killing other components. The poorer the PSU quality - the higher of a chance of component killing.

 

[BananaMonkey]

Something to try...

Haven't had a crash just browsing, though I haven't spent too much time doing so. Have had no crashes that actually restart my PC either, it's just the applications that crash.

Have tried windowed, full-screen and borderless modes.

Well, went out and got a "Corsair RM750e 750W Gold ATX Modular PSU" which sounded much better than my previous PSU when the PC was starting up. Went straight to test it in the BattleBit shooting range and unfortunately crashed after just about 30 seconds, not to a black screen, it just closed the application. I tried another 'easy-to-run' game "SpiderHeck" which crashed when I tabbed out after about a minute. However, on my second try of SpiderHeck, I was actually able to play for around 5 minutes, I tabbed in and out once and that also didn't crash it, so I tried tabbing in and out a bunch to see if it would crash, and not only did it crash but it brought back the artifacts!!! Was just about to confirm that I've still had no artifacting since the last clean install too, what the! (see images)

Restarting made the artifacts go away, I tried BattleBit one last time, went into the range and to my surprise a specific wall in-game had the artifact look on it for a brief moment before my game crashed. No idea how that even happens.

At this point this must surely be a Graphics Card hardware issue, right? I do wonder what started this issue in the first place.

Well, what are the thoughts? Am I in need of a graphics card replacement and if so are there any recommendations? Was thinking of getting a second-hand 2080... is it even possible to use a 30-series or 40-series card with my current specs? Unless there's some miracle way of fixing this.

 

[Aeacus]

Well, went out and got a "Corsair RM750e 750W Gold ATX Modular PSU" which sounded much better than my previous PSU

While RMe is better, RMi or RMx would've been even better. But i digress.

At this point this must surely be a Graphics Card hardware issue, right?

Yeah, GPU seems to be the problem.

One, last ditch effort, would be running your PC in Safe Mode.
If artifacts also appear in Safe Mode, it would be further evidence of hardware (GPU) failure. But if no artifacts appear, issue could be with 3rd party software, which is automatically disabled when running Safe Mode.

I do wonder what started this issue in the first place.

IF your GPU is on it's way out, i'd look towards inferior PSU, used thus far with the GPU. Poor PSUs can feed out of spec voltages to hardware, degrading and damaging them over time, until hardware fails.

In similar example: if you eat contaminated food, you won't drop dead on an instant. Instead you get food poisoning, resulting in stomach aches and vomiting. Eat contaminated food on daily basis and you'll end up 6 feet under. Essentially the same is with your GPU and other hardware, which has been fed the "bad" power by your poor PSU you used.

Am I in need of a graphics card replacement and if so are there any recommendations? Was thinking of getting a second-hand 2080... is it even possible to use a 30-series or 40-series card with my current specs?

For new GPU, i, personally, would not buy used GPU. I'm not rich enough to buy used hardware.

This is age old debate, if to buy used hardware. Now, some hardware is completely okay to buy used, e.g CPUs. But since we are talking about GPUs, people usually fall into two categories;
1. Those who swore by it, to buy used.
2. Those who avoid it like plague.

Since there was a cryptocurrency mining craze, that bought up almost all RTX 30-series GPUs, 2nd hand market is currently flooded with used mining GPUs. Buying used GPU is questionable and buying mining GPU is even worse idea.

To give you insight, look what Steve, from GamersNexus says about it;
at 11:42;

View: https://www.youtube.com/watch?v=g1Yp6pQRdns


Coming back to why i said that i'm not rich enough to buy used;
To me, used GPU price is 50% off from MSRP. That, for the sole reason that used GPU doesn't come with warranty and i have to take a risk. Sadly, most used GPUs aren't sold as 50% less than MSRP, but instead 20% less or even at MSRP prices. <- That, to me, is too expensive to buy a used GPU. Since when used GPU should die, i have to pay far more for 2nd, used GPU, than going with new GPU off the bat.

Example:
Let's say brand new GPU is $400. Used GPU is $300. <- If i were to buy used, spending $300 and it dies, i have 0 redundancy (no warranty) and i have to pay another $300 to get 2nd, used GPU. That makes total of $600, far more than brand new GPU would've costed, if i would've gone with new one. Diff would be $200. At these prices, better to go with brand new GPU off the bat.
But let's say used GPU is $200. Then, when it dies, i need to pay another $200 to get 2nd GPU. That, in total, would be exactly same ($400) as it would've been when going with new GPU.

And if the used GPU is old mining GPU, my price would be 50% of what used GPU would've costed. Since there are huge risks with mining GPUs. Those can die within days. So, if brand new GPU would be $400, i'd pay $200 for used one but only $100 if the used GPU was used in mining.

I'd rather pay full price and have a warranty, just in case something happens; rather than seemingly paying less at first but most likely paying far more down the line. I'd rather have redundancy than taking a risk and hoping for good luck.

is it even possible to use a 30-series or 40-series card with my current specs?

Your Corsair RM750e can sustain up to RTX 3060 Ti or RTX 4070 Ti.

Performance wise, equal to your RTX 2080, would be: 1080 Ti, 3060 Ti, 4060.

Price wise, with brand new prices;
from ~$300 - RTX 4060
from ~$350 - RTX 3060 Ti
from ~$600 - RTX 4070
from ~$800 - RTX 4070 Ti

RTX 4070 Ti would be best you can put in your build, but it will also cost the most;
RTX 2080 vs RTX 4070 Ti comparison: https://gpu.userbenchmark.com/Compare/Nvidia-RTX-2080-vs-Nvidia-RTX-4070-Ti/4026vs4146

RTX 30-series and 40-series need the new +12VHPWR connector but when buying brand new, you'll get the adapter with the GPU.

 

[realpao]

Try checking your SSD.

 

[BananaMonkey]

Wowie! First off, I really appreciate such a detailed response, thanks so much.

While RMe is better, RMi or RMx would've been even better. But i digress.

Ah okay, I wasn't sure if there was too much difference. Though I'll admit I wanted it fixed for the weekend so I wanted something I could pick up rather than get delivered (I was getting a bit impatient with it all) and that was quite literally the only option in the "A-tier" of PSUs at the physical stores near where I live.

One, last ditch effort, would be running your PC in Safe Mode.
If artifacts also appear in Safe Mode, it would be further evidence of hardware (GPU) failure. But if no artifacts appear, issue could be with 3rd party software, which is automatically disabled when running Safe Mode.

So I tried this. When starting up the GPU was very loud and it must have taken at least 5 minutes to actually boot. It seems like the artifacting mainly gets triggered by a game crashing but also doesn't happen that often (then reboot seems to get rid of it) so I wasn't sure what I was going to do to actually try and get the artifacts to occur. Now I'm sure this isn't the greatest strategy but my 'stress' test method involved opening SpiderHeck (BattleBit wouldn't work in Safe Mode), 10 YouTube videos and alt-tabbing quickly (which previously caused a crash and then artifacts right after) and nothing crashed or created artifacts.

Restarted to go back to normal and roughly 5 minutes went by of the fan on the GPU just spinning with nothing happening. So I unplugged and plugged GPU back in and it started. Out of Safe Mode, I crashed trying to launch BattleBit with no artifacts and that's the latest update from me...

has been fed the "bad" power by your poor PSU you used.

Well, this has definitely taught me not to gloss over the PSU, never had a PSU problem before.

I'd rather pay full price and have a warranty, just in case something happens; rather than seemingly paying less at first but most likely paying far more down the line. I'd rather have redundancy than taking a risk and hoping for good luck.

Yeah, I thought I would hear something like this. Interesting way of looking at it too with the "why i said that i'm not rich enough to buy used" explanation. Something to consider for sure, not sure which way I'll go yet. Also, I did see someone selling an "LHR" card, looking that up it says a driver update to reduce profit for miners, but since it was an update wouldn't all the RTX cards be LHR now? Ah I think I see, that must be one of the main reasons for there being lots of used mining cards in the market. So LHR really doesn't mean anything, since they could've been used for mining before that update.

Your Corsair RM750e can sustain up to RTX 3060 Ti or RTX 4070 Ti.

Performance wise, equal to your RTX 2080, would be: 1080 Ti, 3060 Ti, 4060.

Awesome, thanks for this info.

RTX 30-series and 40-series need the new +12VHPWR connector but when buying brand new, you'll get the adapter with the GPU

I believe this came with my PSU, correct? (see screenshot for one of the cables that were in the box with the new PSU) or is there an adapter thing that the "12VHPWR" plugs into

 

[Aeacus]

Ah okay, I wasn't sure if there was too much difference. Though I'll admit I wanted it fixed for the weekend so I wanted something I could pick up rather than get delivered (I was getting a bit impatient with it all) and that was quite literally the only option in the "A-tier" of PSUs at the physical stores near where I live.

With PSUs, there is a difference. And depending on which PSUs you compare, diff can be huge.

For example; can you tell the diff between these two units,
pcpp: https://pcpartpicker.com/products/compare/QKNypg,49VG3C/

One is EVGA 650 N1 and another is EVGA 650 GT. Also, which of the two would you buy?

Now I'm sure this isn't the greatest strategy but my 'stress' test method involved opening SpiderHeck (BattleBit wouldn't work in Safe Mode), 10 YouTube videos and alt-tabbing quickly (which previously caused a crash and then artifacts right after) and nothing crashed or created artifacts.

Hmm....

My best guess is that the driver chip on your GPU is on it's way out. This would explain why in Safe Mode, everything works fine (since none of the GPU drivers are loaded, instead Win built-in display drivers are used), while it acts up when booting normally (once GPU drivers are loaded).

With this, GPU replacement would be in order.

Well, this has definitely taught me not to gloss over the PSU, never had a PSU problem before.

That's the thing with PSUs, when they act up, they will damage or outright kill the hardware. And since it will impact your wallet, people then, after the fact, take notice. But not before, when it actually matters. Still, it is a great lesson for people.
Or in other words, most people won't believe in importance of PSU, until they, themselves have experienced hardware failure caused by poor PSU.

Also, I did see someone selling an "LHR" card, looking that up it says a driver update to reduce profit for miners, but since it was an update wouldn't all the RTX cards be LHR now?

LHR can be either software or hardware. With RTX 30-series, GPU manufacturers initially made hardware version of LHR. It looks like so,
pcpp: https://pcpartpicker.com/products/compare/FZH7YJ,7MPQzy/

Two essentially same GPUs, while one is LHR. For gaming or any other use, LHR doesn't matter. But for miner, LHR cuts the hash rate in half, with 50% less profits. Hence why LHR was made in the first place.
RTX 40-series has it's LHR software based. And if it is software based, it can be disabled as well.

Though, saving grace is that after the latest mining craze was over, legislation made it so, that biggest cryptocurrency coins (Bitcoin, Etherum) can't be mined with consumer grade GPUs anymore. Now, only small/less known coins can be mined with consumer GPUs, with only peanuts as profits,
further reading if interested: https://whattomine.com/gpus
while CPUs can still be used to mine everything. Also, latest hardware to mine is ASIC chip or dedicated mining GPU (e.g Nvidia CMP 50HX specs).

Ah I think I see, that must be one of the main reasons for there being lots of used mining cards in the market.

Used mining GPUs on 2nd hand market aren't mainly because software LHR, but instead because Bitcoin and Etherum moved away from GPU mining. E.g Etherum was the hot cookie when RTX 30-series launched.

I believe this came with my PSU, correct?

Yes, this is the cable.

Though, do note that this cable is compatible only with RTX 40-series GPUs, since it's 16-pin (12+4) and not the 12-pin, used on RTX 30-series GPUs.

12-pin is compatible only with RTX 30-series GPUs,
The one on the right:

While 16-pin is compatible only with RTX 40-series GPUs,
The one at the top:

Also, since the cable only has two 8-pin connectors at the other end, it is compatible with up to 300W GPUs. Meaning that RTX 4070 Ti, at 285W is tops you can use. RTX 4080 is 320W GPU and it would need three 8-pin connectors on the other end of that +12VHPWR cable. And RTX 4090 would need four 8-pin connectors on the other end of the power cable.