Oh, so it's only a dollar hold? That tempts me again to do the cross ship RMA even though doing the thermal pad replacement isn't complicated and it'd let me put some better thermal paste on the GPU. Today/Friday is day 30 since I received my 1070 FTW so I imagine they'd still be cool with sending a new one instead of a refurb. Only thing is it'd suck to get a replacement that turns out to be worse off in the "silicon lottery" than the one I currently have which wasn't a big winner in that regard in the first place.
Guess I should've made a topic on it since I posted updates on this in the 1060, 1070 and 1080 review topics a couple days ago. ¯\_(ツ
_/¯
This is the card I have. Let's hope I'm alright.
But how do I even go about telling if I have a problem? My temps have never exceeded 78 degrees during gameplay. I've seen the card idle to as low as 28.
Unless you have a thermal imaging device you likely won't be able to really tell since there's no temperature sensor on the VRM. You really should apply the pads or RMA.
I didn't even know this was a thing until I found an offhand remark about it on Reddit earlier today. I've never had any heat problems with my card, but I have been using a custom fan curve since day one, so it's always run pretty cool. I might do the BIOS update at some point, but if I'm using a custom fan curve it probably won't affect anything anyway.
The thing is you can't really tell if there's a problem until it happens unless you're looking for it and the higher fan curve is just a band aid fix.
I have a evga 1070 sc, I thought only the 1080 ftw's had issues? I'll download the update, but would it be a better idea for me to RMA my card? Don't want any chance at all of my computer catching fire.
All of the cards with the ACX 3.0 cooler is having this issue because in someone's brilliance they decided to not add thermal pads while every other board partner and the Founders Editions cards had them.
EVGA is easily one of the best brands on the market. This is the first time I can remember something like this happening with them and they are handling it very well.
Two years ago their initial run of 970s with the ACX 2.0 cooler had an issues as well with heatpipes not making proper contact with the GPU die.
Hmm. Did the memory issue ever get fixed? I'm guessing I'll be giving up my Samsung RAM and getting a card with Micron.
The BIOS update for that came out a couple weeks ago and did fix the issue but while it did improve overclocking (at least it did in my case allowing me to go from a max of 8500MHz to 8900MHz) it's still not as good as the Samsung memory.
That then goes to the further issue of this VRM mess, it being so close to the memory can lead to problems as well if the VRM gets hot enough to increase memory temps above its spec which with how heat can apparently get insulated by the backplate is a likely possibility. THEN there's an issue in some cases where the thermal pads for the VRAM aren't actually making contact with the memory and are in turn useless.
EVGA dropped the ball. I have one EVGA 1080 too. But this video is fucked. He blocked one of the two fans and doesn't connect power cables to the gpu increasing the current from the pci.
The fan is "blocked" by a PCI card, another video card in an SLI set up would do more to hinder airflow than that thing and even without the connectors plugged in that should not happen by just going off the PCIE slot power, at worst what should happen is either the computer not starting with a beep or/and an error message popping up on the display about the cables not being plugged in or the GPU being severely down clocked, not having it plugged in would lead to card shooting off a flare. The VRM was already fucked when his display "black screen'd", had the power connectors been plugged in this still would've happened.
