dexters
stalinx20
dexters
3 week old build, ran through the benches (3D mark, Unigine) fine, but is crashing on Civ 6 after about 30 minutes of play. Black screen + fan at 100%
At first I though it was a CPU issue, but after the 2nd crash, it was clear it was the GPU as I could hear the music and my menu selections still going through.
Is RMA my only option or should i pull back my OC?
Current offset is +105 Memory +300
Try debug mode and see if you can still run the game. If it runs fine within debug mode, then it's not a overheat issue, but an unstable overclock.
Just a quick follow up from my earlier post. I have a window for RMA with the retailer, which is good news, but I wanted to give the card another go before I pull the trigger with the retailer (who has been exceedingly nice about it)
I haven't used Debug mode yet , but Haven't had any black screen + fan at 100% since I changed memory offset from 300 to 100 & rechecked the power pins (one of the 8 pin connectors was a bit loose) and there was an audible clicking when I pressed down.
Played Civ6 for a good 5 hours (2x 2.5hour sessions) w/ Xsplit running and recording for extra load
Ran unigine valley for about 30 mins (though it never crashed with this before)
I haven't had an EVGA card since the GTX260 so my question is, if it's stable now (obviously will look at it some more over the weekend) is my GTX 1080 FTW card likely not part of the defective batch?
Put differently, is the black screen + fan 100% a 'feature' of EVGA cards when it is having issues and not specific to this one hardware issue? (ie: it could happen also due to bad overclocks, power connectors being a bit loose etc)
Thanks.
Don't quote me on this but from reading around it seems that
some EVGA GTX 1080 FTW cards have an issue which means that the VRMs and VRAM are not being cooled properly and so are getting very hot. This is what is apparently causing the black screen and the fans to ramp up to 100% (a preventative measure to stop the card from being damaged or, worst, catching fire). A few websites have done testing (these are referred to in this thread, including links) and revealed that those areas are hotspots with alarmingly high temperatures. One I looked at measured the VRAM at 94.7 C and the Micron GDDR5X memory is only rated to 95 C! Whether this is just a production line fault on a number of cards or a design flaw of all the EVGA GTX 1080 FTW cards, I obviously cannot say as I can only comment on my experience. However, the fact remains that there are a number of GTX 1080 FTW cards out there with this issue.
For me, the black screen issue pretty much started a week or so after switching from a custom fan profile in MSI Afterburner to using the default one, which spins the fans only once the GPU core reaches 65 C. I did initially have a couple of black screens in the first week of testing the card but that was likely while I was using the default fan profile. With the custom profile I was able to use my card for 6 weeks or so without the black screen. I switched the default one a few weeks back due to see how effective it was then a week or so later I started getting random black screens in a number of intensive games (Mafia 3, Forza Horizon 3, Gears of War 4, etc.).
When I switched back to using the custom fan profile while I waited for the replacement card to arrive, I found that I was able to play games fine and without seeing the black screen issue. This confirms to me that the issue is almost certainly related to some components of the card overheating due to inadequate cooling, compounded by the fact the fans only spin when required (above 65 C on the core) by default. Since temps in my game were always under 77 C even in the heat of the British summer and usually between 65 and 75 C at load then it clearly wasn't an issue with the GPU itself. Some of the black screen issues occurred in games while watching pre-rendered cutscenes while the GPU was underclocked to low-power 2D speeds (in Mafia 3 for example)!
I saw the Reddit thread link earlier and found it interesting reading, although I realise that is a mix of opinions and experiences rather than 100% facts. Why EVGA continue to remain silent on this issue by refusing to issue an official statement is a mystery. I would have at least expected some kind of sticky on here so that people who encounter the issue will have a general idea of what it might be so they don't, say, assume incorrectly as I did that this is a driver bug or a result of an unstable overclock. This thread does help but having to read through 45+ pages to get the facts is not really ideal. The longer they remain silent then the more it appears that they are trying to sweep the issue under the carpet so to speak. Maybe they think that by issuing an official statement it would damage their sales/reputation or result in a high number of returns?
By the way, an unstable overclock on the GPU would most likely result in the game crashing, Windows freezing or visual artifacts (white dots, coloured squares, stretched polygons, etc). A loose power connector would most likely cause a BSOD, freeze your PC (certainly when trying to run a game) or even prevent it booting. The black screen + 100% fan issue is a result of a hardware issue, confirmed earlier in the thread by EVGA, which IMO is due to overheating components on the PCB that are not being properly cooled. Overclocking the VRAM could also contribute to the black screen issue as well since it is one of the components reported by a few sites to get hot.
There are reports of black screen issues on the NVIDIA forum with their cards (not just the 1000 series) but these are not accompanied by the fans ramping up to 100%.