New GPU issues [RTX 2070 Super]

[VariableK]

Hello,

 

I recently got myself a 2070 Super by Gigabyte and ever since I've installed it my PC has been crashing seemingly at random, during gameplay. It seems random to me in which games this happens and after how much time since it is never immediate.

These crashes come in the form of my PC just turning off as if someone pulled the power plug. After that it immediately reboots without any error messages.

Windows Event viewer only tells me that "The Shutdown was unexpected" and nothing more as far as I can tell.

 

The rest of my PC specs are this:

 

ASRock Z67 Pro Motherboard

Intel i7 4770k

16GB DDR3 Ram

A 650W powersupply by Corsair

 

My previous videocard was an AMD R9 290, and if I go back to it everything is fine again.

I have made sure to properly uninstall the drivers before installing the new video card, but since then I have also completely wiped my entire PC, which still didn't fix it.

 

Other than that I have:

 

Unplugged and reseated the videocard 

Replugged every power connector in the PC

Unplugged the reset button on my case (which would cause similar behaviour if pressed)

Run Benchmark tests (mainly by Unigine) to specifically put load on the GPU which strangely enough did not lead to crashes.

 

It is important to note that I got myself a regular 2070 a couple months back (just before the Super cards were announced) and I had the same issues. I just filed that under "must've got a faulty one", returned it and waited for the Super Cards to release and become available at reasonable pricing.

 

At this point I don't know what I can do (without immediately replacing the rest of my hardware) so please do share any advice you may have.

 

Thank you in advance

[AngryBeaver]

23 minutes ago, VariableK said:

Hello,

 

I recently got myself a 2070 Super by Gigabyte and ever since I've installed it my PC has been crashing seemingly at random, during gameplay. It seems random to me in which games this happens and after how much time since it is never immediate.

These crashes come in the form of my PC just turning off as if someone pulled the power plug. After that it immediately reboots without any error messages.

Windows Event viewer only tells me that "The Shutdown was unexpected" and nothing more as far as I can tell.

 

The rest of my PC specs are this:

 

ASRock Z67 Pro Motherboard

Intel i7 4770k

16GB DDR3 Ram

A 650W powersupply by Corsair

 

My previous videocard was an AMD R9 290, and if I go back to it everything is fine again.

I have made sure to properly uninstall the drivers before installing the new video card, but since then I have also completely wiped my entire PC, which still didn't fix it.

 

Other than that I have:

 

Unplugged and reseated the videocard 

Replugged every power connector in the PC

Unplugged the reset button on my case (which would cause similar behaviour if pressed)

Run Benchmark tests (mainly by Unigine) to specifically put load on the GPU which strangely enough did not lead to crashes.

 

It is important to note that I got myself a regular 2070 a couple months back (just before the Super cards were announced) and I had the same issues. I just filed that under "must've got a faulty one", returned it and waited for the Super Cards to release and become available at reasonable pricing.

 

At this point I don't know what I can do (without immediately replacing the rest of my hardware) so please do share any advice you may have.

 

Thank you in advance

Have you checked to see if you have any new Bios Updates for your motherboard? Sometimes (rarely) you can have a component on a newer architecture that just doesn't play nice with older bios's.

[adm0n]

If the bios flashing doesn't help, it is probably the GPU. If your PC just shuts down, without a warning or anything, it is usually a hardware error.

If you would have to give a mean time it takes for this to happen, what would it be and how long have you actually tested it with your old GPU?

[VariableK]

1 hour ago, AngryBeaver said:

Have you checked to see if you have any new Bios Updates for your motherboard? Sometimes (rarely) you can have a component on a newer architecture that just doesn't play nice with older bios's.

Haven't tried that yet, will do.

 

1 hour ago, adm0n said:

If the bios flashing doesn't help, it is probably the GPU. If your PC just shuts down, without a warning or anything, it is usually a hardware error.

If you would have to give a mean time it takes for this to happen, what would it be and how long have you actually tested it with your old GPU?

Generally it happens after around 20-45 minutes of gameplay. With my old AMD GPU this has never happened in over 5 years of regular use. 

[VariableK]

2 hours ago, AngryBeaver said:

Have you checked to see if you have any new Bios Updates for your motherboard? Sometimes (rarely) you can have a component on a newer architecture that just doesn't play nice with older bios's.

Done now, didn't help.

I forgot to mention earlier: I have kept an eye on the temperatures as well of course, where CPU and GPU both sit well within the 60°C range under regular use, so that shouldn't trip anything either.

[AngryBeaver]

57 minutes ago, VariableK said:

Haven't tried that yet, will do.

 

Generally it happens after around 20-45 minutes of gameplay. With my old AMD GPU this has never happened in over 5 years of regular use. 

Out of curiosity is your CPU overclocked? If so I would try reverting that to stock and testing too.

 

I also don't think this is necessarily a problem with the GPU since you had this issue with a previous nvidia card. I would have suggest this was a driver issue, but you claim you did a fresh install. Now if that isn't the case I would recommend you give that a try or at least nuke the old drivers.

 

Now if you did the windows refresh or fresh start or w/e it is called.Then you might not have fully removed all of those drivers. So consider that as well.

 

This could also be a problem with your power supply. It might be defective and is shutting down because of any number of issues. I am guessing your old card probably uses a single PCI-E power connectior and the nvidia cards use 2. It could be a power delivery issues, a problem with the connector or cables causing a short... which is making the machine instantly power off in protect mode.  So My next avenue if the above aren't an issue is to try a new PSU.

[adm0n]

By Nuking the driver he means to use DDU, just so you know.

[VariableK]

Yes, my CPU is overclocked, not all that much though. 

For the reinstall I did a full wipe because that Windows thing which keeps your files just causes issues in my experience.

[AngryBeaver]

1 minute ago, VariableK said:

Yes, my CPU is overclocked, not all that much though. 

For the reinstall I did a full wipe because that Windows thing which keeps your files just causes issues in my experience.

I would remove the OC and go to stock and do some testing. You need to remove as many variable as possible to see where the issue is.

[VariableK]

Sorry, misspoke. Updating the bios reset all setting to the defaults so the latest crash was already without an overclock.

[adm0n]

Pull out your GPU and run it with the integrated graphics card. See if anything happens. You could also try a different PCIe slot ... but if that is the problem, buy a new mainboard

 

[hollyh88]

doesn't that board have pci gen 2.0? ive been reading some stuff about the rtx 2070 /super cards not working correctly with those boards due to the "connection port" not being the right fit of a 2.0. 

try to plug your gpu in a friends pc that has a 3.0 board. 

[adm0n]

11 minutes ago, hollyh88 said:

doesn't that board have pci gen 2.0? ive been reading some stuff about the rtx 2070 /super cards not working correctly with those boards due to the "connection port" not being the right fit of a 2.0. 

try to plug your gpu in a friends pc that has a 3.0 board. 

Can you provide a source for that? I haven't heard of any GPU not working in even gen1 slots.

[VariableK]

According to the Manufacturer Page it has one 3.0 port, I also realise I typo'd my motherboard earlier. Its the Z87 Pro3 and not 67. 
Running onboard grpahics yielded nothing but low frames. 

I can't test it on a different motherboard for now. I was planning to upgrade the rest of my PC anyways but I guess it needs to happen a bit quicker.

 

Anyways, thank you all for your time already, much appreciated.

[hollyh88]

7 minutes ago, adm0n said:

Can you provide a source for that? I haven't heard of any GPU not working in even gen1 slots.

well i can provide you with some forum posts and stuff like that. i have one from toms hardware that i still remember. although it's just me speculating but that is what ive heard. them either working bad or simply not working at all due to 3.0 cards having their eh.. pci  "connector" being a bit different than that of a pc 2.0. 

 

 

[adm0n]

4 minutes ago, VariableK said:

According to the Manufacturer Page it has one 3.0 port, I also realise I typo'd my motherboard earlier. Its the Z87 Pro3 and not 67. 
Running onboard grpahics yielded nothing but low frames. 

I can't test it on a different motherboard for now. I was planning to upgrade the rest of my PC anyways but I guess it needs to happen a bit quicker.

 

Anyways, thank you all for your time already, much appreciated. 

That just shows you, that your GPU is definitely related to your problem. Your PSU might be a problem too (but this is very improbable)

Just RMA it.

[hollyh88]

3 minutes ago, VariableK said:

According to the Manufacturer Page it has one 3.0 port, I also realise I typo'd my motherboard earlier. Its the Z87 Pro3 and not 67. 
Running onboard grpahics yielded nothing but low frames. 

I can't test it on a different motherboard for now. I was planning to upgrade the rest of my PC anyways but I guess it needs to happen a bit quicker.

 

Anyways, thank you all for your time already, much appreciated.

ah okey then yes that's not the issue here. quick question.. what is your monitor res? if it's 1440p do you perhaps have a 1080p monitor you can hook up onto it and try it?

[hollyh88]

3 minutes ago, adm0n said:

That just shows you, that your GPU is definitely related to your problem. Your PSU might be a problem too (but this is very improbable)

Just RMA it.

well a 650 psu should provide more than enough on the 12v rail. since the bare minimum listed on the nvidia site is a 650 watt psu. 

could ofc be a dead card. but he has to try everything before returning it tho  

[adm0n]

1 minute ago, hollyh88 said:

well a 650 psu should provide more than enough on the 12v rail. since the bare minimum listed on the nvidia site is a 650 watt psu. 

The PSU could still be faulty

[hollyh88]

6 minutes ago, adm0n said:

The PSU could still be faulty

true but then his old gpu wouldnt work either right ? 

[VariableK]

I have 2 1080p monitors.

It's strange because according to the respective manufacturers my old card actually uses 60W more than the new one.

 

Edit: 1080p 60hz, both

[adm0n]

I said it was very unlikely : D

 

[hollyh88]

16 hours ago, VariableK said:

I have 2 1080p monitors.

It's strange because according to the respective manufacturers my old card actually uses 60W more than the new one.

 

Edit: 1080p 60hz, both

hdmi cable connected to them? 

it does sound more and more like a faulty card. i would suggest trying it in one of a friend to see if it works then or doesn't.

[VariableK]

It's one with HDMI one Display Port. I suppose I could try and unplug one of them and see if it happens