Question Games crashing after GPU reaches 1965Mhz

[Leplivo]

Very recently I bought a 3060 Ti, MSI Ventus 3X to be exact and just now I started to have problems, every game I tested in they crash when the GPU reaches 1965Mhz. In War Thunder because they have made to show what exactly the error is, it says video driver hung, so I am assuming the video driver crashes and then the game crashes. When I installed the GPU first thing I did was clean the drivers with DDU and after it started crashing I cleaned the drivers again and even installed older drivers 4 months back old.
My specs are:
Gigabyte X570 Aorus Elite
Ryzen 9 3900X
MSI 3060 Ti VENTUS 3X
16GB DDR4 3200Mhz
EVGA 650W
256GB SSD
2TB 7200rpm HDD
500GB 7200rpm HDD
Update: The card has been sent to the service shop, it will probably be replaced.

 

[Ralston18]

Look in Microsoft's Reliability History and Event Viewer for any error codes, warnings, or even informational events that may be being captured just before or at the time of the crashes.

Check Update History for any failed or problem updates. Or any update that may have been installed just prior to the start of the crashes.

Manually download the GPU drivers directly from the manufacturer's website. Reinstall and reconfigure. No third party tools, uninstallers, and installers.

And consider that 1965 MHz may not be a viable option for the build. And, even if so, may not really provide any meaningful increase in performance.

What was the last stable configuration?

 

[renz496]

Is it factory overclock card?

 

[Leplivo]

Is it factory overclock card?

Yes, the card is factory overclocked to 1965Mhz, I underclocked it by -100MHz and now it has 0 issues. The card is MSI 3060 Ti Ventus 3X OC LHR.

 

[Leplivo]

Look in Microsoft's Reliability History and Event Viewer for any error codes, warnings, or even informational events that may be being captured just before or at the time of the crashes.

Check Update History for any failed or problem updates. Or any update that may have been installed just prior to the start of the crashes.

Manually download the GPU drivers directly from the manufacturer's website. Reinstall and reconfigure. No third party tools, uninstallers, and installers.

And consider that 1965 MHz may not be a viable option for the build. And, even if so, may not really provide any meaningful increase in performance.

What was the last stable configuration?

The card is factory overclocked to 1965Mhz, I underclocked it by -100MHz and now it has 0 issues. The card is MSI 3060 Ti Ventus 3X OC LHR. Tomorrow I will be sending it to the service, so they can test it. It just doesn't make any sense how someone can factory OC a card to a point that it can't handle the OC and then ship it that way.

 

[rayzeeu]

Have you tried to undervolt the card? And did you monitor thermals before the crashes?

 

[Leplivo]

Have you tried to undervolt the card? And did you monitor thermals before the crashes?

Thermals are fine I just underclocked it and it worked, I just left at the service. They can see what is wrong with it and if it's a factory issue then I will replace it.

 

[Leplivo]

I don't know how to close this thread or if I can close it, so I am going to leave it like this. Maybe I will update it when I get my card back.

 

[Ralston18]

Thread does not need to be closed.

There may be nothing wrong with the card. Overclocking, in general, is simply a way of (hopefully) pushing the limits of any given system and/or its' components. Objective being able to noticably or quantifiably improve performance. Which might not happen anyway.

When some limit is hit, the affected components can roll back to some supported configuration. Could be, for example, that the PSU is the limiting factor.

1965 MHz being "too much" and 1865 MHz being viable.

Which could change + or - later if there are other system changes made. Intentionally or otherwise.

 

[kerberos_20]

Thread does not need to be closed.

There may be nothing wrong with the card. Overclocking, in general, is simply a way of (hopefully) pushing the limits of any given system and/or its' components. Objective being able to noticably or quantifiably improve performance. Which might not happen anyway.

When some limit is hit, the affected components can roll back to some supported configuration. Could be, for example, that the PSU is the limiting factor.

1965 MHz being "too much" and 1865 MHz being viable.

Which could change + or - later if there are other system changes made. Intentionally or otherwise.

that could be tested if he adds voltage, if it improves stability, then its factory binning and QC fault
if it doesnt change anything, then on the other hand undervolting and or underclocking may improve PSU stability

 

[renz496]

Yes, the card is factory overclocked to 1965Mhz, I underclocked it by -100MHz and now it has 0 issues. The card is MSI 3060 Ti Ventus 3X OC LHR.

This is problem with overclocked gpu. Sometimes not even factory overclocked card can have guaranteed stability. Though since it is new card maybe you can try to do RMA.