Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
Lazerface

RX 570 (4GB) Graphical glitches and crashes

Recommended Posts

Posted · Original PosterOP

Hello. So about two weeks ago I put together a Ryzen 2600 + RX 570 build. It has worked just fine until the last couple of days.

 

Two days ago the PC bluescreened while idle. After a reboot things seemed fine until I watched youtube for a few minutes, and graphical glitches such as this appeared. I first thought it might be an issue with Firefox since I didn't notice the issue in Chrome. However after trying to play some games it was clear the issue was bigger.

 

Sims 4 ran fine for a couple of minutes before going black, my monitor displaying the "no signal" message, and the computer freezing up (keyboard and mouse included). Some times it crashed to desktop, other times I had to force the PC to shutdown. After a crash or reboot the Radeon Software always displayed the "Default radeon wattman settings have crashed and been restored due to unexpected system failure" Far Cry 5 did the same during the loading screen to get into the game, same for DA: Inquisition. So I don't think what games I play matter.

 

So, I re-install my graphics drivers - no difference. I uninstall the graphics drivers in safe mode and install an older version of them - no difference.

 

I do a clean install of windows 10 and I let windows deal with the drivers (so I didn't download drivers from AMD). And hey, it worked! For about 1 day.

 

At this point it's even worse, as the graphical glitches I experienced on YouTube now appear on different sites too. However it still seemed isolated to Firefox (maybe it uses GPU acceleration?). Games still crash, but now tend to show those glitches before doing so, and I no longer get the "no input" from my monitor, and it seems to only crash to desktop. The Radeon Wattman message does no longer appear. Here is an example of me trying to play Skyrim.

 

Things I've tried:

  •  Driver reinstall & older driver
  •  Clean OS install
  •  Windows RAM Error test thingy (no issues)
  •  GPU stress test - Furmark (no issues, no glitches no crashes, fine thermals)
  •  CPU Stress test - Prime95 (no issues, good thermals)
  •  RAM Stress Test -  VMT (no issues)
  •  Disabled RAM overclock in BIOS (no difference)
  •  Updated BIOS 

So I'm scratching my head here. Especially since there are no issues in Furmark, yet in games it just doesn't seem to work. Since the card worked fine for almost 2 weeks I'm fearing that it's a hardware issue (it also worked fine for 1 week before that in an older system) As I'm writing this my screen just flashed black for a second, oh god help me. Anyone have any ideas?

 

Full system specs:

  • ASUS ROG STRIX B450-I GAMING
  • AMD Ryzen 5 2600
  • Crucial Ballistix Sport LT 2x8GB
  • Sapphire Radeon RX 570 Pulse Mini 4GB
  • Corsair CX600 600W PSU

 

 

 

 

Link to post
Share on other sites

Don't use Furmark.

Use something like Unigine Superposition.


CPU: Ryzen 5 1600 (3.6GHz; 1.25v) Cooler:  Noctua NF-A9x14 on Wraith Spire Heatsink Mobo: Asus Strix B350-i GPU: EVGA XC Gaming RTX 2070 RAM: G.Skill Trident Z 3200MHz 14CL 2x8GB SSD0: Crucial MX300 525GB SSD1: ADATA Ultimate SU800 128GB M.2 PSUCorsair SF450 Case: Fractal Design Node 202 Monitor1ViewSonic XG2401 144Hz 24" / AOC G2460PF 144Hz 24"

Laptop

Microsoft Surface

I usually edit my posts immediately after posting them, as I don't check for typos before pressing the shiny SUBMIT button. 

Spoiler

Retired Devices and Inventory:

  • Asus NV56 17" (i7-3360QM, NVIDIA GT 650M, DDR3 8GB) laptop
  • Lenovo Thinkpad X220 12" (i5 vPro 2520M, Intel HD Graphics 3000, Samsung DDR3 1300MHz 9CL 2x4GB) laptop
  • Biostar X370GTN Racing Mini-ITX motherboard
  • Sapphire Pulse RX 580 8GB gpu

 

Link to post
Share on other sites
Posted · Original PosterOP
58 minutes ago, Nocte said:

Don't use Furmark.

Use something like Unigine Superposition.

I see, I couldn't get Superposition to work but I tried VALLEY instead, which instantly got artifacts and crashed.

Link to post
Share on other sites
Just now, Lazerface said:

I see, I couldn't get Superposition to work but I tried VALLEY instead, which instantly got artifacts and crashed.

An RMA seems in order then.


CPU: Ryzen 5 1600 (3.6GHz; 1.25v) Cooler:  Noctua NF-A9x14 on Wraith Spire Heatsink Mobo: Asus Strix B350-i GPU: EVGA XC Gaming RTX 2070 RAM: G.Skill Trident Z 3200MHz 14CL 2x8GB SSD0: Crucial MX300 525GB SSD1: ADATA Ultimate SU800 128GB M.2 PSUCorsair SF450 Case: Fractal Design Node 202 Monitor1ViewSonic XG2401 144Hz 24" / AOC G2460PF 144Hz 24"

Laptop

Microsoft Surface

I usually edit my posts immediately after posting them, as I don't check for typos before pressing the shiny SUBMIT button. 

Spoiler

Retired Devices and Inventory:

  • Asus NV56 17" (i7-3360QM, NVIDIA GT 650M, DDR3 8GB) laptop
  • Lenovo Thinkpad X220 12" (i5 vPro 2520M, Intel HD Graphics 3000, Samsung DDR3 1300MHz 9CL 2x4GB) laptop
  • Biostar X370GTN Racing Mini-ITX motherboard
  • Sapphire Pulse RX 580 8GB gpu

 

Link to post
Share on other sites
Posted · Original PosterOP
52 minutes ago, onlybuilt4cubanxlinx said:

Try pushing up the voltage on the GPU.

Hi, I'd like to try that but it might void warranty right? To have to adjust the voltage on an almost brand-new card just for it to work seems like a bad omen.

Link to post
Share on other sites
1 minute ago, Lazerface said:

Hi, I'd like to try that but it might void warranty right? To have to adjust the voltage on an almost brand-new card just for it to work seems like a bad omen.

You shouldn't need to increase voltage at stock clocks. Your GPU is defective if you had to do this. Just RMA it.


CPU: Ryzen 5 1600 (3.6GHz; 1.25v) Cooler:  Noctua NF-A9x14 on Wraith Spire Heatsink Mobo: Asus Strix B350-i GPU: EVGA XC Gaming RTX 2070 RAM: G.Skill Trident Z 3200MHz 14CL 2x8GB SSD0: Crucial MX300 525GB SSD1: ADATA Ultimate SU800 128GB M.2 PSUCorsair SF450 Case: Fractal Design Node 202 Monitor1ViewSonic XG2401 144Hz 24" / AOC G2460PF 144Hz 24"

Laptop

Microsoft Surface

I usually edit my posts immediately after posting them, as I don't check for typos before pressing the shiny SUBMIT button. 

Spoiler

Retired Devices and Inventory:

  • Asus NV56 17" (i7-3360QM, NVIDIA GT 650M, DDR3 8GB) laptop
  • Lenovo Thinkpad X220 12" (i5 vPro 2520M, Intel HD Graphics 3000, Samsung DDR3 1300MHz 9CL 2x4GB) laptop
  • Biostar X370GTN Racing Mini-ITX motherboard
  • Sapphire Pulse RX 580 8GB gpu

 

Link to post
Share on other sites
Posted · Original PosterOP
23 minutes ago, Nocte said:

You shouldn't need to increase voltage at stock clocks. Your GPU is defective if you had to do this. Just RMA it.

Thank you, I'll contact the seller

Link to post
Share on other sites
2 hours ago, Lazerface said:

Hello. So about two weeks ago I put together a Ryzen 2600 + RX 570 build. It has worked just fine until the last couple of days.

Possible that the GPU just died.

Its not uncommon that it dies within a couple of weeks for example because something went wrong in Transportation.

2 hours ago, Lazerface said:

Two days ago the PC bluescreened while idle. After a reboot things seemed fine until I watched youtube for a few minutes, and graphical glitches such as this appeared.

That could mean that it might not necessarily be the GPU but the PSU.

But either way, I'd rather not wait and RMA the GPU anyway.

2 hours ago, Lazerface said:

Full system specs:

  • Corsair CX600 600W PSU

How old?
The initial CX600 was released almost 10 years ago...


"Hell is full of good meanings, but Heaven is full of good works"

Link to post
Share on other sites
1 hour ago, onlybuilt4cubanxlinx said:

Try pushing up the voltage on the GPU.

No, don't!
 

The GPU has to work out of the box without firing up Wattman!
If it doesn't there is something wrong!

Either the GPU is damaged or the PSU is dying.


"Hell is full of good meanings, but Heaven is full of good works"

Link to post
Share on other sites
Posted · Original PosterOP
2 hours ago, Stefan Payne said:

Possible that the GPU just died.

Its not uncommon that it dies within a couple of weeks for example because something went wrong in Transportation.

That could mean that it might not necessarily be the GPU but the PSU.

But either way, I'd rather not wait and RMA the GPU anyway.

How old?
The initial CX600 was released almost 10 years ago...

Honestly have no clue how old the PSU is, got it with my case a few years ago from a friend. Sadly I don't have access to another PSU to see if that's the root of the problem. But if the PSU was the issue would not other components be affected as well?

Link to post
Share on other sites
Just now, Lazerface said:

Honestly have no clue how old the PSU is, got it with my case a few years ago from a friend. Sadly I don't have access to another PSU to see if that's the root of the problem. But if the PSU was the issue would not other components be affected as well?

Might be that they are less susceptible to that issue, have more Capacitance and other stuff.

I think its a good idea to also replace the PSU with the RMA of the GPU.

 

Look or ask here, get a Tier B or better:

 


"Hell is full of good meanings, but Heaven is full of good works"

Link to post
Share on other sites
Posted · Original PosterOP
On 4/29/2019 at 10:58 PM, Stefan Payne said:

Might be that they are less susceptible to that issue, have more Capacitance and other stuff.

I think its a good idea to also replace the PSU with the RMA of the GPU.

 

Look or ask here, get a Tier B or better:

 

I have now tried switching to my ancient XFX HD 4870 1GB, and I'm having zero issues. Considering this card has a higher TDP than the rx 570 (and is being pushed harder), is it an ok assumption that the PSU is not part of the issue? I want to eventually switch out the PSU either way since it's a spiderweb of cables, I mostly switched to make sure that the RX 570 is actually defect and needs to be sent back. Anyway, just some random info!

Link to post
Share on other sites
31 minutes ago, Lazerface said:

I have now tried switching to my ancient XFX HD 4870 1GB, and I'm having zero issues.

The 4870 doesn't have the Power Saving stuff the RX570 has...

 

31 minutes ago, Lazerface said:

Considering this card has a higher TDP than the rx 570 (and is being pushed harder), is it an ok assumption that the PSU is not part of the issue?

No, because the Changes the RX570 has when in idle, that caused it.

With the 570 you don't have a forced idle, you have a lose idle clock that gets increased as soon as you do something. Just check it with GPU-Z -> Sensor.

THere you see the clock jumping to something hiher when you for example scroll.

 

And My guess would be that the CX600 might not like that. The Age of the PSU is also something that might or might not have something to do with this.

31 minutes ago, Lazerface said:

I want to eventually switch out the PSU either way since it's a spiderweb of cables, I mostly switched to make sure that the RX 570 is actually defect and needs to be sent back. Anyway, just some random info!

Yeah, it might be defect, might be the PSU.

 

Though if you can RMA the GPU, its not the worst thing to do...


But also keep in mind that the PSU could be the issue...


"Hell is full of good meanings, but Heaven is full of good works"

Link to post
Share on other sites
Posted · Original PosterOP
21 minutes ago, Stefan Payne said:

The 4870 doesn't have the Power Saving stuff the RX570 has...

 

No, because the Changes the RX570 has when in idle, that caused it.

With the 570 you don't have a forced idle, you have a lose idle clock that gets increased as soon as you do something. Just check it with GPU-Z -> Sensor.

THere you see the clock jumping to something hiher when you for example scroll.

 

And My guess would be that the CX600 might not like that. The Age of the PSU is also something that might or might not have something to do with this.

Yeah, it might be defect, might be the PSU.

 

Though if you can RMA the GPU, its not the worst thing to do...


But also keep in mind that the PSU could be the issue...

I see, thank you for all the info, you learn something new every day.

 

So yeah, guess I'll send the GPU back if the store has no issues with that, get a new one and if the problem persists I'll get a new PSU whenever I get the chance!

Link to post
Share on other sites
Posted · Original PosterOP

Update for anyone with a similar issue: Got a new 570 today, seems to be working fine so far, so at this point it seems the card was the issue. If it starts acting up I'll post another update!

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Buy VPN

×