EVGA FTW 1080 and 1070 Have Overheating Issues
Click here to post a comment for EVGA FTW 1080 and 1070 Have Overheating Issues on our message forum
BarryB
Maybe SC means SuperCOOKED?
EdInk
Will installing thermal pads void the warranty?
Agent-A01
maize1951
Simple fix, remove heatsink and put Artic Silver heatsink compound on GPU and reinstall heatsink.
Agent-A01
Rassel
Darren Hodgson
I pre-ordered an EVGA GTX 1080 FTW back in June from Amazon although it was almost two months before it was finally delivered mid-August due to severe stock shortages here in the UK.
Apart from a couple of black screens initially, which I put down at the time to over-ambitious overclocking attempts, the card continued to work for six weeks with no real issues. Then about two weeks ago, coincidentally at the same time as I installed the v373.06 driver, I started experiencing numerous black screens across a number of games; sometimes within 5 minutes of playing and other times after several hours or none at all. They could occur during gameplay or during pre-rendered cutscenes. At first I thought it might be a driver issue or even a bug in Windows 10 v1607 (I mean it has plenty, right?) but then I found a 40+ page thread on the EVGA forum from people experiencing the same issue. When the black screen occurs it is accompanied by the fans ramping up to 100%. The only thing I could do was restart my PC.
I RMAed the card early last week and received my replacement on Friday. I had to use the Advanced RMA option otherwise I would have had to return my old card first before they would have shipped out a new one, which might have taken 1-2 weeks. As I did not have a spare graphics card to use I ended up having to pay an obnoxious collateral payment (refundable) of £722, which is £122 more than I paid for my card in the first place. As far as I can tell, EVGA have yet to post an official statement on the issue on their forum as a sticky and instead any information has to be found by wading through (currently) 45 pages of posts. It's ridiculous really.
I have been using my card all weekend and so far, touch wood, it seems to be working fine (I hope I haven't tempted fate saying that...!!!). My previous card (now on its way back to EVGA so I can get my money back) would not overclock by more than +250 MHz on the VRAM but this new card seems to have no problems with +400 MHz. I suspect the issue may be have been due to the VRMs and VRAM overheating as at no point did the card's core overheat. I saw it heat 82 C during summer but with the default fan profile (which only spins at over 65 C) the temperatures are typically between 65 and 75 C with a fan speed of 30-40%. However, I was using a custom fan profile for the first two months of owning the card and it's telling that the black screen issue started happening shortly after I switched to the default profile. While I waited for my replacement to arrive I switched back to the custom fan profile and did not see the black screen issue again for three days before I swapped the card for the new one.
I assume that my new card has the thermal pads fitted (at least I would hope so) but if this card develops the same fault again in two months then I will return the card to Amazon and ask for a full refund. I guess I've been lucky because I have been buying graphics card since 1998 and this is the first "bad" one that I've had.
Darren Hodgson
P.S. I am really glad that websites such as Guru3D are reporting on this issue as it feels it needs to be out in the open if only to get EVGA to make an official statement on their forum and sticky it.
They should also be replacing affected cards via the Advanced RMA option at NO extra cost. Asking their customers to lay out large amounts as collateral is just damn cheeky IMO, especially as the issue is down to a design flaw in the cards themselves that is no fault of the customer. My card was faulty from the day I bought it; it just took two months for the issue to make itself known.
Noisiv
Legacy-ZA
I actually waited so long for a GTX 1070 FTW review from Guru3D but it never arrived. I saw the SC review and I didn't like that hot spot one bit, but I figured since the VRM area is custom and the cooler looks different on the FTW, being more beefy that is, it wouldn't have the issue, I guess I was wrong. *sigh*
I ordered my thermal pads... let's see if it brings those temps down.
ifixit
lucidus
ifixit
er_wendigo
That's non sense, the cards (as example SC one) that the people are commenting that they "have" without thermalpads on VRMs in fact they have them:
Go to the techpowerup review of the SC card and see by yourself that the card has thermalpads between the VRMs and the backplate (I can't post any link yet).
About the backplate reaching "tremendous" temperatures sooooo high as 70ºC os a little more, well:
If you don't have any thermalpad conecting the VRMs with the backplate, that backplate is USELESS as a thermal buffer, and only serves as a heat trap. If you see that backplate as "cold", then you have a real issue (heat trap), if don't and the backplate is warming, it's working properly. That's much better than without any heatsink, and posibly better than with little and dedicated heatsink to the VRMs.
And don't forget, many graphic cards are without any heatsink/backplate helping to extract heat of VRMs, and works flawlessly.
That's looks as an assembling issue with some FTWs and the lack of these thermalpads associated with VRMs and the backplate as SC. I don't see that the FTW series don't have that as "normal" while the SC one and others ACX cards have their thermalpads.
PD: Ok, I see that the thermalpads are only with the "frontplate" of the card, but that should be more than enough to cool the VRMs, the difference with connecting or not the backplate to the VRMs, when they are connected to the frontplate (that receives fresh air of the fans) should be negligible.
If you have certain mass of aluminium and that is receiving air (convection) the thermal performance should be more than enough, the backplate only contributes minimally.
Cards like the Direct CU II cards of Asus had backplates (at least the kepler ones that I saw) without thermalpads too, hell, the vast majority of cards with backplates have that only for aesthetical reasons.
ruthan
Dont worry, JayZ will recommend it to his fans.
ifixit
Reddoguk
The back-plate is not there to aid in cooling, it's only there for protecting components and to help make the PCB more rigid.
I doubt very much it would behave like a heat-sink unless i'm wrong and it's metal.
The back-plate on my 980G1 looks like some sort of plastic or carbon like material like a mobo is made of.
I love my 980G1 back plate as it's done right unlike some i see with no cut outs, the G1 has like 40% of it cut away for heat dissipation.
EVGA shame on you. Why on earth on a top of the line product like the FTW series would you leave out heat pads especially for overclocked cards.
EVGA = Within tolerance so it's fine, err but it's not fine is it or people wouldn't be having these simply fixed errors would they.
I know they'll make it right but this kind of behavior is off-putting to say the least.
Back plates are in fact brush aluminum but still doubt it's effectiveness in heat dissipation.
TheDeeGee
fry178
@holler
not sure what a design flaw (never planned for thermal pad) and QC issues (faulty cards without pads were shipped) has to do with nv and/or their lifecycle for chips.
stevevnicks
only card's NV need to worry about will be manufactured by NV any others like EVGA, MSI, ASUS etc are their own design all they do is buy the gpu from NVidia and make their own custom boards/lay out but governed by NVidia with volt limit's for the GPU.
so it's down to EVGA's custom board design/lay out and their own spec's based on the spec's of the VRM they use or any other IC's they use, this is down to EVGA no if's or but's, their choice not to use any extra thermal protection for there VRM design.