[SOLVED] Graphic card stops sending a picture

[Veiler]

Hi,

So my graphic card have been annoying me for last couple of weeks, it just started to stop sending a picture to my 2 monitors (also happens when i use 1). I figured it was the graphic card since i would still be able to talk to people on discord while there was no picture, so it was not frozen.
It started a couple weeks after i upgraded from a GTX 970 4gb to the one i currently have (ZOTAC Geforce RTX 2070 AMP Extreme), it would do it once a week, but i always could just reboot my PC and move on. Then it started slowly happen more often, and also do it on reboots. The state im currently in is that it happens every day, with a around 1-2 hours wait time before i can reboot it without loosing the picture upon reaching desktop.

Things is tried:

• Do DDU multiple time in row, for then install the correct driver
• Do a reinstall windows through windows it self, with the possibility of keeping my "personally files

I have to try switch to a different port on graphic card, but don't think it would work.

UPDATE

Hardware is here
PSU: Coolmaster G650 (Model no. RS-650-AMAA-B1)

CPU: Intel Core i7 7700K 4.20GHz

RAM: 16,0GB Dual-Channel Unknown @ 1066MHz (15-15-15-36)

GPU: ZOTAC Geforce RTX 2070 AMP Extreme

Motherboard: MSI Z270-A PRO

SSD: 465GB Samsung SSD 860 EVO 500GB

 

[Darkbreeze]

What is your EXACT power supply model number?

Knowing the rest of your hardware specs would be helpful too.

 

[Avoton]

While knowing what power supply you have is very important, it is also important to know what temperature your VRAM and gpu core were at during loads and idle. Would you happen to know?

 

[Veiler]

While knowing what power supply you have is very important, it is also important to know what temperature your VRAM and gpu core were at during loads and idle. Would you happen to know?

Can't remember precisely it is at those to state but, i can say for sure that i haven't been over 77C

 

[Veiler]

What is your EXACT power supply model number?

Knowing the rest of your hardware specs would be helpful too.

PSU: Coolmaster G650 (Model no. RS-650-AMAA-B1)
CPU: Intel Core i7 7700K 4.20GHz
RAM: 16,0GB Dual-Channel Unknown @ 1066MHz (15-15-15-36)
GPU: ZOTAC Geforce RTX 2070 AMP Extreme
Motherboard: MSI Z270-A PRO
SSD: 465GB Samsung SSD 860 EVO 500GB

 

[Darkbreeze]

I'd start by replacing the power supply with a GOOD quality model. For two reasons.

Well, three actually, potentially.

One, that is not a particularly good model. It's ok, but it's VERY mediocre.

Two, if it is older than the build itself, which is highly probable, but not guaranteed, since that model is rather older, then it's a good bet that it has already seen it's useful life pass by since the hardware itself is already a few years old and the PSU is likely even older than that.

Three, Cooler master power supplies tend to not be very reliable in most cases. The reliability goes down, sharply, when used with high draw graphics cards, as does the reliability for most less than desirable quality PSUs. Even if that wasn't the problem, I'd probably say to replace it as it's just not on par with the quality of the rest of the build. Certainly there are much worse power supplies out there so don't replace it with a worse quality unit.

If the PSU turns out to not be the problem then I'd start talking to Zotac about an RMA because it's unlikely to be anything other than the power supply or graphics card, although I WOULD definitely make sure your BIOS on the motherboard is up to date before assuming the graphics card is to blame.

 

[Veiler]

I'd start by replacing the power supply with a GOOD quality model. For two reasons.

Well, three actually, potentially.

One, that is not a particularly good model. It's ok, but it's VERY mediocre.

Two, if it is older than the build itself, which is highly probable, but not guaranteed, since that model is rather older, then it's a good bet that it has already seen it's useful life pass by since the hardware itself is already a few years old and the PSU is likely even older than that.

Three, Cooler master power supplies tend to not be very reliable in most cases. The reliability goes down, sharply, when used with high draw graphics cards, as does the reliability for most less than desirable quality PSUs. Even if that wasn't the problem, I'd probably say to replace it as it's just not on par with the quality of the rest of the build. Certainly there are much worse power supplies out there so don't replace it with a worse quality unit.

If the PSU turns out to not be the problem then I'd start talking to Zotac about an RMA because it's unlikely to be anything other than the power supply or graphics card, although I WOULD definitely make sure your BIOS on the motherboard is up to date before assuming the graphics card is to blame.

I have now tried with a Corsair AX860 PSU and that did not help, but now i also tried to only use the onboard GPU, and that works perfectly, at this point it seems to be GPU might be the issue?

Do you have something to add?

 

[Darkbreeze]

If you are 100% certain that the Corsair PSU you used is good, then I'd contact Zotac and get an RMA started.

Just to verify though, you DO have the GPU card installed in the primary x16 slot which is the closest of the long PCI slots to the CPU socket, right?

And you do have the required power connections from the PSU to the graphics card attached?

And you are NOT trying to use cables that came with a different power supply, WITH your installed power supply (Or with the Corsair unit) right? You MUST use the cables that come with a power supply if it is modular or semi-modular unless you can confirm that the cables you are using are from a compatible series from the same brand and that they are compatible with the same pinouts. Pinouts can be different on one, two, or more of the cables from any given power supply.

It might also be a good idea to try different cables if possible. A cable that is broken or has a short inside can cause similar problems. Also, make certain that both ends of the cables are 100% seated. A cable that is inserted 95% of the way might work, and then not work, intermittently.

But probably it's just a bad card although there is always an outside chance this is a motherboard issue.

 

[Veiler]

If you are 100% certain that the Corsair PSU you used is good, then I'd contact Zotac and get an RMA started.

Just to verify though, you DO have the GPU card installed in the primary x16 slot which is the closest of the long PCI slots to the CPU socket, right?

And you do have the required power connections from the PSU to the graphics card attached?

And you are NOT trying to use cables that came with a different power supply, WITH your installed power supply (Or with the Corsair unit) right? You MUST use the cables that come with a power supply if it is modular or semi-modular unless you can confirm that the cables you are using are from a compatible series from the same brand and that they are compatible with the same pinouts. Pinouts can be different on one, two, or more of the cables from any given power supply.

It might also be a good idea to try different cables if possible. A cable that is broken or has a short inside can cause similar problems. Also, make certain that both ends of the cables are 100% seated. A cable that is inserted 95% of the way might work, and then not work, intermittently.

But probably it's just a bad card although there is always an outside chance this is a motherboard issue.

I can confirm, that i have made sure, that my cables are correct. i have not tried with new cables.
But if it is motherboard there has issue won't i feel it something when i use the onboard (except if it is 16x slot)

 

[Darkbreeze]

No, you probably wouldn't. There are really no commonalities between the integrated graphics and the PCI graphics aside from the CPU and PCI bus. Many systems that have a broken PCI or something related going to the PCI x16 slot work perfectly fine with integrated graphics.

You could try using one of the other x16 PCI slots to see if you have the same issue.

You could also download HWinfo (NOT HWmonitor, Open hardware monitor, Speccy, or any other utility) and install it, run "sensors only" with "summary" option disabled and monitor your CPU and GPU thermal sensors to make sure that this isn't a thermal issue.