Support NeoGAF

[Übermatik]

So since I built this machine I've been getting some very weird problems. First, my specs:

AMD Ryzen 7 1700 CPU
Asus X-370 PRIME PRO Motherboard
Corsair RM550X PSU (slight coil whine)
Corsair LPX 16GB DDR4 3200Mhz RAM
ASUS RX480 8GB Dual OC

After playing Battlefield 4 for a while, my computer would lock up and crash to a grey screen. I'd kill it, restart and sometimes (though not always) see AMD Wattman pop up telling me it has reset to default settings because of a problem.

Riiight... well looking into this issue, I found a loooong thread on AMD's community forums full of replies from people with the same issue (along with many similar threads): https://community.amd.com/thread/204322?start=0&tstart=0

Now everything else seemed to run okay, until today PUBG did this crazy shit:

https://thumbs.gfycat.com/HardQuestionableAsiantrumpetfish-size_restricted.gif

My first guess was a memory issue... am I close? What's going on?
It's worth noting that all my temps are fine (CPU, GPU etc.), voltage is fine (I think) and fans are all in working order etc. I'd also previously done several stress tests with OCCT and Unigine, with no issues.

Up until now I'd suspected my PSU - it emits a very high frequency buzz whenever I do anything (scrolling on a webpage, even, vs. a constant buzz when running games/3D apps), but now I'm not so sure...

GAF please, for my sanity and my wallet, any advice? Achieving a first time build only to find problems like this down the line is very disheartening!

 

[Kindekuma]

To me, it looks like your GPU is starting to shit itself. I know you said your PSU is acting up, but try to feed a little bit more power to your GPU to see if it might be dying.

 

[Transistor]

Artifacting sounds more like a GPU issue, but try replacing the PSU first as it's undoubtably cheaper

 

[digitalrelic]

Yeah, definitely a GPU problem. I would try downclocking it a bit and seeing if that fixes the problem.

 

[bloodforge]

Always try replacing the PSU first. It is always the most likely culprit.

 

[Randomizer]

Sounds like a GPU problem most likely. The cause of which could be the GPU itself or even as you suggested a PSU issue. I suggest you try and get a hold of another PSU to troubleshoot before RMA'ing anything. If the issue is gone it was the PSU, if it remains the GPU is the problem.

 

[motorised mallard]

I'll throw my two cents in the GPU hat also.

My WoW started to look just like that back in the day when my GPU was about to give up. I managed to get some amazing screenshots because of it though.

 

[Duxxy3]

Yeah, definitely a GPU problem. I would try downclocking it a bit and seeing if that fixes the problem.

Yup. Try this. Try default settings for a regular 480.

 

[Theblackrain]

First thing to do is take everything back to stock. Use default settings in the UEFI, remove any CPU overclock and keep the GPU at stock. If the problem persists, it's a process of elimination. If you have access to another known working GPU, try that and see if cures the issue. If it doesn't then try replacing PSU. You are probably hearing inductor whine from the PSU, it's not uncommon and it's harmless, though annoying.

 

[J_Viper]

Don't mean to hijack your thread OP, but I was having the same problem on my PC, a pre-built PowerSpec

For the first few months, it was smooth sailing, but then I started to get hard locks on a few games. It all ended with my SSD becoming corrupted.

I'd also like to know where these issues come from.

I mean, I assume the SSD was the problem, but could a bad CPU/GPU lead to a SSD failure like that?

 

[Theblackrain]

Don't mean to hijack your thread OP, but I was having the same problem on my PC, a pre-built PowerSpec

For the first few months, it was smooth sailing, but then I started to get hard locks on a few games. It all ended with my SSD becoming corrupted.

I'd also like to know where these issues come from.

I mean, I assume the SSD was the problem, but could a bad CPU/GPU lead to a SSD failure like that?

I would say that was just a faulty SSD. What brand/model was it?

 

[Budi]

I'll join others in saying it's probably GPU. Also had artifacts popping in when my GPU busted.

 

[bluntspoon]

It seems trite but my standard answer is 90% of the time it's the PSU 100% of the time.

For me at least PSU's fail more often than any other component and I buy decent mid level units.

Except for my 420W OCZ Powerstream. That thing has been a tank. 12 years later it's sitting in my daughters computer happy as a little lark.

*edited for spelling

 

[Cthuluhoop]

If the PSU is fucked and it's causing the GPU to artifact rendered scenes, doesn't that mean the PSU has fried the GPU? Probably an internal voltage issue in the PSU that's letting too much power through? I don't entirely know but the high pitch noise of the PSU and now the GPU dying it's more than likely that the PSU has killed the GPU. Imo you should stop using immediately and contact the PSU manufacturers to report the problem and state you want a new GPU.

I would say that was just a faulty SSD. What brand/model was it?

Faulty or dying SSD would cause stuttering though wouldn't it? Download HD tune or something to test.

 

[Übermatik]

If the PSU is fucked and it's causing the GPU to artifact rendered scenes, doesn't that mean the PSU has fried the GPU? Probably an internal voltage issue in the PSU that's letting too much power through? I don't entirely know but the high pitch noise of the PSU and now the GPU dying it's more than likely that the PSU has killed the GPU. Imo you should stop using immediately and contact the PSU manufacturers to report the problem and state you want a new GPU.


Faulty or dying SSD would cause stuttering though wouldn't it? Download HD tune or something to test.

Is this possible? It's something I feared. Can anyone else corroborate on this?

Also I should stress that nothing is overclocked. Everything is at default settings, and using up to date BIOS version.

 

[Jackpot]

Check your GPU temperatures.

 

[Übermatik]

Check your GPU temperatures.

Temperatures are fine.


Is there any way I could read a faulty PSU? I'm not too savvy on voltage readings - what should I look out for? The voltages seems to jump up and down quite a bit, is this a problem?

 

[opticalmace]

Probably GPU... or maybe drivers. Do you have another video card you could swap in there to test it out?

 

[Übermatik]

Probably GPU... or maybe drivers. Do you have another video card you could swap in there to test it out?

Unfortunately not, nor another PSU, so my diagnosis is really limited

Drivers are up to date, and this problem has persisted through many driver updates.

 

[opticalmace]

Unfortunately not, nor another PSU, so my diagnosis is really limited

Drivers are up to date, and this problem has persisted through many driver updates.

Local computer shop could probably check your PSU for you (I think many of them have special PSU tester machines--just ask nicely and they'll probably check it for you). Though, I think GPU is more likely.

What about a friend with an extra video card? Or, where did you buy your card?

If driver rollbacks aren't helping then my guess would be video card...

edit: you could always underclock your GPU to see if that helps. Might not be stable at its 'stock' settings.

 

[Übermatik]

Local computer shop could probably check your PSU for you (I think many of them have special PSU tester machines--just ask nicely and they'll probably check it for you). Though, I think GPU is more likely.

What about a friend with an extra video card? Or, where did you buy your card?

If driver rollbacks aren't helping then my guess would be video card...

edit: you could always underclock your GPU to see if that helps. Might not be stable at its 'stock' settings.

PC shop is a good shout, I could give that a go. What's the tool I need to ask for?
Both PSU and GPU came from Amazon... I'm actually not too keen on getting a fix/replacement, I'd far rather get a new make/model (maybe try fork out for a 1070 instead), but Amazon probably won't reward that request... What do you think? Was about March this year I bought them.

Underclocking, maybe... but I'm not happy using a product that doesn't work as advertised.

 

[Duxxy3]

PC shop is a good shout, I could give that a go. What's the tool I need to ask for?
Both PSU and GPU came from Amazon... I'm actually not too keen on getting a fix/replacement, I'd far rather get a new make/model (maybe try fork out for a 1070 instead), but Amazon probably won't reward that request... What do you think? Was about March this year I bought them.

Underclocking, maybe... but I'm not happy using a product that doesn't work as advertised.

Assuming Amazon won't let you return it and Asus won't fix it - If you have to choose between underclocking or buying a new card, which would you prefer?

 

[Übermatik]

Assuming Amazon won't let you return it and Asus won't fix it - If you have to choose between underclocking or buying a new card, which would you prefer?

Considering it's well within warranty and only 2 months old, I'll make them fix it.

 

[jryeje29]

If the card isn't overclocked then either the card is bad or the PSU isn't providing the proper amount of power, first try a new PSU or throw the card in a friend's PC and see if the problem repeats in either scenario. If the card is overclocked then revert it. Also try a fresh driver install after cleaning the drivers off with ddu, I doubt this is the issue but it's free to try so why not.

 

[BigDes]

HAve you checked your GPU temperatures when under load?

It does sound like it is starting to die

 

[dsk1210]

If you are getting artifacts then my money is on the gpu, the psu would probably just hard lock or reboot the system.

Amazon are fine for returning stuff. Send the gpu back.

 

[Bydobob]

Did I miss the part where OP said he hasn't overclocked his GPU?

First thing I'd check. Just because an OC runs fine in synthetic benches doesn't mean it'll run well in all games. I can run stupid overclocks in Heaven for example, and then have an artefacting light-fest in games. Reset to factory and see if the problem goes away.

 

[J_Viper]

I would say that was just a faulty SSD. What brand/model was it?

My bad, forgot I posted here

It was the stock SSD that came with the computer, I actually can't find the brand

 

[Übermatik]

HAve you checked your GPU temperatures when under load?

It does sound like it is starting to die

Temps are fine, high 70s under load.

If you are getting artifacts then my money is on the gpu, the psu would probably just hard lock or reboot the system.

Amazon are fine for returning stuff. Send the gpu back.

My thoughts too. What happens though is my screen goes grey, there's a buzzing and my computer just dies but stays powered on. I then have to kill it by holding down the power button before rebooting.
The fact the screen goes grey and Wattman pops up sometimes after has me believing it's a GPU issue.
I'm not sure Amazon will play ball as the return option under My Orders has disappeared.

Did I miss the part where OP said he hasn't overclocked his GPU?

First thing I'd check. Just because an OC runs fine in synthetic benches doesn't mean it'll run well in all games. I can run stupid overclocks in Heaven for example, and then have an artefacting light-fest in games. Reset to factory and see if the problem goes away.

No overclocking of any sort, all at stock clocks/factory settings.
I've tried fresh installs of drivers with DDU too.

 

[big_z]

I'm on mobile so hopefully another gaffer can provide links but there are programs that will check ram for errors, stress test only the gpu or CPU Etc.

When I was having issues A few years back I used furmark to make sure it was the gpu. Not sure if it's still recommended or if there are better alternatives to check now.

Edit: if you do use furmark, keep an eye on the temps

 

[JaseC]

I'm on mobile so hopefully another gaffer can provide links but there are programs that will check ram for errors, stress test only the gpu or CPU Etc.

When I was having issues A few years back I used furmark to make sure it was the gpu. Not sure if it's still recommended or if there are better alternatives to check now.

Edit: if you do use furmark, keep an eye on the temps

- MemTest86 (for RAM testing)
- FurMark (for GPU stress testing)
- AIDA64 (for CPU stress testing)

 

[PleaseBeGentle_]

If it's ur GPU u can always use the oven trick to try and salvage it

 

[big_z]

- MemTest86 (for RAM testing)
- FurMark (for GPU stress testing)
- AIDA64 (for CPU stress testing)

- MemTest86 (for RAM testing)
- FurMark (for GPU stress testing)
- AIDA64 (for CPU stress testing)

Perfect, thanks JaceC

Run those OP. Start with the gpu.

 

[Übermatik]

Thanks big_z/JaseC, but Amazon have agreed to take returns of both my GPU and PSU! Saves the hassle of diagnosing either. I did some brief testing last night and came to the conclusion that my GPU was certainly dying. As for the PSU, I'm not 100% sure, but the coil whine is reason enough to send it back.

I'm getting a replacement RM550X and a refund for the GPU as it's out of stock... now I need to shop for an alternative. Bit wary of Asus now, and a little put off of AMD cards, but the 580 is looking like the suitable replacement. I'd love to fork out for a 1070, but can't quite afford it... hmmm.

 

[Cthuluhoop]

Is this possible? It's something I feared. Can anyone else corroborate on this?

Also I should stress that nothing is overclocked. Everything is at default settings, and using up to date BIOS version.

Thanks big_z/JaseC, but Amazon have agreed to take returns of both my GPU and PSU! Saves the hassle of diagnosing either. I did some brief testing last night and came to the conclusion that my GPU was certainly dying. As for the PSU, I'm not 100% sure, but the coil whine is reason enough to send it back.

I'm getting a replacement RM550X and a refund for the GPU as it's out of stock... now I need to shop for an alternative. Bit wary of Asus now, and a little put off of AMD cards, but the 580 is looking like the suitable replacement. I'd love to fork out for a 1070, but can't quite afford it... hmmm.

Nice! I used to work with computer parts quite a lot including refurbishment and new builds. We've seen this quite a lot with faulty PSU's that don't regulate voltage properly. It's a rare occurrence though and happend probably 1% of the hundreds of computers we built.

Happy that Amazon has been good to you!

 

[Übermatik]

Nice! I used to work with computer parts quite a lot including refurbishment and new builds. We've seen this quite a lot with faulty PSU's that don't regulate voltage properly. It's a rare occurrence though and happend probably 1% of the hundreds of computers we built.

Happy that Amazon has been good to you!

Me too! They've really cut the time out of getting these parts replaced too, Corsair weren't really playing ball.

Am I at risk using my computer now? Or is this just an issue that materialises with high load use? I don't experience any issues until I start to play games.
And yeah, it looks like I lost out in the lottery... hoping the replacement runs fine and that my next GPU, whatever it is, operates without a hitch.

 

[Übermatik]

Just want to put one more issue out there - occasionally my PC seems to 'stick' or 'hang'. Example: I'll be playing music in Chrome (YouTube) when I notice the sound suddenly stops and buzzes, almost like the record is stuck. I'm not sure if anything else freezes (it's momentary, less than a second), but it has me worrying. Could that at all be PSU related? Or do I just have an audio driver (recently updated)/browser issue?

My guess would be that if there were a serious PSU issue my PC would crash/fail to start.
Basically I don't want to risk further damage. New one comes on Monday, so might turn my PC off until then...

 

[TransTrender]

Downclock your GPU, take the side off of your case, and get an extra fan to blow in the or on your GPU.
Also override your fan settings for the GPU and crank them up all the way.

 

[dsk1210]

Me too! They've really cut the time out of getting these parts replaced too, Corsair weren't really playing ball.

Am I at risk using my computer now? Or is this just an issue that materialises with high load use? I don't experience any issues until I start to play games.
And yeah, it looks like I lost out in the lottery... hoping the replacement runs fine and that my next GPU, whatever it is, operates without a hitch.

The last time I had a dodgy psu it fucked everything basically, when I hooked up the new psu I found out the previous one had fried 3 hard drives and the Cpu was not playing ball all anymore. Thankfully the gpu survived.


Hopefully you have not had surges causing damage to the motherboard/Cpu.

Good luck dude, hope it all works out for you.