Question Zotac RTX 2060 stops giving display output when pushed hard ?

[Moru Zerinho6]

This card was working fine to me when I got it, played Satisfactory with Lumen enabled and Fort nite with RT. The problem I'm having now is that, if I push the card a little bit hard, it will stop giving video to my monitors, I'm able to hear the game for a few more seconds but then it stops, I suppose it has crashed (same with Discord, but that's supposed to happen gives it uses Hardware acceleration).

Another issue is that, randomly, the fan of the card stops for half a second and then starts to spin again, this does not happen if I force a high (>75%) fan speed in Afterbunner. In resume, I can easily make that happen again by running Fort nite in dx11 with everything maxed or Fort nite in dx12 with Lumen epic + Hardware RT enabled.

One time that randomly happened while playing Project Zomboid, which doesn't make sense to me as that game barely uses GPU at all.

I have tried:

- Used DDU to uninstall the driver and install again. (This system had a RX 560 before, but I also used DDU to remove amd driver before installing this one)
- Used MSI's Afterburner to adjust fan speed and curve. (This only solves the problem if I force the fans to spin at >75%, but I don't think I should do that)
- Reset motherboard bios to factory configs and then re-enabled 4G Decoding, CPU Virtualization, Disabled PBO

One important thing to note is that I have a friend with a RTX 3080 who has also installed the new NVIDIA APP and is also having this problem, I was not having such problem before having that app but it also hasn't even announced at the time. I gave this card to a parent of mine to test, he also installed the new NVIDIA APP as soon as he got the card and had the exact same problem when running BF42/BFV or running GPU benchmark tests.

My specs:

- AMD Ryzen 5 4500

- 64GB DDR 4 2400MHz (4 sticks)
- GIGABYTE A520M DS3H
- PSU GAMEMAX gp400a (400w 80 Plus Bronze)
- 5 Storage Drives (2 NVMEs, 2 HDDs, 1 SSD)
- RTX 2060 Zotac
- NVIDIA Driver 552.22 (Installed with factory reset options enabled)
- 2 Monitors (1080p 100Hz HDR Enabled [Using RTX HDR], 1080p 60Hz)
- Windows 11 Pro 23H2

 

[rgd1101]

test with a different psu?

 

[Moru Zerinho6]

test with a different psu?

Already did

> I gave this card to a parent of mine to test, he also installed the new NVIDIA APP as soon as he got the card and had the exact same problem when running BF42/BFV or running GPU benchmark tests.

 

[beyondlogic]

This card was working fine to me when I got it, played Satisfactory with Lumen enabled and Fort nite with RT. The problem I'm having now is that, if I push the card a little bit hard, it will stop giving video to my monitors, I'm able to hear the game for a few more seconds but then it stops, I suppose it has crashed (same with Discord, but that's supposed to happen gives it uses Hardware acceleration).

Another issue is that, randomly, the fan of the card stops for half a second and then starts to spin again, this does not happen if I force a high (>75%) fan speed in Afterbunner. In resume, I can easily make that happen again by running Fort nite in dx11 with everything maxed or Fort nite in dx12 with Lumen epic + Hardware RT enabled.

One time that randomly happened while playing Project Zomboid, which doesn't make sense to me as that game barely uses GPU at all.

I have tried:

- Used DDU to uninstall the driver and install again. (This system had a RX 560 before, but I also used DDU to remove amd driver before installing this one)
- Used MSI's Afterbunner to adjust fan speed and curve. (This only solves the problem if I force the fans to spin at >75%, but I don't think I should do that)
- Reset motherboard bios to factory configs and then re-enabled 4G Decoding, CPU Virtualization, Disabled PBO

One important thing to note is that I have a friend with a RTX 3080 who has also installed the new NVIDIA APP and is also having this problem, I was not having such problem before having that app but it also hasn't even announced at the time. I gave this card to a parent of mine to test, he also installed the new NVIDIA APP as soon as he got the card and had the exact same problem when running BF42/BFV or running GPU benchmark tests.

My specs:

- AMD Ryzen 5 4500

- 64GB DDR 4 2400MHz (4 sticks)
- GIGABYTE A520M DS3H
- PSU GAMEMAX gp400a (400w 80 Plus Bronze)
- 5 Storage Drives (2 NVMEs, 2 HDDs, 1 SSD)
- RTX 2060 Zotac
- NVIDIA Driver 552.22 (Installed with factory reset options enabled)
- 2 Monitors (1080p 100Hz HDR Enabled [Using RTX HDR], 1080p 60Hz)
- Windows 11 Pro 23H2

your about 100w under the requirement to run that card
that aside

run hw monitor in the background see if anything is getting hot
https://www.cpuid.com/softwares/hwmonitor.html

 

[35below0]

your about 100w under the requirement to run that card

Possibly more if the PSU is old. It may not be PSU related though, because it does not explain the other PCs having a similar issue with the same nvidia driver.

At any rate a 550w or more PSU would be more appropriate. Remember that a more efficient PSU both delivers more power and also saves more energy when it's not needed. A 400w Bronze seems puny for that system.


Can you test with a different GPU?

Also, you're not only pushing the GPU, you're pushing the CPU too. The 4500 isn't a powerhouse CPU. It's quite weak for a 64Gb, multimonitor HDR gaming rig.

 

[beyondlogic]

Possibly more if the PSU is old. It may not be PSU related though, because it does not explain the other PCs having a similar issue with the same nvidia driver.

At any rate a 550w or more PSU would be more appropriate. Remember that a more efficient PSU both delivers more power and also saves more energy when it's not needed. A 400w Bronze seems puny for that system.


Can you test with a different GPU?

Also, you're not only pushing the GPU, you're pushing the CPU too. The 4500 isn't a powerhouse CPU. It's quite weak for a 64Gb, multimonitor HDR gaming rig.

we dont know what the pcs power supply was either which doesnt help.

other thing to try in nvidia control panel under manage settings
option named «low latency mode» is this is put to ultra it force G-sync so either turn it off or on high.

 

[Moru Zerinho6]

Possibly more if the PSU is old. It may not be PSU related though, because it does not explain the other PCs having a similar issue with the same nvidia driver.

At any rate a 550w or more PSU would be more appropriate. Remember that a more efficient PSU both delivers more power and also saves more energy when it's not needed. A 400w Bronze seems puny for that system.


Can you test with a different GPU?

Also, you're not only pushing the GPU, you're pushing the CPU too. The 4500 isn't a powerhouse CPU. It's quite weak for a 64Gb, multimonitor HDR gaming rig.

Yes I have tested with a different GPU, with a RX 560, it does not have this issue, the 4500 isn't the problem as I've even undervolted it to the point where it doesn't even use more than 20w to see if that would fix the problem, but it didn't. This CPU is indeed weak if I want to play BF42 or things with heavy RT, but I play mostly 6 year old games so I'm rarely CPU limited.

The PSU can't be the problem because my parent PSU is a cooler master 600w 80 plus bronze and he had the exact same problem.

I've also ran HWMonitor to look at the temperature of things, but I didn't see anything out of the ordinary.

 

[MJS WARLORD]

You mentioned satisfactory , i use this every day , i have a gtx 1080 card and on ultra settings the cards fans kick in at around 70 degrees , have you looked to see if the cards fans are working.

during a game use the windows key so you can open task manager whilst a game is running , what does it say the temp of the card is.

 

[rgd1101]

what the model of this cooler master 600w. can't keep testing with low quality psu

 

[Moru Zerinho6]

what the model of this cooler master 600w. can't keep testing with low quality psu

Can't really check the model as I won't tell my parent to mess with his whole setup just to check which model it is, but why are we still insistent that the problem is the PSU?

This system already ran fine with a RX 560 and a GTX 1060, even when both puxed to the max, my parent also ran both of those cards and he also has a 5700x. I've already undervolted my CPU to reduce power draw and the issue happens even on non-GPU demanding tasks such as Project Zomboid, so no the problem is not the PSU.

You mentioned satisfactory , i use this every day , i have a gtx 1080 card and on ultra settings the cards fans kick in at around 70 degrees , have you looked to see if the cards fans are working.

during a game use the windows key so you can open task manager whilst a game is running , what does it say the temp of the card is.

Yes, the fans are working, I've watched the temps of the card and the max I've seen it reach on those heavy moments is around 73ºC, I'm also able to force the fans to spin more with msi afterbunner, as commented before. But I has presented the same problem even when it's mostly idle, such as when running a emulator.

 

[rgd1101]

Can't really check the model as I won't tell my parent to mess with his whole setup just to check which model it is, but why are we still insistent that the problem is the PSU?

could be your whole system

 

[Moru Zerinho6]

could be your whole system

How could it be my whole system when this happens only when using the RTX 2060?

 

[rgd1101]

then the solution is not use the RTX 2060

 

[Cellie]

Is your OS set to auto-update drivers? You should try to rollback GPU the driver to the version released around the time you got the card, disable auto-updates if possible, and remove that Nvidia App for good measure.

 

[MJS WARLORD]

i found this on a web search for the power supply you should be using

Nvidia recommends a power supply of at least 550 watts if you’re using a GeForce RTX 2060 Super graphics card. Ideally, you should aim a little higher than this minimum – 650 watts or more is enough to cope with power spikes and give you room to upgrade.

So i think it fair to say the size of the power supply your using is the problem.
If you want to find out about everything on the pc you dont need to take the side off and look inside it.
Download a free program called belarc advisor and you will get information that you would not be able to find yourself.

NOTE only get it from the oficial site , other sites will give you junk you dont want or need.

https://www.belarc.com/products/belarc-advisor/download-eda4a40b4916

 

[DSzymborski]

Yes I have tested with a different GPU, with a RX 560, it does not have this issue, the 4500 isn't the problem as I've even undervolted it to the point where it doesn't even use more than 20w to see if that would fix the problem, but it didn't. This CPU is indeed weak if I want to play BF42 or things with heavy RT, but I play mostly 6 year old games so I'm rarely CPU limited.

The PSU can't be the problem because my parent PSU is a cooler master 600w 80 plus bronze and he had the exact same problem.

I've also ran HWMonitor to look at the temperature of things, but I didn't see anything out of the ordinary.

Unless this is the Bronze V2 that is *not* 230V only, I wouldn't consider the PSU tested; you can't test a low quality PSU with another low-quality PSU.

What's especially concerning is the possibility that the GPU is failing *because* of the junk PSU you've been using it on. That's how bad your current PSU is; for all intents and purposes, it's not even a *300*W PSU, given that the +12V rail is only specced for 288W. If you've been using your GPU for years on something this low-quality, you may be seeing the consequences, which is a common thing we see around here. Even if your parents have an appropriate budget PSU, it doesn't rule out this extremely plausible scenario.

If you're going to diagnose further, you may need both an appropriate PSU *and* a GPU that's confirmed to be working otherwise. That's a real pain, of course, but cheap PSUs are one of the most expensive components.

 

[Moru Zerinho6]

If you've been using your GPU for years on something this low-quality, you may be seeing the consequences, which is a common thing we see around here. Even if your parents have an appropriate budget PSU, it doesn't rule out this extremely plausible scenario.

I've not been using this 2060 Zotac for years, I've only had it for about 2 months, and at the start it wasn't having such problem.

If you're going to diagnose further, you may need both an appropriate PSU *and* a GPU that's confirmed to be working otherwise. That's a real pain, of course, but cheap PSUs are one of the most expensive components.

I've already written that both a RX 560 and a GTX 1060 are working just fine in both computers, and in both computers the 2060 Zotac is the only one showing issues, how can you realistically say the problem is not something to do with the GPU in that case?

Even if my PSU does not have the recommended wattage, my system is not even close to consume even 300w as I've undervolted my CPU (See my posts above) to the point where it's using 21w max and the GPU shows the same problem even when it's not doing anything which will cause it to use more than 50w.

I have been using the 1060 since its launch at this point and the RX560 for 3 years and none have ever showed the same problem.

 

[rgd1101]

on you first post, you said you fix it by adjust fan speed and curve. (This only solves the problem if I force the fans to spin at >75%, but I don't think I should do that)
what the temp if you don't set it to spin at > 75% ?
what was the default?

 

[Moru Zerinho6]

on you first post, you said you fix it by adjust fan speed and curve. (This only solves the problem if I force the fans to spin at >75%, but I don't think I should do that)
what the temp if you don't set it to spin at > 75% ?
what was the default?

The highest I've seen the temps reach is 73ºC before setting it to spin at >75%, but in all cases (forcing or not forcing it) it stays at around 70ºC, and at that point it's already at the danger of doing the thing.

But more recently it's starting to show the same issue even when it's mostly idle (43ºC to 50ºC), already tried to uninstall the NVIDIA App Beta and rollback the driver but no luck, I just came to the conclusion that the card itself might have some physical damage so I went back to use the 1060, a lot lower performance but at least I don't have the problem anymore.

 

[DSzymborski]

I've already written that both a RX 560 and a GTX 1060 are working just fine in both computers, and in both computers the 2060 Zotac is the only one showing issues, how can you realistically say the problem is not something to do with the GPU in that case?

By knowing how much wattage these use? An RX 560 maxes out at 75W and a 1060 at 120W. The RTX 2060 peaks around 160W if you have the 6 GB one and can get near 190W with the 12 GB one with momentary spikes up to 180W and 210W respectively. You're not isolating a GPU problem by testing cards that use a lot less power and have a lot less aggressive boost behavior.

You have symptoms that are very common to a PSU issue. You've already ruled out temperatures, the other primary cause of your symptoms. You have a junk PSU and tested it with another PSU that might also be junk. So with temperatures ruled out, you have four most likely scenarios:

1. The crashes are due to a junk PSU.
2. The crashes are due to the GPU failing, caused by a junk PSU.
3. The crashes are due to the GPU failing and it was already heading there when you bought it used.
4. The crashes are due to the GPU failing coincidentally.

Since you're apparently unwilling to meaningfully test for any of these scenarios, I'm not sure what else there is to do at this point. You're a bit in the position of someone who goes to the emergency room with chest pain but insists that you *know* that it's not your heart and you forbid anyone from checking your heart.

I hope you have the best of fortune and are able to solve this problem.