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

Can't figure out what the problem is!!!

Recommended Posts

Posted · Original PosterOP

I bought a used 980ti from CeX, when I plugged it in it booted up fine, but as soon as it went underload my entire pc would crash, I assumed it was the power supply so I bought a new one and the same thing would happen, I don't get a BSOD or anything it just shuts off and powers back up like nothing ever happend,

If i downclock the gpu its would still crash so i tried setting the power target going down in 5% intervals, and by the time i reached 60% power target it was stable for about 15 minutes in heaven and crashed.

When using a 1050ti from zotac the system runs fine.....

 

I have tried new power supply.

Messed about with driver versions.

re-flashed gpu bios.

reinstalled windows.

reinstalled all my chipset drivers e.t.c.

updated my mobo bios.

resetted everything to make sure that isn't a problem.

tried putting the card into a completely different system to rule out mobo, cpu, or ram problems.

Temps are completely fine, sits at about 70 before crashing, the highest ive seen it go up to was 80 once so i doubt that could be a problem.

Only crashes when loading 3D objects, I can watch 4k videos all day just fine for as long as i want without it crashing.

 

But yet the problem is still there, I don't know what else it could be other than a bust gpu but there is no BSOD or any artifacting it just crashes 20 seconds after going underload or sometimes even straight away in some titles such as bo4.

 

My system: 

Mobo- Asus prime x470-pro

CPU- Ryzen 5 3600x

Ram- 2x Trident z 8Gb 3200 Mhz

GPU- GTX 980ti (I think is at fault)

The two PSU's ive tried- CX450M (I know its not powerfull thats why it's the first thing I thought could be causing the problem), and 80+ EVGA 600w

 

P.S i noticed that the card had some missing backplate screws when it arrived, I ended up taking it apart after everything i tried, and there was i'd say about a gallon of thermal paste around the die which i wiped off and reapplied new paste.

Link to post
Share on other sites
Posted · Original PosterOP
Just now, homeap5 said:

Try to underclock your card using MSI Afterburner.

 

already tried that

Link to post
Share on other sites
27 minutes ago, Mikudze said:

tried putting the card into a completely different system to rule out mobo, cpu, or ram problems.

So, did the card work perfectly in the other machine? What are the specs of that one? (specifically CPU and PSU)


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
1 minute ago, DJ46 said:

So, did the card work perfectly in the other machine? What are the specs of that one? (specifically CPU and PSU)

no it didn't work in the other pc, the other pc's specs where, fx8320 on a 550w psu

 

Link to post
Share on other sites
4 minutes ago, Mikudze said:

no it didn't work in the other pc, the other pc's specs where, fx8320 on a 550w psu

 

I don't know anything about CeX but from looking at their website they seem to sell used stuff or something like that.

I'd guess someone sold them that GPU already broken, they checked it by booting to desktop and thought it was fine.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
1 minute ago, DJ46 said:

I don't know anything about CeX but from looking at their website they seem to sell used stuff or something like that.

I'd guess someone sold them that GPU already broken, they checked it by booting to desktop and thought it was fine.

I figured that but I still want to know what the problem could be, because from searching about forum's and such nothing comes up with the situation I'm in, like they are very similear but they have all been solved by changing psu or putting into a different pc, which I have tried and dosn't help, which has just made me even more curious to what the problem in the gpu could actually be. cause the temps are fine, it dosnt hit power limit or temp limit just crashes the system, the fans don't stop spinning or anything, just reboots, its odd

Link to post
Share on other sites
2 minutes ago, Mikudze said:

I figured that but I still want to know what the problem could be, because from searching about forum's and such nothing comes up with the situation I'm in, like they are very similear but they have all been solved by changing psu or putting into a different pc, which I have tried and dosn't help, which has just made me even more curious to what the problem in the gpu could actually be. cause the temps are fine, it dosnt hit power limit or temp limit just crashes the system, the fans don't stop spinning or anything, just reboots, its odd

The fact that it happens only in 3D workloads is a pretty typical issue, but they way it is crashing is quite odd indeed.

I had plenty of very interesting issues with my GPU and right now it's somehow working 100% fine after the roughly 5th repaste.

Spoiler

I gave up on it at first but then I noticed it wasn't crashing anymore for some reason, I thought the chip was ruined for sure. So you could always try a repaste or something if you need to get it working (I did it the same way everytime, but slowly with less and less paste, which was pretty scary on a bare die chip).

 

It would just crash completely randomly, but it was always stable at high usage. I tried upping the voltage of the lower power states and it crashed faster, lowering the voltage somehow helped. It would just keep freezing on the desktop or while showing a browser, showed artifacts in just one game... Genuinely the most infuriating and weird issue I've had. I ended up finding a reddit thread complaining about the stock thermal paste application on Vega cards where the OP described the exact same issues I was having. As I said, repasting the GPU did nothing so I gave up. For some reason, I did one more and now it's fine.

You could also try lowering the power limit even more and seeing if it ends up being completely stable at like 50% or something.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
13 minutes ago, DJ46 said:

The fact that it happens only in 3D workloads is a pretty typical issue, but they way it is crashing is quite odd indeed.

I had plenty of very interesting issues with my GPU and right now it's somehow working 100% fine after the roughly 5th repaste.

  Reveal hidden contents

I gave up on it at first but then I noticed it wasn't crashing anymore for some reason, I thought the chip was ruined for sure. So you could always try a repaste or something if you need to get it working (I did it the same way everytime, but slowly with less and less paste, which was pretty scary on a bare die chip).

 

It would just crash completely randomly, but it was always stable at high usage. I tried upping the voltage of the lower power states and it crashed faster, lowering the voltage somehow helped. It would just keep freezing on the desktop or while showing a browser, showed artifacts in just one game... Genuinely the most infuriating and weird issue I've had. I ended up finding a reddit thread complaining about the stock thermal paste application on Vega cards where the OP described the exact same issues I was having. As I said, repasting the GPU did nothing so I gave up. For some reason, I did one more and now it's fine.

You could also try lowering the power limit even more and seeing if it ends up being completely stable at like 50% or something.

lowest i can go is 60% on power limit -90 on core and -201 on mem, on afterburner, it works fine for a small while but then eventually crashes, the one thing i found that did help was setting temp limit to around 65-70 but once it reaches that temp the card becomes useless and frames drop drasticly, i may aswell just be using the igpu on a ryzen 3 3300 g at that point to be fair.

Link to post
Share on other sites
9 minutes ago, Mikudze said:

lowest i can go is 60% on power limit -90 on core and -201 on mem, on afterburner, it works fine for a small while but then eventually crashes, the one thing i found that did help was setting temp limit to around 65-70 but once it reaches that temp the card becomes useless and frames drop drasticly, i may aswell just be using the igpu on a ryzen 3 3300 g at that point to be fair.

I wasn't sure how low you could set the power limit. Bummer.

 

If you need to get the card to work, you could try repasting it. I thought I knew how to do it but the previous comment of mine says otherwise...

The mounting pressure of my GPU's cooler is probably weaker than I'm used to from CPUs, as the issue ended up being (probably) too much paste.

When it comes to GPUs, I am an advocate of spreading the paste. You have to be careful and check the coverage (I do it by checking if any part of it reflects light like naked silicon, preferably use a magnifying glass to look for small gaps). I just stick my finger in a plastic bag and spread the paste with my finger.

I even tried the method EK recommends (looks like a plus AND a cross combined) and only accomplished making a mess. The last time I put on barely enough paste and worked on spreading it for a long time, making sure every part of the core was covered but with as thin a layer as possible. It's hard to describe, but hey, you can just try a few times. Not like there's much to lose if the GPU is useless anyway.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
1 minute ago, DJ46 said:

I wasn't sure how low you could set the power limit. Bummer.

 

If you need to get the card to work, you could try repasting it. I thought I knew how to do it but the previous comment of mine says otherwise...

The mounting pressure of my GPU's cooler is probably weaker than I'm used to from CPUs, as the issue ended up being (probably) too much paste.

When it comes to GPUs, I am an advocate of spreading the paste. You have to be careful and check the coverage (I do it by checking if any part of it reflects light like naked silicon, preferably use a magnifying glass). I just stick my finger in a plastic bag and spread the paste with my finger.

I even tried the method EK recommends (looks like a plus AND a cross combined) and only accomplished making a mess. The last time I put on barely enough paste and worked on spreading it for a long time, making sure every part of the core was covered but with as thin a layer as possible. It's hard to describe, but hey, you can just try a few times. Not like there's much to lose if the GPU is useless anyway.

yeah i guess so, i just dont see how it would help since nothing is wrong with the temps for the workload the gpu is actually pushing before crashing, i just changed temp limit to keep voltages and clock rates down i didnt relise by how much it actually puts it down by, ends up running like 400 Mhz on the core when hitting temp limit, but yeah I guess I could give it another go on repasting not like it would break the card more since its pretty much useless right now....last night i put it on like a cpu so like a grain of rice in the middle and then small tiny blobs around it becuase it didnt look like it would be enough, im guessing that wouldnt be right for gpu dye from what youre saying but I'll give it a go none the less

Link to post
Share on other sites
23 minutes ago, Mikudze said:

yeah i guess so, i just dont see how it would help since nothing is wrong with the temps for the workload the gpu is actually pushing before crashing, i just changed temp limit to keep voltages and clock rates down i didnt relise by how much it actually puts it down by, ends up running like 400 Mhz on the core when hitting temp limit, but yeah I guess I could give it another go on repasting not like it would break the card more since its pretty much useless right now....last night i put it on like a cpu so like a grain of rice in the middle and then small tiny blobs around it becuase it didnt look like it would be enough, im guessing that wouldnt be right for gpu dye from what youre saying but I'll give it a go none the less

You have to remember that the temperature sensor is somewhere on the die, specifically on the edge. New AMD cards have multiple sensors, one on the edge and then a bunch spread around the core, the hottest of which gets reported as the "hotspot" temperature.

When you think about the size of a GPU die, that sensor is useless if you have an air gap the size of a dust particle somewhere, with a bunch of transistors under it getting basically no cooling.

After I realised that my issues probably came from something like that, I was terrified. The GPU was showing artifacts and I knew I tried upping the voltage first and I had been playing VR games in the mean time. I was completely sure I ruined it. There was a tiny dent that looked like a spec of dust in the die, and I thought it was a burned out spot of silicon or something. I guess it's just a scratch or something, since the card works fine now.

 

That's why you have to take great care when spreading the paste and exactly the reason EK recommends that giant amount of paste - their mounting system can probably create enough pressure to actually push the extra out from the gap, like a CPU cooler would.

My issue could've also been caused by the fact that I was using the Arctic MX-2 thermal paste. It's a pretty thick one. MX-4 is easier to spread as far as I know, same with the Thermal Grizzly stuff I've seen.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
2 minutes ago, DJ46 said:

You have to remember that the temperature sensor is somewhere on the die, specifically on the edge. New AMD cards have multiple sensors, one on the edge and then a bunch spread around the core, the hottest of which gets reported as the "hotspot" temperature.

When you think about the size of a GPU die, that sensor is useless if you have an air gap the size of a dust particle somewhere, with a bunch of transistors under it getting basically no cooling.

After I realised that my issues probably came from something like that, I was terrified. The GPU was showing artifacts and I knew I tried upping the voltage first and I had been playing VR games in the mean time. I was completely sure I ruined it. There was a tiny dent that looked like a spec of dust in the die, and I thought it was a burned out spot of silicon or something. I guess it's just a scratch or something, since the card works fine now.

 

That's why you have to take great care when spreading the paste and exactly the reason EK recommends that giant amount of paste - their mounting system can probably create enough pressure to actually push the extra out from the gap, like a CPU cooler would.

My issue could also be caused by the fact that I was using the Arctic MX-2 thermal paste. It's a pretty thick one. MX-4 is easier to spread as far as I know, same with the Thermal Grizzly stuff I've seen.

well im using some coolermaster "mastergel" since its the only i have on hand, ive made sure that the entire die is covered before putting the heat sync back, I'm putting all the screws back in now, I hope this works if not then ahh well 

Link to post
Share on other sites
14 minutes ago, Mikudze said:

well im using some coolermaster "mastergel" since its the only i have on hand, ive made sure that the entire die is covered before putting the heat sync back, I'm putting all the screws back in now, I hope this works if not then ahh well 

You can always put it on the shelf, get some thermal paste later and try it a couple times. I'm not saying it will 100% save the card, but for me it worked on like the 5th attempt, as I said before.

I've had a reason to take the card apart since but honestly I'm too scared to do it. I don't want to anger the demon in it that seems to be chilling for now :D


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
1 minute ago, DJ46 said:

You can always put it on the shelf, get some thermal paste later and try it a couple times. I'm not saying it will 100% save the card, but for me it worked on like the 5th attempt, as I said before.

I've had a reason to take the card apart since but honestly I'm too scared to do it. I don't want to anger the demon in it that seems to be chilling for now :D

yeah it still crashed after the repaste, my suspicion now is that whoever had the card before must have watercooled it because the shroud and the heat sync are in very good condition no dust or scratches nothing, so probably was watercooled overclocked and pushed to the limit and blown something, but the problem with that is, that there is no artifacting or nothing just goes yeet and restarts.

Link to post
Share on other sites
3 hours ago, Mikudze said:

my suspicion now is that whoever had the card before must have watercooled it because the shroud and the heat sync are in very good condition no dust or scratches nothing, so probably was watercooled overclocked and pushed to the limit and blown something

If it was watercooled it also could've gotten wet. There might've been a leak from the CPU block onto the back of the PCB and only specific components/traces got damaged. So it works fine until those get utilised.

You could thoroughly inspect the PCB or even clean it with a soft brush and some isopropyl alcohol, let it dry completely and try the repaste a few times.

Since the alternative is throwing it away, you can just view it as a project/experiment and work on it whenever you feel like it.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
3 hours ago, DJ46 said:

If it was watercooled it also could've gotten wet. There might've been a leak from the CPU block onto the back of the PCB and only specific components/traces got damaged. So it works fine until those get utilised.

You could thoroughly inspect the PCB or even clean it with a soft brush and some isopropyl alcohol, let it dry completely and try the repaste a few times.

Since the alternative is throwing it away, you can just view it as a project/experiment and work on it whenever you feel like it.

well the backplate for some reason has tape on the underside, like some weird certain area's that arnt covering anything and its not like the backplate is touching the resistors either, i'll attach a photo, but thats really it, the pcb seems fine other than that there dosnt seem to be any damage to it from what i can tell except for some light discolouration around the die and memory chips but that's probably just from heat and nothing to worry about i'll attach a photo anyway, maybe its one of the memory chips thats gone bad because even if its not at temp limit or a power limit once one of those memory chips gets asked for information it crashes, could I be right in thinking that.

whatever if it is bust CeX offers 24 month warrenty on everything so it's whatever just dont want to go throught the hassle of sending something waiting for them to recieve, then either get a refund or a replacement.

IMG_1018.jpg

IMG_1008.jpg

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

well the backplate for some reason has tape on the underside, like some weird certain area's that arnt covering anything and its not like the backplate is touching the resistors either, i'll attach a photo, but thats really it, the pcb seems fine other than that there dosnt seem to be any damage to it from what i can tell except for some light discolouration around the die and memory chips but that's probably just from heat and nothing to worry about i'll attach a photo anyway, maybe its one of the memory chips thats gone bad because even if its not at temp limit or a power limit once one of those memory chips gets asked for information it crashes, could I be right in thinking that.

whatever if it is bust CeX offers 24 month warrenty on everything so it's whatever just dont want to go throught the hassle of sending something waiting for them to recieve, then either get a refund or a replacement.

The tape does seem pretty odd.

 

I assumed there was no warranty. Since there is, I'd just send it back. Hopefully they don't mind you taking the card apart.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
Link to post
Share on other sites
Posted · Original PosterOP
33 minutes ago, DJ46 said:

The tape does seem pretty odd.

 

I assumed there was no warranty. Since there is, I'd just send it back. Hopefully they don't mind you taking the card apart.

The card came with missing screws both on the pcb and backplate so im pretty sure they dont care haha

Link to post
Share on other sites
Posted · Original PosterOP

For whoever comes across this thread in the future with a similear problem, I've had the card sitting on a shelf for around 24hours since I last touched it, which was when I done the second repaste......and I decided ahh whatever may aswell try the card again, I ddu'd and dowloaded 388.71 driver from nvidia instead of the latest couple of drivers for a change and the card has completed a full benchmark of FurMark, it was hitting 101% tdp at max and around 75 degree's, so my only guess right now would be the driver I dont see how leaving the card would fix anything, I'll run a few more runs of FurMark and do some more testing in general, but so far the repaste and 388.71 drivers seemed to have fixed the card (for now), I'll keep this thread updated as I go along testing the card.

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

For whoever comes across this thread in the future with a similear problem, I've had the card sitting on a shelf for around 24hours since I last touched it, which was when I done the second repaste......and I decided ahh whatever may aswell try the card again, I ddu'd and dowloaded 388.71 driver from nvidia instead of the latest couple of drivers for a change and the card has completed a full benchmark of FurMark, it was hitting 101% tdp at max and around 75 degree's, so my only guess right now would be the driver I dont see how leaving the card would fix anything, I'll run a few more runs of FurMark and do some more testing in general, but so far the repaste and 388.71 drivers seemed to have fixed the card (for now), I'll keep this thread updated as I go along testing the card.

called it too soon haha it crashed again 

 

Link to post
Share on other sites
Posted · Original PosterOP

I'm gonna run a cuda memcheck, to see narrow down the problems, I guess something in the board when it gets hot is causing the system to crash, because when i leave the card sitting on a shelf to cool down to room temp its runs fine for around 5-10 minutes then crashes, and then after that I can't run the card for even 30 seconds underload, so i'm guessing it is a heat related problem, even tho its showing that temps are completely fine something is getting up to operating temps and crashing, because when its still at room temp it runs fine

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

called it too soon haha it crashed again 

 

I know that feeling too well :D

1 hour ago, Mikudze said:

runs fine for around 5-10 minutes then crashes, and then after that I can't run the card for even 30 seconds underload, so i'm guessing it is a heat related problem, even tho its showing that temps are completely fine something is getting up to operating temps and crashing, because when its still at room temp it runs fine

I'm not sure if uneven thermal paste would cause that, a part of the die would probably heat up a lot faster than that.

Without knowing the state of the thermal pads on the VRAM and VRM it's hard to tell. It could also be a different component that isn't cooled directly, unstable due to some kind off damage.


PC: CPU: Intel i7-4790 MB: Gigabyte B85N RAM: Adata 4GB + Kingston 8GB SSD: Samsung 850 EVO 500GB GPU: XFX GTR RX 480 8GB Case: Lian Li PC-A05NB PSU: Corsair RM1000i KB: Drevo Gramr (modded with real Cherry MX Blues) Mouse: Logitech G502 Proteus Core Mousepad: Steelseries QcK Headphones: Sennheiser HD555
Phone: Samsung Galaxy S7
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


×