I've been having an issue lately where my graphics card (Radeon RX-480) appears to crash at random times when playing games. This happens most frequently when playing Overwatch, where it will either scramble the display and force a complete restart of the computer, or it will cut the signal briefly and give an error that says "Your rendering device has been lost". Other games have had various graphics crashes as well (VRChat, The Long Dark, Just Cause 2, Sea of Thieves, etc.).
A computer repair shop tested the graphics card in a different machine, and it ran fine, so it's not the card itself. After this, I reset Windows 10 and tried playing Overwatch again afterwards; it seemed to run fine for a couple of hours, but then the display scrambled, started turning off and on, and finally I got a blue screen as the whole computer crashed and force restarted. I have run a memory test on my computer and found no errors, so now I'm concerned it's the motherboard (I figure it is most likely the motherboard and not the processor, since I don't seem to crash when I'm not running 3D games).
My system specs:
Processor: AMD FX(tm)-6350 Six-Core Processor 3.90 gHz
Memory: 10.0 GB RAM DDR3
Computer model: HP Envy h8-1534
Motherboard: M3970AM-HP (Angelica2), with AM3-b processor socket
Graphics: AMD Radeon RX-480
You forgot to list the specs of the most important piece of hardware... your power supply! ...but a quick search suggests that it runs a 460W, which is kind of the bare minimum for a discrete card and 6-core processor.
The RX 480 doesn't require a lot of power to run properly, but if you've got a cheap OE power supply you could run into issues after hours of prolonged and heavier power draw. I'd recommend picking up a quality 500W unit and giving it another go.
It's not the power supply that came with the computer. I've upgraded it recently to a 650W.
Actually, I'm not sure what the wattage is, I was reading it off a box but I forgot that my wife also upgraded her power supply recently, so that might have been hers. Mine was more than hers though. I tried to look it up on the power supply itself but I couldn't find the label. Suffice to say, it's more than enough to run everything though.
if you susspect you could have an issue with the power supply. Test it with a free utility called, OCCT from OCBASE_com. It has a great PS and GPU test.
Do the following things and you will likely see your issues go away.
1.) In Radeon Settings / Gaming Tab / Global Settings (top left) / Global Wattman: down bellow the fan settings, slide the red "Power Limit" horizontal slider, all the way to the right to it's maximum. This will allow the card to receive more power. Don't worry this is still in the safe limits of the card and is not an over volt.
2.) Windows Power Plan setting to "High Performance" if it is not already.
4.) Disable "Fast Startup" (Windows 10 Only)
5.) Disable all Microsoft Game Features like "DVR" in Settings Game (Windows 10 Only)
You can Google any of these on how to do it.
After this you will have a much better gaming experience with your AMD product, as this fixes about 95% of any current problems.
1.) Ok, I have it set now. I'll test to see how this affects it overall.
2.) Already on High Performance.
4.) Already disabled.
5.) Already off.
Okay, while it seems like it's improved, I just had a crash in Sea of Thieves a few moments ago. It gave the following error:
Could you post a DX diag report of that?
How do I do that?
Google directx diagnostic tool. You will find lots of help. If I send you a link it will get held up to by the moderators till late tonight.
Also, I finished running a 1-hour GPU test in OCCT, no errors found. About to try it again with the power supply test.
No errors on the power supply test either. It could be that increasing the power to the fan worked, and that the Sea of Thieves error I got is unrelated? Hard to tell.
I wouldn't worry too much about some random errors as long as overall stability is very good. Even on the best build you still run into random issues.
Ok, I haven't played Overwatch for a few days and was just now testing it again, and I got the crash again. "Your rendering device has been lost."
No idea what to do from here.
I've made a DxDiag.txt, but I'm pretty sure I shouldn't be posting that publicly. What should I do with it?
Only machine name/machine id, you can edit/delete that part.
Send it as an attachment, top right when you open reply, click advanced editor, then bottom right is 'Attach'.
I don't see any advanced editor; I don't even see an edit option.
I'll just upload it to Dropbox: Dropbox - DxDiag_modified.txt
I gotta say mostly when I see errors like that they are power related. If raising the power limit to the max doesn't fix it then, you might add a small under clock. If that doesn't do it. I think I would suspect it is a power supply. Just sucks when you do all that and it still isn't fixed. Unfortunately Overwatch just isn't a game I play. I can tell you historically it is one I see complained about constantly and my kid plays it and has had issues with it on his Nvidia card. I don't think many of the issues are driver related more game engine related. Now not saying that is your issue either as your error is one we see commonly and usually links to a power issue.
Well I did get a new power supply, and that's when the issue started. I was having issues where my computer was shutting itself down entirely, and it was traced to the power supply, so I got a new one (with more wattage than the last, just to be safe). That actually caused the circuit breaker in my den to trip frequently, which I (temporarily) solved by running an extension cord to an outlet in a different room. But then these graphics card issues started.
Near as I can tell, there's no pattern for WHEN it will happen. I can play Overwatch for a few hours just fine and then it will crash, or I could play it for 5 minutes and crash. It is also crashing with seemingly any other program that runs 3D graphics; since the issue might not happen for hours, I can't confirm it on all programs, but what I can confirm is that it has never once crashed when NOT running 3D graphics (for example, I sometimes watch YouTube or browse websites for hours at a time and it never crashed).
These are surely tough situations. Not impossible either a new supply isn't bad. In fact new stuff is more likely to be bad than old stuff. Usually it is either bad right away or it lasts for years. That being said it can be motherboard, it could even still be just a registry corruption. All I know is normally a run of DDU installing a last known good driver and giving that power limit a slide to the right mostly fixes this stuff.
I might have found the problem. Remember earlier when I said I ran an extension cord to another room to avoid tripping the breaker for my den? Well, I put it back in the regular outlet, and now the crashes appear to have stopped. I just played about 3 1/2 hours of Overwatch without a crash, my longest in a long time. I don't understand WHY that might have fixed it, but I'm grateful either way if it does.
Now I just have to figure out how to solve the den breaker problem.
Figure out what else is on that line. Could be creating line noise or spiking the power somehow. You might want to feel those power lines if you can get to them make sure they aren't hot.
Glad you are gaming though!
I take it back. I had Sea of Thieves and Overwatch crash on me again, so that was not the problem.
Currently I'm trying a different power supply. I swapped mine (550W) with my wife's (650W). If it works, then I know the problem was a bad power supply. It may take some time though; the crashes do appear to be completely at random, and it went at least 6 or 7 hours of gameplay before it happened again. That being the case, when the computer store tested the video card for 4 hours, maybe it just didn't have the issue at that time, meaning that it might still be the video card...
I haven't posted in a while, but it's time for an update.
The power supply was not the issue, the game continues to crash. I have one recourse for testing left, and that's to swap it with the video card on my wife's computer to see whether it crashes or not. The big problem, however, is that she doesn't run any graphically intensive games like I do, sticking mostly to MMOs. To test to see if they would trigger the crash, I loaded one of the MMOs she plays (Mabinogi) on my computer and left it running for almost a full day. No crashes. But then I started up Fallout 4 and got a crash within 5 minutes, and another after about 2 hours of gameplay.
I'm going to try and see if a utility like OCCT, if left running long enough, will trigger it. If it does, then I can just load that utility on her computer.
I just ran OCCT for 4 hours with no crash. This has become really frustrating, because I can't seem to find a reliable way to make it crash, which makes testing nearly impossible.
When I get the chance, I'm just going to install Overwatch on my wife's computer, put the video cards in it, and give it a good 8 to 10 hours of gameplay. If it does the video crash even once, I know the card's the problem.
I don't know about see of thieves, I do not that historically Overwatch is a game that has lots of crashes even the Nvidia guys complain about it. I think for such a big game it has a less than stellar optimized engine. Not trying to point the finger elsewhere either as I fully realize it can be related to just about anything. As you already discovered figuring out what is tough!
Another thing worth trying is to disable Full Screen Optimizations for that game. Right click on the .exe file for that game and the check box is on the compatibility tab.
I think we found the answer!
So I put the card in my wife's computer, and hers in mine, and her computer did the same crash, meaning it had to do with the card and not the motherboard. However, she noticed something in the way it crashed that led her to believe something was happening with the drivers, like Windows was trying to update the driver automatically and failing. She messed around with the Windows 10 settings, and found something that may have fixed it, so I put the card back in my computer and changed the same setting. Here are the steps:
Since I changed this about day and a half ago, I haven't had a single crash (that's the longest it's gone so far without one). If it goes the rest of the week, I'm going to declare success.
Nevermind, just did it again. Aargh...
Well now I'm completely out of ideas.
Same problem,games keep crashing but im using R9 390,took the PC to 2 shops for diagnostics,both cant find the problem,gonna send it to 3rd place this week.
Starting to think amd messed up the drivers,cause this happened 3 weeks ago when I updated the drivers.
Okay, I think I finally found the trigger.
I was beginning to suspect my HTC Vive had something to do with it, since I had gotten it around the same time that the issues started. I decided to try a test. I reinstalled my drivers and restarted the computer, and started using it normally except that I did not power on the Vive (I usually have it unplugged from power when not in use, since it overrides my speakers and microphone). After doing this, I went for about 3 to 4 days without a crash before deciding to power on the Vive again. As soon as I powered it on (which I did for just a minute, and then powered it off and completely unplugged it again), I had a crash within about 20 minutes, and another one about 15-20 minutes after that. I restarted again, leaving the Vive unplugged, and everything worked normal again.
For some reason, the Vive coming online is doing something to the system that is causing it to start having this glitch. All it took was the system recognizing the Vive for merely 1 minute for it to start acting this way. I figure it has to be something with the way the system is treating the monitors, like the Vive coming onto the system is confusing it somehow. The video card feed shuts down momentarily as if it is switching monitors or installing new drivers (even though nothing actually installs), even once the Vive is completely removed from the system (both the USB and the HDMI unplugged).
I know what the trigger is now, but as for fixing it, I have no idea. My only option currently is to just restart the computer after using the Vive as a workaround, unless someone can suggest something else to try.
Now you've mentioned VIVE
https://www.game-debate.com/games/index.php?g_id=23700&game=HTC%20Vive
which has a 4-6Gig requirement as well (and thats on win7), its no wonder its crashing
this set up would pretty much wipe out my sons specs... he would be looking for something with min 64gig ram.
So I have further refined the Vive issue. I've been running everything fine for quite a while without any crashes simply by not using the VR system, or by restarting after using it. Recently, I started up SteamVR by accident and closed it; keep in mind that the VR system was not even plugged in or powered on. Shortly after this happened, I had a crash, and after restarting it went back to normal as usual.
I now believe that it's not the Vive specifically that's causing the trouble, but SteamVR. I will be focusing my attention on that instead, to see if maybe there's some aspect of SteamVR that stays active even after it is closed.
sorry if i gave the impression it was vive, i was trying to suggest that i believe the culprit to be the system ram limitation.
overclocking and page file writes can only do so much for the lack of ram available.
"I will be focusing my attention on that instead, to see if maybe there's some aspect of SteamVR that stays active even after it is closed."
let us know how this goes, because I would still be interested....... mainly because I know how normal steam upsets my computer (being a weaker build than my sons running win7 with 8gig ram) currently, even with overlays and stuff turned off.
My private opinion only
I dont know enough about the differences between the cards and their capabilities, however Im almost willing to bet there just isnt enough ram between the OS, Video Card, Virus Firewall protection to run with only 10gigs and maybe a dust blow out too would help.
your system specs:
Processor: AMD FX(tm)-6350 Six-Core Processor 3.90 gHz
Memory: 10.0 GB RAM DDR3
Computer model: HP Envy h8-1534
Motherboard: M3970AM-HP (Angelica2), with AM3-b processor socket
Graphics: AMD Radeon RX-480
Im Wondering if it isnt a RAM issue
My sons Specs:
Processor: AMD FX 8320 Eight-Core Black Edition CPU 3.5GHZ
Memory: 32 GB RAM 1600mhz DDR3 (27% free)
MotherBoard: Gigabyte GA-970A -D3P
Graphics:
Now Primary: MSI TwinFrozr vi Radeon RX470 8G GDDR5 (requires 8gig ram to run, says it on box)
secondary (origonal NVidea Ge Force GTX 740 2Gig): GeForce GTX 740 2Gig
runs all games on high since the rx470 was put in
Sample of normal apps running: (ie constantly running)
Win10
Discord
Steam
Opera Browser
Norton Security suite
Warframe
other background junk unsure
Hi, 10GB ram… Do you have two memory sticks? Thanks.
I have the same problem all the 3D games are crashing, GTA5, Project AURA, NO MAN'S SKY, etc.
I have done clean install with DDU. My settings are
GPU: SAPPHIRE RADEON 360
MOTHERBOARD: ASROCK FM2A88M EXTREME4+ R2.0 BIOS VERSION 1.80
DESKTOP
WINDOWS 10 64 BIT HOME version1809
RADEON 18.12.3 ADRENALIN EDITION
DISPLAY HDMI TO VGA, LG 20KM400A,1366X768 60HZ ONE SCREEN
APU AMD A10-7860K
RAM 8GB DUAL CHANNEL DDR3
POWER SUPPLY UNIT 230V FORCE MODEL SL-X750EPS MAX
I have checked it with OCCT, no problem.
I have included my dxdiag.txt
Mine pc was crashing with blue screens, I solved it by updating the pcI slot, sata atapi drivers of the system.
Having the same issue with my pc , x370 gaming plus motherboard,ryzen 7 2700x ,16gb sniper x ram running at 2133 ghz,730 watt power supply. Crash on game with the gtx 1070 ftw , but does not with the rx 580 8gb golden red devil...thinking the ram is set to low?
Also the crash just bring me back to window screen ,it does not crash the whole pc