Hey everyone,
So, I put together this rad new PC not too long ago, all AMD-powered. But dang, I've been dealing with GPU crashes ever since, and it's driving me up the wall.
Started with Palword crashing on me, though it's been hit or miss since then. Helldivers 2? Well, let's just say it's been a comedy of errors – crashing left and right, especially as I'm getting into the groove. Remnant 2 is mostly cool, but sometimes starts up with some crazy lag, making me restart just to get things smooth again.
Then Destiny 2 decided to throw me a curveball with a grey screen that just hung there until I gave up and restarted.
I've tried everything under the sun: clean driver installs, updating all my drivers, tweaking undervolting settings – you name it. But nada. Still dealing with crashes in all shapes and sizes – grey screens, black screens, driver timeouts, and even some funky artifacts before it kicks me out of a game.
Here's what's under the hood:
If anyone's got any tricks up their sleeve or some sage advice, I'm all ears. Help a gamer out, will ya?
Same specs here. You are not the only one having crashing issues. Just try to google "7900xtx crashing".
It's unbelievable that they haven't still been able to fix crashing problems with this really expensive GPU. I'm wondering if it's even driver related issue or is the card just so **bleep**ty? It's really frustrating and I really regret buying AMD. Going to buy Nvidia from now on.
Sorry that I can't help. I have tried almost everything and haven't found a solution.
I've spent the past month diving deep into "7900 XTX crashes" search results, and it's disappointing, to say the least. I mean, I shelled out a pretty penny for this card, coming from a 1080 Ti (which, yeah, had its own issues, but nothing like this).
Tonight, I'm gonna give swapping out the cables for the GPU a shot, just to rule out any power-related issues. Though, honestly, I'm not holding my breath on that front. I mean, this card's been chugging along for 12+ hours on weekends, pulling around 400W, so it's unlikely to be a power problem, right?
I've also reached out to the store about a potential replacement. Gonna give that a whirl, and worst-case scenario, I might have to bite the bullet and trade up for a 4080 Super. Really didn't want to go down that route, but AMD seems hell-bent on testing my patience lately.
Good luck! Make sure you have separate cables from PSU to GPU. Don't use chained ones. I have also tried swapping cables and reseatting GPU but it didn't help me.
Funny that I actually found a possible solution today after complaining here. I tried something new and after that I was able to play PUBG whole night (6 hours) without a single black screen/crash.
I did two things:
1. I uninstalled chipset drivers first and then GPU drivers with DDU in safe mode. I disabled Windows automatic driver update and disconnected internet just to make sure that Windows update doesn't mess with my clean driver installation. Then I installed GPU drivers first, restarted PC and installed latest chipset drivers.
2. After clean driver insallations I adjusted GPU clock limit to 2750 and GPU power limit to +5% in radeon software. I have read that some people have noticed GPU clock speed spiking to +3000 MHZ and that could possibly cause instability. Too low power limit could also cause crashing during those spikes if GPU doesn't get enough power.
I think the solution here was adjusting GPU clock limit. But of course it's also possible that windows update had messed up with my drivers. I didn't notice any fps drops after lowering clock speed limit (default was 3000) and seems like there is much less stuttering now.
I hope this is going to help you too!
Hey there, thanks for sharing your solution, I really appreciate it. It's interesting to hear that adjusting the GPU clock limit seems to have made a difference for you.
Just yesterday, I was in direct contact with an AMD employee trying to troubleshoot my issues. They also recommended reinstalling all drivers using the AMD cleanup utility, which I'll be doing shortly. Additionally, I'll follow your lead and reinstall the GPU drivers in safe mode with the internet disconnected. It's worth a shot.
It's notable that the AMD employee suggested running the System File Checker in Windows to identify any corrupted data. Even though my PC is relatively new, it's surprising to hear that it found some corrupted data.
Hi guys,
Sorry late to the party but I have this card with over a year and spend 6 months trying all the different solutions and only found one that makes the system stable.
As mentioned above you have to go into the adrenalin software and hard set the GPU clock. You can only do this after you start the game tho. If you set it before the game starts then the clock setting gets resets to the default and the card with start overclocking until the shaders go to high and crash the syatem. For some reason if you load up the game and the go into adrenalin to hard set the GPU clock the system stays stable.
It is a joke that after spending over 1200 euro to get the card, then spend 6 months sending it back to the company that I bought it from and they saying everything is fine with it that this problem still exists. I will never buy an AMD card again. 0 help from the company and they cant fix there own **bleep**.
Hope this will help other people out.
I have tried System File Checker too and it found corrupted data even though I also have a quite new PC. I had to reinstall Windows in order to fix that. That wasn't a solution to my problems though.
I have played couple of days with clocks and volts and I'm getting close to finding optimal and most stable settings for my GPU. You can actually adjust upper clock limit down to 2600 MHZ and it doesn't really affect to performance.
If you are not very familiar with GPU tuning, Check this tutorial: https://www.youtube.com/watch?v=mH9NGfZKU9I
Hey, thanks a lot for the tips and suggestions. I'm happy to report that after following your advice, I managed to play Helldivers 2 for 5 hours straight without experiencing any crashes, stutters, or other issues. It seems that using DDU in safe mode without internet and reinstalling all drivers for both the GPU and CPU has done the trick, at least for now.
I also tried your tuning suggestions for the GPU, and so far, everything seems to be working smoothly. I'll definitely check out the video you recommended to learn more about GPU tuning. Thanks again for your help!
Okay, never mind. Helldivers 2 just crashed. I managed to play for 6 hours, but this time, I didn't get a driver timeout; instead, the application stopped working.
Well at least that sounds positive if you were able to play that long before crashing and didn't even get the usual driver timeout. Application error could be something else than GPU related issue. I would recommend to keep trying different GPU tunings to find the most stable one. Some games seems to handle higher clocks and lower volts better than others.
My volts are now 1140 mV and GPU high limit is 2700 MHz. Haven't had crashes with this tuning yet. I have also raised my fan speeds so that my GPU temps doesn't go much over 70 C.
Yeah, there's definitely been an improvement. I was starting to worry my GPU was faulty, but thankfully, no other games have crashed except for Helldivers 2. It seems like the issues with that game are more about its coding than my hardware.
Interestingly, I played with some friends yesterday, each with different GPUs, and we encountered similar problems after a patch was added to the game. It suddenly closed twice in 30 minutes, affecting all of us. However, games like Destiny 2 ran smoothly throughout the weekend without any crashes.
I'm planning to give Palworld another shot since I noticed a crash of drivers there too. Overall, I'm pleased with the progress. Thanks a lot for your support
So far, no other game has crashed except for Helldivers 2. It appears that the issues with that game stem from the developers' side. However, it's worth noting that now the game won't crash within 2 minutes when running on DirectX 12 without using Process Lasso on the .exe or employing other methods to limit the GPU, or by using DirectX 11.
I hope that it is till only helldrivers 2 that is crashing for you. It crashes for every game I have. COD, Dying light 2. Greyzone warfare, 6 days in fallujah.
I ordered a cyberpowerpc last year. Intel I9 32 gb ram. Very nice rig.
First new computer in 10 years so i spent some money. Computer worked flawlessly for 4 months. I mostly play MW2 and MW3. I was getting 260 fps at max setting at 1440 super fast and smooth.
Then came the problems. Screen would go black. Saying no input detected. I would have unistall the drivers and reinstall drivers. Kept happening. Cyberpower sent me a new 7900 xtx. Cool. Installed it. Worked great. Then same issues. Then card completely failed. It wouldn’t even light up. Contacted cyberpower.
My 2 year warranty on parts had just expired. I was extremely pissed. I contacted Powercolor the maker of the card. The warranty guy was super cool and understanding. Filled out the Rma warranty info.
and sent him the cyberpower online receipt and he warrantied me for another 2 years. Sent me a new card.
this one seems to work great. The last month.
Some changes i made. THIS CARD RUNS HOT. IMO the default fan speed is not adequate for the heat.
i went to tuning. Disabled zero rpm for fan tuning. Enabled advanced tuning. And opened up fine tuning controls for the fan. I put fan speed to up to 100.
for temp 25c i put 60. 50c 75. 64 79 88 c i put fan speed to 100.
this.keeps the GPU hot spot under 45c even when playing MW2 for 3 hrs.
previously before i tuned the fan. I would check. And the hot spot would be 80c sometimes. I would put my hand on the computer blow out fan. You could feel the heat radiating out it.
hipefully. This helps some people out there. I was about ready to go Nvidia 7080.
im running HYPR~RX for the card setting. And its staying cool. Happy. For now