It is the card. I talked to Microcenter and they told me they have had multiple returns (more than 3 before me) of this card. I find that kind of aggravating. I have never worked retail, so I am not sure if that is common practice. It seems like very bad practice though. I would think at some point you have to stop selling the hardware. I think 3 returns would be enough. Maybe they need 5.
3 returns for the actual gpu you got? It's definitely not common practice but you did say it takes hours for it to BSOD. They probably weren't able to replicate it so determined it fit for resale.
Chances are the built in overclocking is pushing too far over time and/or pushing your own overclock even further if you used one. I've had a card that did that before and the only way to stop it from doing it was to underclock it back to stock settings to stop the dynamic overclocking.
The multiple returns and the fact I got a PAGE_FAULT_IN_NONPAGED_AREA means it is most likely a RAM issue on the card.
Maybe, but if the dynamic overclocking (Boost Clock) is eventually causing corruption then it will corrupt the data that is in the RAM and would return a RAM related error.
If it was a problem with RAM itself it is unlikely it would take hours to create a problem and extremely unlikely you would have been able to stress test it without it throwing tons of errors at you, which you said you have done.
If you still have it, try underclocking it to the standard 2080 clock, as in GPU 1515 Memory 1750, and see if there's still a problem. Some cards have that issue with Boost Clocks, is all I'm saying. I had an Inno3D GTX 770 that did it. The Boost Clock boosted higher than the specified boost clock for that card in certain games (usually newer ones at the time), for some unknown reason.
Only if you're curious about finding out what the problem is of course...
Either way, it ain't what you want from a new card and needs to be returned.
I don't suffer from insanity, I enjoy every minute of it.
It sounds like you have a bad memory sector. And this totally comes from my expert opinion and not simply reading the thread. Ram makes sense. In a synthetic benchmark, it could be accessing the same ram repeatedly. In real world testing it will randomize the memory while constantly adding more into memory.
Still seem to be ignoring the big gorilla in the picture, time to dump Win 7.
Why do you think then that his 1080 worked just fine if Win7 is such a problem?
He has a new card and a new driver. That's where you start troubleshooting not the freaking OS, fffs. Troubleshooting 101.
Really? How well do you think they test Win 7 drivers these days? My guess only superficially. Since Win 7 support is on the wane you would think upgrading the OS would be something that should be considered whether a bad card or not.
Still seem to be ignoring the big gorilla in the picture, time to dump Win 7.
Why do you think then that his 1080 worked just fine if Win7 is such a problem?
He has a new card and a new driver. That's where you start troubleshooting not the freaking OS, fffs. Troubleshooting 101.
Really? How well do you think they test Win 7 drivers these days? My guess only superficially. Since it seems to be a driver issue it would behoove him to upgrade the OS. I have built multiple systems with 2080's in them and never had an issue with Windows 10 drivers.
It's the same driver for 7, 8 and 8.1 so yeah they do make them work just fine. I run Win10 myself and I do think it's far better than 7 or 8 but that's beside the point.
When you make one change to a computer, especially something as simple as going from one nVidia card to another one, and you have problems with it, suggesting an OS re-install or OS upgrade as the first step in troubleshooting is what I'm having problems with here.
It's nonsensical advice even if it's generally good advice long term... long term as in sometime after he fixes this one problem that has nothing to do with WIN7 lol.
"Social media gives legions of idiots the right to speak when they once only spoke at a bar after a glass of wine, without harming the community ... but now they have the same right to speak as a Nobel Prize winner. It's the invasion of the idiots”
― Umberto Eco
“Microtransactions? In a single player role-playing game? Are you nuts?” ― CD PROJEKT RED
Comments
If it was a problem with RAM itself it is unlikely it would take hours to create a problem and extremely unlikely you would have been able to stress test it without it throwing tons of errors at you, which you said you have done.
If you still have it, try underclocking it to the standard 2080 clock, as in GPU 1515 Memory 1750, and see if there's still a problem. Some cards have that issue with Boost Clocks, is all I'm saying. I had an Inno3D GTX 770 that did it. The Boost Clock boosted higher than the specified boost clock for that card in certain games (usually newer ones at the time), for some unknown reason.
Only if you're curious about finding out what the problem is of course...
Either way, it ain't what you want from a new card and needs to be returned.
Ram makes sense. In a synthetic benchmark, it could be accessing the same ram repeatedly. In real world testing it will randomize the memory while constantly adding more into memory.
When you make one change to a computer, especially something as simple as going from one nVidia card to another one, and you have problems with it, suggesting an OS re-install or OS upgrade as the first step in troubleshooting is what I'm having problems with here.
It's nonsensical advice even if it's generally good advice long term... long term as in sometime after he fixes this one problem that has nothing to do with WIN7 lol.
“Microtransactions? In a single player role-playing game? Are you nuts?”
― CD PROJEKT RED